fix(codeql): fix high-severity CodeQL issues

- Replace MD5 with SHA-256 for finding ID generation (4 locations) - Use crypto.randomBytes() instead of Math.random() for temp files (2 files) - Fix TOCTOU race conditions by using try/catch instead of existsSync (5 files) - Add __dir__() to lazy import module for static analysis Fixes the following CodeQL alerts: - Use of broken/weak cryptographic hashing algorithm - Insecure temporary file creation - Potential file system race conditions (TOCTOU) - Explicit export not defined in __all__ All 3841 frontend tests pass.
test: fix issue-create-handler test mock
2026-02-18 08:30:55 +01:00 · 2026-02-18 07:53:38 +01:00 · 2026-02-18 07:42:46 +01:00 · 2026-02-18 07:32:45 +01:00 · 2026-02-17 22:16:33 +01:00 · 2026-02-17 19:31:25 +01:00
2159 changed files with 274123 additions and 102008 deletions
@@ -42,14 +42,18 @@ reviews:

  # Path-specific review instructions
  path_instructions:
-    - path: "apps/desktop/**/*.{ts,tsx}"
+    - path: "apps/backend/**/*.py"
+      instructions: |
+        Focus on Python best practices, type hints, and async patterns.
+        Check for proper error handling and security considerations.
+        Verify compatibility with Python 3.12+.
+    - path: "apps/frontend/**/*.{ts,tsx}"
      instructions: |
        Review React patterns and TypeScript type safety.
        Check for proper state management and component composition.
-        Verify Vercel AI SDK v6 usage patterns and tool definitions.
-    - path: "apps/desktop/**/*.test.{ts,tsx}"
+    - path: "tests/**"
      instructions: |
-        Ensure tests are comprehensive and follow Vitest conventions.
+        Ensure tests are comprehensive and follow pytest conventions.
        Check for proper mocking and test isolation.

 chat:
@@ -41,7 +41,7 @@ runs:
      shell: bash
      # Run npm ci from root to properly handle workspace dependencies.
      # With npm workspaces, the lock file is at root and dependencies are hoisted there.
-      # Running npm ci in apps/desktop would fail to populate node_modules correctly.
+      # Running npm ci in apps/frontend would fail to populate node_modules correctly.
      run: |
        if [ "${{ inputs.ignore-scripts }}" == "true" ]; then
          npm ci --ignore-scripts
@@ -51,12 +51,12 @@ runs:

    - name: Link node_modules for electron-builder
      shell: bash
-      # electron-builder expects node_modules in apps/desktop for native module rebuilding.
+      # electron-builder expects node_modules in apps/frontend for native module rebuilding.
      # With npm workspaces, packages are hoisted to root. Create a link so electron-builder
      # can find the modules during packaging and code signing.
      # Uses symlink on Unix, directory junction on Windows (works without admin privileges).
      #
-      # IMPORTANT: npm workspaces may create a partial node_modules in apps/desktop for
+      # IMPORTANT: npm workspaces may create a partial node_modules in apps/frontend for
      # packages that couldn't be hoisted. We must remove it and create a proper link to root.
      run: |
        # Verify npm ci succeeded
@@ -65,42 +65,42 @@ runs:
          exit 1
        fi

-        # Remove any existing node_modules in apps/desktop
+        # Remove any existing node_modules in apps/frontend
        # This handles: partial directories from npm workspaces, AND broken symlinks
-        if [ -e "apps/desktop/node_modules" ] || [ -L "apps/desktop/node_modules" ]; then
+        if [ -e "apps/frontend/node_modules" ] || [ -L "apps/frontend/node_modules" ]; then
          # Check if it's a valid symlink pointing to root node_modules
-          if [ -L "apps/desktop/node_modules" ]; then
-            target=$(readlink apps/desktop/node_modules 2>/dev/null || echo "")
-            if [ "$target" = "../../node_modules" ] && [ -d "apps/desktop/node_modules" ]; then
-              echo "Correct symlink already exists: apps/desktop/node_modules -> ../../node_modules"
+          if [ -L "apps/frontend/node_modules" ]; then
+            target=$(readlink apps/frontend/node_modules 2>/dev/null || echo "")
+            if [ "$target" = "../../node_modules" ] && [ -d "apps/frontend/node_modules" ]; then
+              echo "Correct symlink already exists: apps/frontend/node_modules -> ../../node_modules"
            else
              echo "Removing incorrect/broken symlink (was: $target)..."
-              rm -f "apps/desktop/node_modules"
+              rm -f "apps/frontend/node_modules"
            fi
          else
            echo "Removing partial node_modules directory created by npm workspaces..."
-            rm -rf "apps/desktop/node_modules"
+            rm -rf "apps/frontend/node_modules"
          fi
        fi

        # Create link if it doesn't exist or was removed
-        if [ ! -L "apps/desktop/node_modules" ]; then
+        if [ ! -L "apps/frontend/node_modules" ]; then
          if [ "$RUNNER_OS" == "Windows" ]; then
            # Use directory junction on Windows (works without admin privileges)
            # Use PowerShell's New-Item -ItemType Junction for reliable path handling
            abs_target=$(cygpath -w "$(pwd)/node_modules")
-            link_path=$(cygpath -w "$(pwd)/apps/desktop/node_modules")
+            link_path=$(cygpath -w "$(pwd)/apps/frontend/node_modules")
            powershell -Command "New-Item -ItemType Junction -Path '$link_path' -Target '$abs_target'" > /dev/null
            if [ $? -eq 0 ]; then
-              echo "Created junction: apps/desktop/node_modules -> $abs_target"
+              echo "Created junction: apps/frontend/node_modules -> $abs_target"
            else
              echo "::error::Failed to create directory junction on Windows"
              exit 1
            fi
          else
            # Use symlink on Unix (macOS/Linux)
-            if ln -s ../../node_modules apps/desktop/node_modules; then
-              echo "Created symlink: apps/desktop/node_modules -> ../../node_modules"
+            if ln -s ../../node_modules apps/frontend/node_modules; then
+              echo "Created symlink: apps/frontend/node_modules -> ../../node_modules"
            else
              echo "::error::Failed to create symlink"
              exit 1
@@ -111,16 +111,16 @@ runs:
        # Final verification - the link must exist and resolve correctly
        # Note: On Windows, junctions don't show as symlinks (-L), so we check if the directory exists
        # and can be listed. On Unix, we also verify it's a symlink.
-        if [ "$RUNNER_OS" != "Windows" ] && [ ! -L "apps/desktop/node_modules" ]; then
-          echo "::error::apps/desktop/node_modules symlink was not created"
+        if [ "$RUNNER_OS" != "Windows" ] && [ ! -L "apps/frontend/node_modules" ]; then
+          echo "::error::apps/frontend/node_modules symlink was not created"
          exit 1
        fi
        # Verify the link resolves to a valid directory with content
-        if ! ls apps/desktop/node_modules/electron >/dev/null 2>&1; then
-          echo "::error::apps/desktop/node_modules does not resolve correctly (electron not found)"
-          ls -la apps/desktop/ || true
-          ls apps/desktop/node_modules 2>&1 | head -5 || true
+        if ! ls apps/frontend/node_modules/electron >/dev/null 2>&1; then
+          echo "::error::apps/frontend/node_modules does not resolve correctly (electron not found)"
+          ls -la apps/frontend/ || true
+          ls apps/frontend/node_modules 2>&1 | head -5 || true
          exit 1
        fi
-        count=$(ls apps/desktop/node_modules 2>/dev/null | wc -l)
-        echo "Verified: apps/desktop/node_modules resolves correctly ($count entries)"
+        count=$(ls apps/frontend/node_modules 2>/dev/null | wc -l)
+        echo "Verified: apps/frontend/node_modules resolves correctly ($count entries)"
@@ -0,0 +1,52 @@
+name: 'Setup Python Backend'
+description: 'Set up Python with uv package manager and cached dependencies for the backend'
+
+inputs:
+  python-version:
+    description: 'Python version to use'
+    required: false
+    default: '3.12'
+  install-test-deps:
+    description: 'Whether to install test dependencies'
+    required: false
+    default: 'false'
+
+outputs:
+  cache-hit:
+    description: 'Whether cache was hit'
+    value: ${{ steps.cache.outputs.cache-hit }}
+
+runs:
+  using: 'composite'
+  steps:
+    - name: Set up Python ${{ inputs.python-version }}
+      uses: actions/setup-python@v5
+      with:
+        python-version: ${{ inputs.python-version }}
+
+    - name: Install uv package manager
+      uses: astral-sh/setup-uv@v4
+      with:
+        version: "latest"
+
+    - name: Cache uv dependencies
+      id: cache
+      uses: actions/cache@v4
+      with:
+        path: |
+          ~/.cache/uv
+          ~/AppData/Local/uv/cache
+          ~/Library/Caches/uv
+        key: uv-${{ runner.os }}-${{ runner.arch }}-${{ inputs.python-version }}-${{ hashFiles('apps/backend/requirements.txt', 'tests/requirements-test.txt') }}
+        restore-keys: |
+          uv-${{ runner.os }}-${{ runner.arch }}-${{ inputs.python-version }}-
+
+    - name: Install dependencies
+      working-directory: apps/backend
+      shell: bash
+      run: |
+        uv venv
+        uv pip install -r requirements.txt
+        if [ "${{ inputs.install-test-deps }}" == "true" ]; then
+          uv pip install -r ../../tests/requirements-test.txt
+        fi
@@ -14,7 +14,7 @@ inputs:
  dmg-path:
    description: 'Path to the dist directory containing the DMG file'
    required: false
-    default: 'apps/desktop/dist'
+    default: 'apps/frontend/dist'

 outputs:
  notarization-id:
@@ -1,8 +1,20 @@
 version: 2
 updates:
+  # Python dependencies
+  - package-ecosystem: pip
+    directory: /apps/backend
+    schedule:
+      interval: weekly
+    open-pull-requests-limit: 5
+    labels:
+      - dependencies
+      - python
+    commit-message:
+      prefix: "chore(deps)"
+
  # npm dependencies
  - package-ecosystem: npm
-    directory: /apps/desktop
+    directory: /apps/frontend
    schedule:
      interval: weekly
    open-pull-requests-limit: 5
@@ -74,11 +74,35 @@ jobs:
          # Use tag for real releases, develop branch for dry runs
          ref: ${{ github.event.inputs.dry_run == 'true' && 'develop' || format('v{0}', needs.create-tag.outputs.version) }}

+      - name: Setup Python
+        uses: actions/setup-python@v6
+        with:
+          python-version: '3.11'
+
      - name: Setup Node.js and install dependencies
        uses: ./.github/actions/setup-node-frontend

+      - name: Install Rust toolchain (for building native Python packages)
+        uses: dtolnay/rust-toolchain@stable
+
+      - name: Cache pip wheel cache (for compiled packages like real_ladybug)
+        uses: actions/cache@v5
+        with:
+          path: ~/Library/Caches/pip
+          key: pip-wheel-${{ runner.os }}-x64-${{ hashFiles('apps/backend/requirements.txt') }}
+          restore-keys: |
+            pip-wheel-${{ runner.os }}-x64-
+
+      - name: Cache bundled Python
+        uses: actions/cache@v5
+        with:
+          path: apps/frontend/python-runtime
+          key: python-bundle-${{ runner.os }}-x64-3.12.8-rust-${{ hashFiles('apps/backend/requirements.txt') }}
+          restore-keys: |
+            python-bundle-${{ runner.os }}-x64-3.12.8-rust-
+
      - name: Build application
-        run: cd apps/desktop && npm run build
+        run: cd apps/frontend && npm run build
        env:
          SENTRY_DSN: ${{ secrets.SENTRY_DSN }}
          SENTRY_TRACES_SAMPLE_RATE: ${{ secrets.SENTRY_TRACES_SAMPLE_RATE }}
@@ -87,7 +111,7 @@ jobs:
      - name: Package macOS (Intel)
        run: |
          VERSION="${{ needs.create-tag.outputs.version }}"
-          cd apps/desktop && npm run package:mac -- --x64 --config.extraMetadata.version="$VERSION"
+          cd apps/frontend && npm run package:mac -- --x64 --config.extraMetadata.version="$VERSION"
        env:
          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
          CSC_LINK: ${{ secrets.MAC_CERTIFICATE }}
@@ -109,9 +133,9 @@ jobs:
        with:
          name: macos-intel-builds
          path: |
-            apps/desktop/dist/*.dmg
-            apps/desktop/dist/*.zip
-            apps/desktop/dist/*.yml
+            apps/frontend/dist/*.dmg
+            apps/frontend/dist/*.zip
+            apps/frontend/dist/*.yml

  # Apple Silicon build on ARM64 runner for native compilation
  build-macos-arm64:
@@ -126,11 +150,32 @@ jobs:
          # Use tag for real releases, develop branch for dry runs
          ref: ${{ github.event.inputs.dry_run == 'true' && 'develop' || format('v{0}', needs.create-tag.outputs.version) }}

+      - name: Setup Python
+        uses: actions/setup-python@v6
+        with:
+          python-version: '3.11'
+
      - name: Setup Node.js and install dependencies
        uses: ./.github/actions/setup-node-frontend

+      - name: Cache pip wheel cache
+        uses: actions/cache@v5
+        with:
+          path: ~/Library/Caches/pip
+          key: pip-wheel-${{ runner.os }}-arm64-${{ hashFiles('apps/backend/requirements.txt') }}
+          restore-keys: |
+            pip-wheel-${{ runner.os }}-arm64-
+
+      - name: Cache bundled Python
+        uses: actions/cache@v5
+        with:
+          path: apps/frontend/python-runtime
+          key: python-bundle-${{ runner.os }}-arm64-3.12.8-${{ hashFiles('apps/backend/requirements.txt') }}
+          restore-keys: |
+            python-bundle-${{ runner.os }}-arm64-3.12.8-
+
      - name: Build application
-        run: cd apps/desktop && npm run build
+        run: cd apps/frontend && npm run build
        env:
          SENTRY_DSN: ${{ secrets.SENTRY_DSN }}
          SENTRY_TRACES_SAMPLE_RATE: ${{ secrets.SENTRY_TRACES_SAMPLE_RATE }}
@@ -139,7 +184,7 @@ jobs:
      - name: Package macOS (Apple Silicon)
        run: |
          VERSION="${{ needs.create-tag.outputs.version }}"
-          cd apps/desktop && npm run package:mac -- --arm64 --config.extraMetadata.version="$VERSION"
+          cd apps/frontend && npm run package:mac -- --arm64 --config.extraMetadata.version="$VERSION"
        env:
          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
          CSC_LINK: ${{ secrets.MAC_CERTIFICATE }}
@@ -161,9 +206,9 @@ jobs:
        with:
          name: macos-arm64-builds
          path: |
-            apps/desktop/dist/*.dmg
-            apps/desktop/dist/*.zip
-            apps/desktop/dist/*.yml
+            apps/frontend/dist/*.dmg
+            apps/frontend/dist/*.zip
+            apps/frontend/dist/*.yml

  build-windows:
    needs: create-tag
@@ -180,11 +225,32 @@ jobs:
          # Use tag for real releases, develop branch for dry runs
          ref: ${{ github.event.inputs.dry_run == 'true' && 'develop' || format('v{0}', needs.create-tag.outputs.version) }}

+      - name: Setup Python
+        uses: actions/setup-python@v6
+        with:
+          python-version: '3.11'
+
      - name: Setup Node.js and install dependencies
        uses: ./.github/actions/setup-node-frontend

+      - name: Cache pip wheel cache
+        uses: actions/cache@v5
+        with:
+          path: ~\AppData\Local\pip\Cache
+          key: pip-wheel-${{ runner.os }}-x64-${{ hashFiles('apps/backend/requirements.txt') }}
+          restore-keys: |
+            pip-wheel-${{ runner.os }}-x64-
+
+      - name: Cache bundled Python
+        uses: actions/cache@v5
+        with:
+          path: apps/frontend/python-runtime
+          key: python-bundle-${{ runner.os }}-x64-3.12.8-${{ hashFiles('apps/backend/requirements.txt') }}
+          restore-keys: |
+            python-bundle-${{ runner.os }}-x64-3.12.8-
+
      - name: Build application
-        run: cd apps/desktop && npm run build
+        run: cd apps/frontend && npm run build
        env:
          SENTRY_DSN: ${{ secrets.SENTRY_DSN }}
          SENTRY_TRACES_SAMPLE_RATE: ${{ secrets.SENTRY_TRACES_SAMPLE_RATE }}
@@ -194,7 +260,7 @@ jobs:
        shell: bash
        run: |
          VERSION="${{ needs.create-tag.outputs.version }}"
-          cd apps/desktop && npm run package:win -- --config.extraMetadata.version="$VERSION"
+          cd apps/frontend && npm run package:win -- --config.extraMetadata.version="$VERSION"
        env:
          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
          # Disable electron-builder's built-in signing (we use Azure Trusted Signing instead)
@@ -218,7 +284,7 @@ jobs:
          endpoint: https://neu.codesigning.azure.net/
          trusted-signing-account-name: ${{ secrets.AZURE_SIGNING_ACCOUNT }}
          certificate-profile-name: ${{ secrets.AZURE_CERTIFICATE_PROFILE }}
-          files-folder: apps/desktop/dist
+          files-folder: apps/frontend/dist
          files-folder-filter: exe
          file-digest: SHA256
          timestamp-rfc3161: http://timestamp.acs.microsoft.com
@@ -228,7 +294,7 @@ jobs:
        if: env.AZURE_CLIENT_ID != ''
        shell: pwsh
        run: |
-          cd apps/desktop/dist
+          cd apps/frontend/dist
          $exeFile = Get-ChildItem -Filter "*.exe" | Select-Object -First 1
          if ($exeFile) {
            Write-Host "Verifying signature on $($exeFile.Name)..."
@@ -252,7 +318,7 @@ jobs:
        shell: pwsh
        run: |
          $ErrorActionPreference = "Stop"
-          cd apps/desktop/dist
+          cd apps/frontend/dist

          # Find the installer exe (electron-builder names it with "Setup" or just the app name)
          # electron-builder produces one installer exe per build
@@ -319,8 +385,8 @@ jobs:
        with:
          name: windows-builds
          path: |
-            apps/desktop/dist/*.exe
-            apps/desktop/dist/*.yml
+            apps/frontend/dist/*.exe
+            apps/frontend/dist/*.yml

  build-linux:
    needs: create-tag
@@ -331,6 +397,11 @@ jobs:
          # Use tag for real releases, develop branch for dry runs
          ref: ${{ github.event.inputs.dry_run == 'true' && 'develop' || format('v{0}', needs.create-tag.outputs.version) }}

+      - name: Setup Python
+        uses: actions/setup-python@v6
+        with:
+          python-version: '3.11'
+
      - name: Setup Node.js and install dependencies
        uses: ./.github/actions/setup-node-frontend

@@ -338,13 +409,29 @@ jobs:
        run: |
          set -e
          sudo apt-get update
-          sudo apt-get install -y flatpak flatpak-builder squashfs-tools
+          sudo apt-get install -y flatpak flatpak-builder libarchive-tools
          flatpak remote-add --user --if-not-exists flathub https://flathub.org/repo/flathub.flatpakrepo
          flatpak install -y --user flathub org.freedesktop.Platform//25.08 org.freedesktop.Sdk//25.08
          flatpak install -y --user flathub org.electronjs.Electron2.BaseApp//25.08

+      - name: Cache pip wheel cache
+        uses: actions/cache@v5
+        with:
+          path: ~/.cache/pip
+          key: pip-wheel-${{ runner.os }}-x64-${{ hashFiles('apps/backend/requirements.txt') }}
+          restore-keys: |
+            pip-wheel-${{ runner.os }}-x64-
+
+      - name: Cache bundled Python
+        uses: actions/cache@v5
+        with:
+          path: apps/frontend/python-runtime
+          key: python-bundle-${{ runner.os }}-x64-3.12.8-${{ hashFiles('apps/backend/requirements.txt') }}
+          restore-keys: |
+            python-bundle-${{ runner.os }}-x64-3.12.8-
+
      - name: Build application
-        run: cd apps/desktop && npm run build
+        run: cd apps/frontend && npm run build
        env:
          SENTRY_DSN: ${{ secrets.SENTRY_DSN }}
          SENTRY_TRACES_SAMPLE_RATE: ${{ secrets.SENTRY_TRACES_SAMPLE_RATE }}
@@ -353,7 +440,7 @@ jobs:
      - name: Package Linux
        run: |
          VERSION="${{ needs.create-tag.outputs.version }}"
-          cd apps/desktop && npm run package:linux -- --config.extraMetadata.version="$VERSION"
+          cd apps/frontend && npm run package:linux -- --config.extraMetadata.version="$VERSION"
        env:
          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
          SENTRY_DSN: ${{ secrets.SENTRY_DSN }}
@@ -361,17 +448,17 @@ jobs:
          SENTRY_PROFILES_SAMPLE_RATE: ${{ secrets.SENTRY_PROFILES_SAMPLE_RATE }}

      - name: Verify Linux packages
-        run: cd apps/desktop && npm run verify:linux
+        run: cd apps/frontend && npm run verify:linux

      - name: Upload artifacts
        uses: actions/upload-artifact@v4
        with:
          name: linux-builds
          path: |
-            apps/desktop/dist/*.AppImage
-            apps/desktop/dist/*.deb
-            apps/desktop/dist/*.flatpak
-            apps/desktop/dist/*.yml
+            apps/frontend/dist/*.AppImage
+            apps/frontend/dist/*.deb
+            apps/frontend/dist/*.flatpak
+            apps/frontend/dist/*.yml

  # Finalize macOS notarization (runs in parallel with Windows/Linux builds)
  finalize-notarization:
@@ -38,7 +38,7 @@ jobs:
        uses: microsoft/setup-msbuild@v2

      - name: Install node-pty and rebuild for Electron
-        working-directory: apps/desktop
+        working-directory: apps/frontend
        shell: pwsh
        run: |
          # Install only node-pty
@@ -52,7 +52,7 @@ jobs:
          npx @electron/rebuild --version $env:ELECTRON_VERSION --module-dir node_modules/node-pty --arch ${{ matrix.arch }}

      - name: Package prebuilt binaries
-        working-directory: apps/desktop
+        working-directory: apps/frontend
        shell: pwsh
        run: |
          $electronAbi = (npx electron-abi $env:ELECTRON_VERSION)
@@ -78,7 +78,7 @@ jobs:
          Get-ChildItem $prebuildDir

      - name: Create archive
-        working-directory: apps/desktop
+        working-directory: apps/frontend
        shell: pwsh
        run: |
          $electronAbi = (npx electron-abi $env:ELECTRON_VERSION)
@@ -93,14 +93,14 @@ jobs:
        uses: actions/upload-artifact@v4
        with:
          name: node-pty-win32-${{ matrix.arch }}
-          path: apps/desktop/node-pty-*.zip
+          path: apps/frontend/node-pty-*.zip
          retention-days: 90

      - name: Upload to release
        if: github.event_name == 'release'
        uses: softprops/action-gh-release@v1
        with:
-          files: apps/desktop/node-pty-*.zip
+          files: apps/frontend/node-pty-*.zip
        env:
          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}

@@ -3,7 +3,8 @@
 # Tests on all target platforms (Linux, Windows, macOS) to catch
 # platform-specific bugs before they merge. ALL platforms must pass.
 #
-# Optimized: Frontend-only matrix, path filters to skip on docs-only changes.
+# Optimized: Reduced matrix (4 jobs vs 6), merged integration tests,
+# coverage on Linux only, path filters to skip on docs-only changes.

 name: CI

@@ -12,7 +13,10 @@ on:
    branches: [main, develop]
    paths:
      - 'apps/**'
+      - 'tests/**'
      - 'package*.json'
+      - 'requirements*.txt'
+      - 'pyproject.toml'
      - 'tsconfig*.json'
      - 'biome.jsonc'
      - '.github/workflows/ci.yml'
@@ -21,7 +25,10 @@ on:
    branches: [main, develop]
    paths:
      - 'apps/**'
+      - 'tests/**'
      - 'package*.json'
+      - 'requirements*.txt'
+      - 'pyproject.toml'
      - 'tsconfig*.json'
      - 'biome.jsonc'
      - '.github/workflows/ci.yml'
@@ -36,6 +43,70 @@ permissions:
  actions: read

 jobs:
+  # --------------------------------------------------------------------------
+  # Python Backend Tests - Optimized Matrix (4 jobs instead of 6)
+  # --------------------------------------------------------------------------
+  test-python:
+    name: test-python (${{ matrix.python-version }}, ${{ matrix.os }})
+    runs-on: ${{ matrix.os }}
+
+    strategy:
+      fail-fast: false
+      matrix:
+        # 3.12 on all OS for cross-platform coverage
+        # 3.13 on Linux only for compatibility check (saves 2 jobs)
+        include:
+          - os: ubuntu-latest
+            python-version: '3.12'
+          - os: ubuntu-latest
+            python-version: '3.13'
+          - os: windows-latest
+            python-version: '3.12'
+          - os: macos-latest
+            python-version: '3.12'
+
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@v4
+
+      - name: Setup Python backend
+        uses: ./.github/actions/setup-python-backend
+        with:
+          python-version: ${{ matrix.python-version }}
+          install-test-deps: 'true'
+
+      - name: Run all tests (including platform-specific)
+        working-directory: apps/backend
+        shell: bash
+        env:
+          PYTHONPATH: ${{ github.workspace }}/apps/backend
+        run: |
+          if [ "$RUNNER_OS" == "Windows" ]; then
+            source .venv/Scripts/activate
+          else
+            source .venv/bin/activate
+          fi
+          pytest ../../tests/ -v --tb=short -x
+
+      - name: Run coverage (Linux + Python 3.12 only)
+        if: matrix.os == 'ubuntu-latest' && matrix.python-version == '3.12'
+        working-directory: apps/backend
+        shell: bash
+        env:
+          PYTHONPATH: ${{ github.workspace }}/apps/backend
+        run: |
+          source .venv/bin/activate
+          pytest ../../tests/ -v --cov=. --cov-report=xml --cov-report=term-missing --cov-fail-under=10
+
+      - name: Upload coverage to Codecov
+        if: matrix.os == 'ubuntu-latest' && matrix.python-version == '3.12'
+        uses: codecov/codecov-action@v4
+        with:
+          file: ./apps/backend/coverage.xml
+          fail_ci_if_error: false
+        env:
+          CODECOV_TOKEN: ${{ secrets.CODECOV_TOKEN }}
+
  # --------------------------------------------------------------------------
  # Frontend Tests - All Platforms
  # --------------------------------------------------------------------------
@@ -58,15 +129,15 @@ jobs:
          ignore-scripts: 'true'

      - name: Run TypeScript type check
-        working-directory: apps/desktop
+        working-directory: apps/frontend
        run: npm run typecheck

      - name: Run unit tests
-        working-directory: apps/desktop
+        working-directory: apps/frontend
        run: npm run test

      - name: Build application
-        working-directory: apps/desktop
+        working-directory: apps/frontend
        run: npm run build

  # --------------------------------------------------------------------------
@@ -75,16 +146,18 @@ jobs:
  ci-complete:
    name: CI Complete
    runs-on: ubuntu-latest
-    needs: [test-frontend]
+    needs: [test-python, test-frontend]
    if: always()
    steps:
      - name: Check all CI jobs passed
        run: |
          echo "CI Job Results:"
+          echo "  test-python:   ${{ needs.test-python.result }}"
          echo "  test-frontend: ${{ needs.test-frontend.result }}"
          echo ""

-          if [[ "${{ needs.test-frontend.result }}" != "success" ]]; then
+          if [[ "${{ needs.test-python.result }}" != "success" ]] || \
+             [[ "${{ needs.test-frontend.result }}" != "success" ]]; then
            echo "❌ One or more CI jobs failed"
            exit 1
          fi
@@ -3,7 +3,6 @@ name: Discord Release Notification
 on:
  release:
    types: [published]
-  workflow_dispatch:

 jobs:
  discord-notification:
@@ -4,23 +4,50 @@ on:
  push:
    branches: [main, develop]
    paths:
-      - 'apps/desktop/**'
+      - 'apps/**'
+      - 'tests/**'
      - '.github/workflows/lint.yml'
      - '.github/actions/**'
-      - 'apps/desktop/biome.jsonc'
+      - 'apps/frontend/biome.jsonc'
+      - '.pre-commit-config.yaml'
  pull_request:
    branches: [main, develop]
    paths:
-      - 'apps/desktop/**'
+      - 'apps/**'
+      - 'tests/**'
      - '.github/workflows/lint.yml'
      - '.github/actions/**'
-      - 'apps/desktop/biome.jsonc'
+      - 'apps/frontend/biome.jsonc'
+      - '.pre-commit-config.yaml'

 concurrency:
  group: lint-${{ github.event.pull_request.number || github.ref }}
  cancel-in-progress: true

 jobs:
+  # Python linting (Ruff) - already fast, no changes needed
+  python:
+    name: Python (Ruff)
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v4
+
+      - name: Set up Python
+        uses: actions/setup-python@v6
+        with:
+          python-version: '3.12'
+
+      # Pin ruff version to match .pre-commit-config.yaml
+      - name: Install ruff
+        run: pip install ruff==0.14.10
+
+      - name: Run ruff check
+        run: ruff check apps/backend/ --output-format=github
+
+      - name: Run ruff format check
+        run: ruff format apps/backend/ --check --diff
+
  # TypeScript/JavaScript linting (Biome) - 15-25x faster than ESLint
  typescript:
    name: TypeScript (Biome)
@@ -36,7 +63,7 @@ jobs:
          version: 2.3.11

      - name: Run Biome
-        working-directory: apps/desktop
+        working-directory: apps/frontend
        # biome ci fails on errors by default; warnings are reported but don't block
        # Use --error-on-warnings when ready to enforce all rules
        run: biome ci .
@@ -47,13 +74,15 @@ jobs:
  lint-complete:
    name: Lint Complete
    runs-on: ubuntu-latest
-    needs: [typescript]
+    needs: [python, typescript]
    if: always()
    steps:
      - name: Check lint results
        run: |
-          if [[ "${{ needs.typescript.result }}" != "success" ]]; then
+          if [[ "${{ needs.python.result }}" != "success" ]] || \
+             [[ "${{ needs.typescript.result }}" != "success" ]]; then
            echo "❌ Linting failed"
+            echo "  Python:     ${{ needs.python.result }}"
            echo "  TypeScript: ${{ needs.typescript.result }}"
            exit 1
          fi
@@ -56,14 +56,15 @@ jobs:

              // Area detection paths
              AREA_PATHS: Object.freeze({
-                frontend: 'apps/desktop/',
+                frontend: 'apps/frontend/',
+                backend: 'apps/backend/',
                ci: '.github/'
              }),

              // Label definitions
              LABELS: Object.freeze({
                SIZE: ['size/XS', 'size/S', 'size/M', 'size/L', 'size/XL'],
-                AREA: ['area/frontend', 'area/ci']
+                AREA: ['area/frontend', 'area/backend', 'area/fullstack', 'area/ci']
              }),

              // Pagination
@@ -116,15 +117,16 @@ jobs:
            /**
             * Detect areas affected by file changes
             * @param {Array} files - List of changed files
-             * @returns {{frontend: boolean, ci: boolean}}
+             * @returns {{frontend: boolean, backend: boolean, ci: boolean}}
             */
            function detectAreas(files) {
-              const areas = { frontend: false, ci: false };
+              const areas = { frontend: false, backend: false, ci: false };
              const { AREA_PATHS } = CONFIG;

              for (const file of files) {
                const path = file.filename || '';
                if (path.startsWith(AREA_PATHS.frontend)) areas.frontend = true;
+                if (path.startsWith(AREA_PATHS.backend)) areas.backend = true;
                if (path.startsWith(AREA_PATHS.ci)) areas.ci = true;
              }

@@ -133,11 +135,13 @@ jobs:

            /**
             * Determine area label based on detected areas
-             * @param {{frontend: boolean, ci: boolean}} areas
+             * @param {{frontend: boolean, backend: boolean, ci: boolean}} areas
             * @returns {string|null} Area label or null
             */
            function determineAreaLabel(areas) {
+              if (areas.frontend && areas.backend) return 'area/fullstack';
              if (areas.frontend) return 'area/frontend';
+              if (areas.backend) return 'area/backend';
              if (areas.ci) return 'area/ci';
              return null;
            }
@@ -10,7 +10,7 @@ on:
  push:
    branches: [main]
    paths:
-      - 'apps/desktop/package.json'
+      - 'apps/frontend/package.json'
      - 'package.json'
  workflow_dispatch:
    inputs:
@@ -50,7 +50,7 @@ jobs:
      - name: Get package version
        id: package
        run: |
-          VERSION=$(node -p "require('./apps/desktop/package.json').version")
+          VERSION=$(node -p "require('./apps/frontend/package.json').version")
          echo "version=$VERSION" >> $GITHUB_OUTPUT
          echo "Package version: $VERSION"

@@ -1,19 +1,24 @@
 name: Quality Security

 # CodeQL runs on all PRs, pushes to main, and weekly schedule
-# Note: CodeQL takes 20-30 min
+# Note: CodeQL takes 20-30 min per language (40-60 min total)
+# Bandit is fast (5-10 min)

 on:
  push:
    branches: [main]
    paths:
-      - 'apps/desktop/**'
+      - 'apps/**'
+      - 'tests/**'
+      - 'pyproject.toml'
      - 'package.json'
      - '.github/workflows/quality-security.yml'
  pull_request:
    branches: [main, develop]
    paths:
-      - 'apps/desktop/**'
+      - 'apps/**'
+      - 'tests/**'
+      - 'pyproject.toml'
      - 'package.json'
      - '.github/workflows/quality-security.yml'
  schedule:
@@ -36,7 +41,7 @@ jobs:
    strategy:
      fail-fast: false
      matrix:
-        language: [javascript-typescript]
+        language: [python, javascript-typescript]
    steps:
      - name: Checkout
        uses: actions/checkout@v4
@@ -55,13 +60,91 @@ jobs:
        with:
          category: "/language:${{ matrix.language }}"

+  # Bandit runs on all PRs - it's fast (5-10 min)
+  python-security:
+    name: Python Security (Bandit)
+    runs-on: ubuntu-latest
+    timeout-minutes: 10
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v4
+
+      - name: Set up Python
+        uses: actions/setup-python@v6
+        with:
+          python-version: '3.12'
+
+      - name: Install Bandit
+        run: pip install bandit
+
+      - name: Run Bandit security scan
+        id: bandit
+        run: |
+          echo "::group::Running Bandit security scan"
+          bandit -r apps/backend/ -ll -ii -f json -o bandit-report.json || BANDIT_EXIT=$?
+          if [ "${BANDIT_EXIT:-0}" -gt 1 ]; then
+            echo "::error::Bandit scan failed with exit code $BANDIT_EXIT"
+            exit 1
+          fi
+          echo "::endgroup::"
+
+      - name: Analyze Bandit results
+        uses: actions/github-script@v8
+        with:
+          script: |
+            const fs = require('fs');
+
+            if (!fs.existsSync('bandit-report.json')) {
+              core.setFailed('Bandit report not found - scan may have failed');
+              return;
+            }
+
+            const report = JSON.parse(fs.readFileSync('bandit-report.json', 'utf8'));
+            const results = report.results || [];
+
+            const high = results.filter(r => r.issue_severity === 'HIGH');
+            const medium = results.filter(r => r.issue_severity === 'MEDIUM');
+            const low = results.filter(r => r.issue_severity === 'LOW');
+
+            console.log(`::group::Bandit Security Scan Results`);
+            console.log(`Found ${results.length} issues:`);
+            console.log(`  HIGH:   ${high.length}`);
+            console.log(`  MEDIUM: ${medium.length}`);
+            console.log(`  LOW:    ${low.length}`);
+            console.log('::endgroup::');
+
+            let summary = `## Python Security Scan (Bandit)\n\n`;
+            summary += `| Severity | Count |\n`;
+            summary += `|----------|-------|\n`;
+            summary += `| High | ${high.length} |\n`;
+            summary += `| Medium | ${medium.length} |\n`;
+            summary += `| Low | ${low.length} |\n\n`;
+
+            if (high.length > 0) {
+              summary += `### High Severity Issues\n\n`;
+              for (const issue of high) {
+                summary += `- **${issue.filename}:${issue.line_number}**\n`;
+                summary += `  - ${issue.issue_text}\n`;
+                summary += `  - Test: \`${issue.test_id}\` (${issue.test_name})\n\n`;
+              }
+            }
+
+            core.summary.addRaw(summary);
+            await core.summary.write();
+
+            if (high.length > 0) {
+              core.setFailed(`Found ${high.length} high severity security issue(s)`);
+            } else {
+              console.log('No high severity security issues found');
+            }
+
  # --------------------------------------------------------------------------
  # Gate Job - Single check for branch protection
  # --------------------------------------------------------------------------
  security-summary:
    name: Security Summary
    runs-on: ubuntu-latest
-    needs: [codeql]
+    needs: [codeql, python-security]
    if: always()
    timeout-minutes: 5
    steps:
@@ -70,15 +153,19 @@ jobs:
        with:
          script: |
            const codeql = '${{ needs.codeql.result }}';
+            const bandit = '${{ needs.python-security.result }}';

            console.log('Security Check Results:');
            console.log(`  CodeQL:        ${codeql}`);
+            console.log(`  Bandit:        ${bandit}`);

            // Only 'failure' is a real failure; 'skipped' is acceptable (e.g., path filters, PR skipping CodeQL)
            const acceptable = ['success', 'skipped'];
            const codeqlOk = acceptable.includes(codeql);
+            const banditOk = acceptable.includes(bandit);
+            const allPassed = codeqlOk && banditOk;

-            if (codeqlOk) {
+            if (allPassed) {
              console.log('\n✅ All security checks passed');
              core.summary.addRaw('## ✅ Security Checks Passed\n\nAll security scans completed successfully.');
            } else {
@@ -29,18 +29,42 @@ jobs:
    steps:
      - uses: actions/checkout@v4

+      - name: Setup Python
+        uses: actions/setup-python@v6
+        with:
+          python-version: '3.11'
+
      - name: Setup Node.js and install dependencies
        uses: ./.github/actions/setup-node-frontend

+      - name: Install Rust toolchain (for building native Python packages)
+        uses: dtolnay/rust-toolchain@stable
+
+      - name: Cache pip wheel cache (for compiled packages like real_ladybug)
+        uses: actions/cache@v5
+        with:
+          path: ~/Library/Caches/pip
+          key: pip-wheel-${{ runner.os }}-x64-${{ hashFiles('apps/backend/requirements.txt') }}
+          restore-keys: |
+            pip-wheel-${{ runner.os }}-x64-
+
+      - name: Cache bundled Python
+        uses: actions/cache@v5
+        with:
+          path: apps/frontend/python-runtime
+          key: python-bundle-${{ runner.os }}-x64-3.12.8-rust-${{ hashFiles('apps/backend/requirements.txt') }}
+          restore-keys: |
+            python-bundle-${{ runner.os }}-x64-3.12.8-rust-
+
      - name: Build application
-        run: cd apps/desktop && npm run build
+        run: cd apps/frontend && npm run build
        env:
          SENTRY_DSN: ${{ secrets.SENTRY_DSN }}
          SENTRY_TRACES_SAMPLE_RATE: ${{ secrets.SENTRY_TRACES_SAMPLE_RATE }}
          SENTRY_PROFILES_SAMPLE_RATE: ${{ secrets.SENTRY_PROFILES_SAMPLE_RATE }}

      - name: Package macOS (Intel)
-        run: cd apps/desktop && npm run package:mac -- --x64
+        run: cd apps/frontend && npm run package:mac -- --x64
        env:
          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
          CSC_LINK: ${{ secrets.MAC_CERTIFICATE }}
@@ -62,10 +86,10 @@ jobs:
        with:
          name: macos-intel-builds
          path: |
-            apps/desktop/dist/*.dmg
-            apps/desktop/dist/*.zip
-            apps/desktop/dist/*.yml
-            apps/desktop/dist/*.blockmap
+            apps/frontend/dist/*.dmg
+            apps/frontend/dist/*.zip
+            apps/frontend/dist/*.yml
+            apps/frontend/dist/*.blockmap

  # Apple Silicon build on ARM64 runner for native compilation
  build-macos-arm64:
@@ -76,18 +100,39 @@ jobs:
    steps:
      - uses: actions/checkout@v4

+      - name: Setup Python
+        uses: actions/setup-python@v6
+        with:
+          python-version: '3.11'
+
      - name: Setup Node.js and install dependencies
        uses: ./.github/actions/setup-node-frontend

+      - name: Cache pip wheel cache
+        uses: actions/cache@v5
+        with:
+          path: ~/Library/Caches/pip
+          key: pip-wheel-${{ runner.os }}-arm64-${{ hashFiles('apps/backend/requirements.txt') }}
+          restore-keys: |
+            pip-wheel-${{ runner.os }}-arm64-
+
+      - name: Cache bundled Python
+        uses: actions/cache@v5
+        with:
+          path: apps/frontend/python-runtime
+          key: python-bundle-${{ runner.os }}-arm64-3.12.8-${{ hashFiles('apps/backend/requirements.txt') }}
+          restore-keys: |
+            python-bundle-${{ runner.os }}-arm64-3.12.8-
+
      - name: Build application
-        run: cd apps/desktop && npm run build
+        run: cd apps/frontend && npm run build
        env:
          SENTRY_DSN: ${{ secrets.SENTRY_DSN }}
          SENTRY_TRACES_SAMPLE_RATE: ${{ secrets.SENTRY_TRACES_SAMPLE_RATE }}
          SENTRY_PROFILES_SAMPLE_RATE: ${{ secrets.SENTRY_PROFILES_SAMPLE_RATE }}

      - name: Package macOS (Apple Silicon)
-        run: cd apps/desktop && npm run package:mac -- --arm64
+        run: cd apps/frontend && npm run package:mac -- --arm64
        env:
          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
          CSC_LINK: ${{ secrets.MAC_CERTIFICATE }}
@@ -109,10 +154,10 @@ jobs:
        with:
          name: macos-arm64-builds
          path: |
-            apps/desktop/dist/*.dmg
-            apps/desktop/dist/*.zip
-            apps/desktop/dist/*.yml
-            apps/desktop/dist/*.blockmap
+            apps/frontend/dist/*.dmg
+            apps/frontend/dist/*.zip
+            apps/frontend/dist/*.yml
+            apps/frontend/dist/*.blockmap

  build-windows:
    runs-on: windows-latest
@@ -125,18 +170,39 @@ jobs:
    steps:
      - uses: actions/checkout@v4

+      - name: Setup Python
+        uses: actions/setup-python@v6
+        with:
+          python-version: '3.11'
+
      - name: Setup Node.js and install dependencies
        uses: ./.github/actions/setup-node-frontend

+      - name: Cache pip wheel cache
+        uses: actions/cache@v5
+        with:
+          path: ~\AppData\Local\pip\Cache
+          key: pip-wheel-${{ runner.os }}-x64-${{ hashFiles('apps/backend/requirements.txt') }}
+          restore-keys: |
+            pip-wheel-${{ runner.os }}-x64-
+
+      - name: Cache bundled Python
+        uses: actions/cache@v5
+        with:
+          path: apps/frontend/python-runtime
+          key: python-bundle-${{ runner.os }}-x64-3.12.8-${{ hashFiles('apps/backend/requirements.txt') }}
+          restore-keys: |
+            python-bundle-${{ runner.os }}-x64-3.12.8-
+
      - name: Build application
-        run: cd apps/desktop && npm run build
+        run: cd apps/frontend && npm run build
        env:
          SENTRY_DSN: ${{ secrets.SENTRY_DSN }}
          SENTRY_TRACES_SAMPLE_RATE: ${{ secrets.SENTRY_TRACES_SAMPLE_RATE }}
          SENTRY_PROFILES_SAMPLE_RATE: ${{ secrets.SENTRY_PROFILES_SAMPLE_RATE }}

      - name: Package Windows
-        run: cd apps/desktop && npm run package:win
+        run: cd apps/frontend && npm run package:win
        env:
          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
          # Disable electron-builder's built-in signing (we use Azure Trusted Signing instead)
@@ -160,7 +226,7 @@ jobs:
          endpoint: https://neu.codesigning.azure.net/
          trusted-signing-account-name: ${{ secrets.AZURE_SIGNING_ACCOUNT }}
          certificate-profile-name: ${{ secrets.AZURE_CERTIFICATE_PROFILE }}
-          files-folder: apps/desktop/dist
+          files-folder: apps/frontend/dist
          files-folder-filter: exe
          file-digest: SHA256
          timestamp-rfc3161: http://timestamp.acs.microsoft.com
@@ -170,7 +236,7 @@ jobs:
        if: env.AZURE_CLIENT_ID != ''
        shell: pwsh
        run: |
-          cd apps/desktop/dist
+          cd apps/frontend/dist
          $exeFile = Get-ChildItem -Filter "*.exe" | Select-Object -First 1
          if ($exeFile) {
            Write-Host "Verifying signature on $($exeFile.Name)..."
@@ -194,7 +260,7 @@ jobs:
        shell: pwsh
        run: |
          $ErrorActionPreference = "Stop"
-          cd apps/desktop/dist
+          cd apps/frontend/dist

          # Find the installer exe (electron-builder names it with "Setup" or just the app name)
          # electron-builder produces one installer exe per build
@@ -261,35 +327,56 @@ jobs:
        with:
          name: windows-builds
          path: |
-            apps/desktop/dist/*.exe
-            apps/desktop/dist/*.yml
-            apps/desktop/dist/*.blockmap
+            apps/frontend/dist/*.exe
+            apps/frontend/dist/*.yml
+            apps/frontend/dist/*.blockmap

  build-linux:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4

+      - name: Setup Python
+        uses: actions/setup-python@v6
+        with:
+          python-version: '3.11'
+
      - name: Setup Node.js and install dependencies
        uses: ./.github/actions/setup-node-frontend

      - name: Setup Flatpak and verification tools
        run: |
          sudo apt-get update
-          sudo apt-get install -y flatpak flatpak-builder squashfs-tools
+          sudo apt-get install -y flatpak flatpak-builder libarchive-tools
          flatpak remote-add --user --if-not-exists flathub https://flathub.org/repo/flathub.flatpakrepo
          flatpak install -y --user flathub org.freedesktop.Platform//25.08 org.freedesktop.Sdk//25.08
          flatpak install -y --user flathub org.electronjs.Electron2.BaseApp//25.08

+      - name: Cache pip wheel cache
+        uses: actions/cache@v5
+        with:
+          path: ~/.cache/pip
+          key: pip-wheel-${{ runner.os }}-x64-${{ hashFiles('apps/backend/requirements.txt') }}
+          restore-keys: |
+            pip-wheel-${{ runner.os }}-x64-
+
+      - name: Cache bundled Python
+        uses: actions/cache@v5
+        with:
+          path: apps/frontend/python-runtime
+          key: python-bundle-${{ runner.os }}-x64-3.12.8-${{ hashFiles('apps/backend/requirements.txt') }}
+          restore-keys: |
+            python-bundle-${{ runner.os }}-x64-3.12.8-
+
      - name: Build application
-        run: cd apps/desktop && npm run build
+        run: cd apps/frontend && npm run build
        env:
          SENTRY_DSN: ${{ secrets.SENTRY_DSN }}
          SENTRY_TRACES_SAMPLE_RATE: ${{ secrets.SENTRY_TRACES_SAMPLE_RATE }}
          SENTRY_PROFILES_SAMPLE_RATE: ${{ secrets.SENTRY_PROFILES_SAMPLE_RATE }}

      - name: Package Linux
-        run: cd apps/desktop && npm run package:linux
+        run: cd apps/frontend && npm run package:linux
        env:
          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
          SENTRY_DSN: ${{ secrets.SENTRY_DSN }}
@@ -297,18 +384,18 @@ jobs:
          SENTRY_PROFILES_SAMPLE_RATE: ${{ secrets.SENTRY_PROFILES_SAMPLE_RATE }}

      - name: Verify Linux packages
-        run: cd apps/desktop && npm run verify:linux
+        run: cd apps/frontend && npm run verify:linux

      - name: Upload artifacts
        uses: actions/upload-artifact@v4
        with:
          name: linux-builds
          path: |
-            apps/desktop/dist/*.AppImage
-            apps/desktop/dist/*.deb
-            apps/desktop/dist/*.flatpak
-            apps/desktop/dist/*.yml
-            apps/desktop/dist/*.blockmap
+            apps/frontend/dist/*.AppImage
+            apps/frontend/dist/*.deb
+            apps/frontend/dist/*.flatpak
+            apps/frontend/dist/*.yml
+            apps/frontend/dist/*.blockmap

  # Finalize macOS notarization (runs in parallel with Windows/Linux builds)
  finalize-notarization:
@@ -533,7 +620,7 @@ jobs:
          draft: false
          prerelease: ${{ contains(github.ref, 'beta') || contains(github.ref, 'alpha') }}
        env:
-          GITHUB_TOKEN: ${{ secrets.PAT_TOKEN || secrets.GITHUB_TOKEN }}
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}

  # Update README with new version after successful release
  update-readme:
@@ -66,10 +66,52 @@ lerna-debug.log*
 .update-metadata.json

 # ===========================
-# Node.js (apps/desktop)
+# Python (apps/backend)
+# ===========================
+__pycache__/
+*.py[cod]
+*$py.class
+*.so
+.Python
+build/
+develop-eggs/
+eggs/
+.eggs/
+*.egg-info/
+.installed.cfg
+*.egg
+MANIFEST
+
+# Virtual environments
+.venv/
+venv/
+ENV/
+env/
+.conda/
+
+# Testing
+.pytest_cache/
+.coverage
+htmlcov/
+.tox/
+.nox/
+coverage.xml
+*.cover
+*.py,cover
+.hypothesis/
+
+# Type checking
+.mypy_cache/
+.dmypy.json
+dmypy.json
+.pytype/
+.pyre/
+
+# ===========================
+# Node.js (apps/frontend)
 # ===========================
 node_modules
-apps/desktop/node_modules
+apps/frontend/node_modules
 .npm
 .yarn/
 .pnp.*
@@ -78,6 +120,7 @@ apps/desktop/node_modules
 dist/
 out/
 *.tsbuildinfo
+apps/frontend/python-runtime/

 # Cache
 .cache/
@@ -89,8 +132,8 @@ out/
 # ===========================
 # Electron
 # ===========================
-apps/desktop/dist/
-apps/desktop/out/
+apps/frontend/dist/
+apps/frontend/out/
 *.asar
 *.blockmap
 *.snap
@@ -110,12 +153,6 @@ test-results/
 playwright-report/
 playwright/.cache/

-# ===========================
-# Python
-# ===========================
-__pycache__/
-*.pyc
-
 # ===========================
 # Misc
 # ===========================
@@ -48,18 +48,26 @@ if git diff --cached --name-only | grep -q "^package.json$"; then
  VERSION=$(node -p "require('./package.json').version")

  if [ -n "$VERSION" ]; then
-    # Sync to apps/desktop/package.json
-    if [ -f "apps/desktop/package.json" ]; then
+    # Sync to apps/frontend/package.json
+    if [ -f "apps/frontend/package.json" ]; then
      node -e "
        const fs = require('fs');
-        const pkg = require('./apps/desktop/package.json');
+        const pkg = require('./apps/frontend/package.json');
        if (pkg.version !== '$VERSION') {
          pkg.version = '$VERSION';
-          fs.writeFileSync('./apps/desktop/package.json', JSON.stringify(pkg, null, 2) + '\n');
-          console.log('  Updated apps/desktop/package.json to $VERSION');
+          fs.writeFileSync('./apps/frontend/package.json', JSON.stringify(pkg, null, 2) + '\n');
+          console.log('  Updated apps/frontend/package.json to $VERSION');
        }
      "
-      git add apps/desktop/package.json
+      git add apps/frontend/package.json
+    fi
+
+    # Sync to apps/backend/__init__.py
+    if [ -f "apps/backend/__init__.py" ]; then
+      sed -i.bak "s/__version__ = \"[^\"]*\"/__version__ = \"$VERSION\"/" apps/backend/__init__.py
+      rm -f apps/backend/__init__.py.bak
+      git add apps/backend/__init__.py
+      echo "  Updated apps/backend/__init__.py to $VERSION"
    fi

    # Sync to README.md - section-aware updates (stable vs beta)
@@ -111,14 +119,126 @@ if git diff --cached --name-only | grep -q "^package.json$"; then
  fi
 fi

-
 # =============================================================================
-# DESKTOP APP CHECKS (TypeScript/React)
+# BACKEND CHECKS (Python) - Run first, before frontend
 # =============================================================================

-# Check if there are staged files in apps/desktop
-if git diff --cached --name-only | grep -q "^apps/desktop/"; then
-  echo "Desktop app changes detected, running checks..."
+# Check if there are staged Python files in apps/backend
+if git diff --cached --name-only | grep -q "^apps/backend/.*\.py$"; then
+  echo "Python changes detected, running backend checks..."
+
+  # Detect if we're in a worktree
+  IS_WORKTREE=false
+  if [ -f ".git" ]; then
+    # .git is a file (not directory) in worktrees
+    IS_WORKTREE=true
+  fi
+
+  # Determine ruff command (venv or global)
+  RUFF=""
+  if [ -f "apps/backend/.venv/bin/ruff" ]; then
+    RUFF="apps/backend/.venv/bin/ruff"
+  elif [ -f "apps/backend/.venv/Scripts/ruff.exe" ]; then
+    RUFF="apps/backend/.venv/Scripts/ruff.exe"
+  elif command -v ruff >/dev/null 2>&1; then
+    RUFF="ruff"
+  fi
+
+  if [ -n "$RUFF" ]; then
+    # Get only staged Python files in apps/backend (process only what's being committed)
+    STAGED_PY_FILES=$(git diff --cached --name-only --diff-filter=ACM | grep "^apps/backend/.*\.py$" || true)
+
+    if [ -n "$STAGED_PY_FILES" ]; then
+      # Run ruff linting (auto-fix) only on staged files
+      echo "Running ruff lint on staged files..."
+      echo "$STAGED_PY_FILES" | xargs $RUFF check --fix
+      if [ $? -ne 0 ]; then
+        echo "Ruff lint failed. Please fix Python linting errors before committing."
+        exit 1
+      fi
+
+      # Run ruff format (auto-fix) only on staged files
+      echo "Running ruff format on staged files..."
+      echo "$STAGED_PY_FILES" | xargs $RUFF format
+
+      # Re-stage only the files that were originally staged (in case ruff modified them)
+      echo "$STAGED_PY_FILES" | xargs git add
+    fi
+  else
+    if [ "$IS_WORKTREE" = true ]; then
+      echo ""
+      echo "⚠️  WARNING: ruff not available in this worktree."
+      echo "   Python linting checks will be skipped."
+      echo "   This is expected for auto-claude worktrees."
+      echo "   Full validation will occur when PR is created/merged."
+      echo ""
+    else
+      echo "Warning: ruff not found, skipping Python linting. Install with: uv pip install ruff"
+    fi
+  fi
+
+  # Run pytest (skip slow/integration tests and Windows-incompatible tests for pre-commit speed)
+  # Run from repo root (not apps/backend) so tests that use Path.resolve() get correct CWD.
+  # PYTHONPATH includes apps/backend so imports resolve correctly.
+  echo "Running Python tests..."
+  (
+    # Tests to skip: graphiti (external deps), merge_file_tracker/service_orchestrator/worktree/workspace (Windows path/git issues)
+    # Also skip tests that require optional dependencies (pydantic structured outputs)
+    # Also skip gitlab_e2e (e2e test sensitive to test-ordering env contamination, validated by CI)
+    IGNORE_TESTS="--ignore=tests/test_graphiti.py --ignore=tests/test_merge_file_tracker.py --ignore=tests/test_service_orchestrator.py --ignore=tests/test_worktree.py --ignore=tests/test_workspace.py --ignore=tests/test_finding_validation.py --ignore=tests/test_sdk_structured_output.py --ignore=tests/test_structured_outputs.py --ignore=tests/test_gitlab_e2e.py"
+    # Determine Python executable from venv
+    VENV_PYTHON=""
+    if [ -f "apps/backend/.venv/bin/python" ]; then
+      VENV_PYTHON="apps/backend/.venv/bin/python"
+    elif [ -f "apps/backend/.venv/Scripts/python.exe" ]; then
+      VENV_PYTHON="apps/backend/.venv/Scripts/python.exe"
+    fi
+
+    # -k "not windows_path": skip tests using fake Windows paths that break
+    # Path.resolve() on macOS/Linux. These are validated by CI on all platforms.
+    if [ -n "$VENV_PYTHON" ]; then
+      # Check if pytest is installed in venv
+      if $VENV_PYTHON -c "import pytest" 2>/dev/null; then
+        PYTHONPATH=apps/backend $VENV_PYTHON -m pytest tests/ -v --tb=short -x -m "not slow and not integration" -k "not windows_path" $IGNORE_TESTS
+      else
+        echo "Warning: pytest not installed in venv. Installing test dependencies..."
+        $VENV_PYTHON -m pip install -q -r tests/requirements-test.txt
+        PYTHONPATH=apps/backend $VENV_PYTHON -m pytest tests/ -v --tb=short -x -m "not slow and not integration" -k "not windows_path" $IGNORE_TESTS
+      fi
+    elif [ -d "apps/backend/.venv" ]; then
+      echo "Warning: venv exists but Python not found in it, using system Python"
+      PYTHONPATH=apps/backend python -m pytest tests/ -v --tb=short -x -m "not slow and not integration" -k "not windows_path" $IGNORE_TESTS
+    elif [ "$IS_WORKTREE" = true ]; then
+      echo ""
+      echo "⚠️  WARNING: Python venv not available in this worktree."
+      echo "   Python tests will be skipped."
+      echo "   This is expected for auto-claude worktrees."
+      echo "   Full validation will occur when PR is created/merged."
+      echo ""
+      exit 77  # GNU convention for 'test skipped' (avoids pytest exit-code collision)
+    else
+      echo "Warning: No .venv found in apps/backend, using system Python"
+      PYTHONPATH=apps/backend python -m pytest tests/ -v --tb=short -x -m "not slow and not integration" -k "not windows_path" $IGNORE_TESTS
+    fi
+  )
+  PYTHON_EXIT=$?
+  if [ $PYTHON_EXIT -eq 77 ]; then
+    echo "Backend checks passed! (Python tests skipped — worktree)"
+  elif [ $PYTHON_EXIT -ne 0 ]; then
+    echo "Python tests failed. Please fix failing tests before committing."
+    exit 1
+  else
+    echo "Backend checks passed!"
+  fi
+fi
+
+# =============================================================================
+# FRONTEND CHECKS (TypeScript/React)
+# =============================================================================
+
+# Check if there are staged files in apps/frontend
+if git diff --cached --name-only | grep -q "^apps/frontend/"; then
+  echo "Frontend changes detected, running frontend checks..."

  # Detect if we're in a worktree and check if dependencies are available
  IS_WORKTREE=false
@@ -132,11 +252,11 @@ if git diff --cached --name-only | grep -q "^apps/desktop/"; then

  # Check if node_modules has actual dependencies by looking for a known package
  # @lydell/node-pty is required for terminal code and is a common source of TypeScript errors
-  # It may be in root node_modules (hoisted) or apps/desktop/node_modules
+  # It may be in root node_modules (hoisted) or apps/frontend/node_modules
  # Note: -d follows symlinks automatically, so this works for both real dirs and symlinks
  # We check for the full package path (@lydell/node-pty) rather than just the namespace
  # for precise detection - ensures the actual dependency is installed, not just any @lydell package
-  if [ ! -d "node_modules/@lydell/node-pty" ] && [ ! -d "apps/desktop/node_modules/@lydell/node-pty" ]; then
+  if [ ! -d "node_modules/@lydell/node-pty" ] && [ ! -d "apps/frontend/node_modules/@lydell/node-pty" ]; then
    DEPS_AVAILABLE=false
  fi

@@ -158,7 +278,7 @@ if git diff --cached --name-only | grep -q "^apps/desktop/"; then
    # Dependencies available - run full frontend checks
    # Use subshell to isolate directory changes and prevent worktree corruption
    (
-      cd apps/desktop
+      cd apps/frontend

      # Run lint-staged (handles staged .ts/.tsx files)
      npm exec lint-staged
@@ -167,14 +287,22 @@ if git diff --cached --name-only | grep -q "^apps/desktop/"; then
        exit 1
      fi

-      # Run TypeScript type check (incremental: only rechecks changed files after first run)
+      # Run TypeScript type check
      echo "Running type check..."
-      NODE_OPTIONS="--max-old-space-size=2048" npm run typecheck
+      npm run typecheck
      if [ $? -ne 0 ]; then
        echo "Type check failed. Please fix TypeScript errors before committing."
        exit 1
      fi

+      # Run linting
+      echo "Running lint..."
+      npm run lint
+      if [ $? -ne 0 ]; then
+        echo "Lint failed. Run 'npm run lint:fix' to auto-fix issues."
+        exit 1
+      fi
+
      # Check for vulnerabilities (only critical severity)
      # Note: Using critical level because electron-builder has a known high-severity
      # tar vulnerability (CVE-2026-23745) that cannot be fixed until electron-builder
@@ -18,17 +18,20 @@ repos:
            VERSION=$(node -p "require('./package.json').version")
            if [ -n "$VERSION" ]; then

-              # Sync to apps/desktop/package.json
+              # Sync to apps/frontend/package.json
              node -e "
                const fs = require('fs');
-                const p = require('./apps/desktop/package.json');
+                const p = require('./apps/frontend/package.json');
                const v = process.argv[1];
                if (p.version !== v) {
                  p.version = v;
-                  fs.writeFileSync('./apps/desktop/package.json', JSON.stringify(p, null, 2) + '\n');
+                  fs.writeFileSync('./apps/frontend/package.json', JSON.stringify(p, null, 2) + '\n');
                }
              " "$VERSION"

+              # Sync to apps/backend/__init__.py
+              sed -i.bak "s/__version__ = \"[^\"]*\"/__version__ = \"$VERSION\"/" apps/backend/__init__.py && rm -f apps/backend/__init__.py.bak
+
              # Sync to README.md - section-aware updates (stable vs beta)
              ESCAPED_VERSION=$(echo "$VERSION" | sed 's/-/--/g')

@@ -67,13 +70,66 @@ repos:
              rm -f README.md.bak

              # Stage changes
-              git add apps/desktop/package.json README.md 2>/dev/null || true
+              git add apps/frontend/package.json apps/backend/__init__.py README.md 2>/dev/null || true
            fi
        language: system
        files: ^package\.json$
        pass_filenames: false

-  # Frontend linting (apps/desktop/) - Biome is 15-25x faster than ESLint
+  # Python encoding check - prevent regression of UTF-8 encoding fixes (PR #782)
+  - repo: local
+    hooks:
+      - id: check-file-encoding
+        name: Check file encoding parameters
+        entry: python scripts/check_encoding.py
+        language: system
+        types: [python]
+        files: ^apps/backend/
+        description: Ensures all file operations specify encoding="utf-8"
+
+  # Python linting (apps/backend/)
+  - repo: https://github.com/astral-sh/ruff-pre-commit
+    rev: v0.14.10
+    hooks:
+      - id: ruff
+        args: [--fix]
+        files: ^apps/backend/
+      - id: ruff-format
+        files: ^apps/backend/
+
+  # Python tests (apps/backend/) - run full test suite from project root
+  # Tests to skip: graphiti (external deps), merge_file_tracker/service_orchestrator/worktree/workspace (Windows path/git issues)
+  - repo: local
+    hooks:
+      - id: pytest
+        name: Python Tests
+        entry: bash
+        args:
+          - -c
+          - |
+            # Run pytest directly from project root
+            if [ -f "apps/backend/.venv/bin/pytest" ]; then
+              PYTEST_CMD="apps/backend/.venv/bin/pytest"
+            elif [ -f "apps/backend/.venv/Scripts/pytest.exe" ]; then
+              PYTEST_CMD="apps/backend/.venv/Scripts/pytest.exe"
+            else
+              PYTEST_CMD="python -m pytest"
+            fi
+            $PYTEST_CMD tests/ \
+              -v \
+              --tb=short \
+              -x \
+              -m "not slow and not integration" \
+              --ignore=tests/test_graphiti.py \
+              --ignore=tests/test_merge_file_tracker.py \
+              --ignore=tests/test_service_orchestrator.py \
+              --ignore=tests/test_worktree.py \
+              --ignore=tests/test_workspace.py
+        language: system
+        files: ^(apps/backend/.*\.py$|tests/.*\.py$)
+        pass_filenames: false
+
+  # Frontend linting (apps/frontend/) - Biome is 15-25x faster than ESLint
  # NOTE: These hooks check for worktree context to avoid npm/node_modules issues
  - repo: local
    hooks:
@@ -84,13 +140,13 @@ repos:
          - -c
          - |
            # Skip in worktrees if node_modules doesn't exist (Biome not installed)
-            if [ -f ".git" ] && [ ! -d "apps/desktop/node_modules" ]; then
+            if [ -f ".git" ] && [ ! -d "apps/frontend/node_modules" ]; then
              echo "Skipping Biome in worktree (node_modules not found)"
              exit 0
            fi
-            cd apps/desktop && npx biome check --write --no-errors-on-unmatched .
+            cd apps/frontend && npx biome check --write --no-errors-on-unmatched .
        language: system
-        files: ^apps/desktop/.*\.(ts|tsx|js|jsx|json)$
+        files: ^apps/frontend/.*\.(ts|tsx|js|jsx|json)$
        pass_filenames: false

      - id: typecheck
@@ -100,13 +156,13 @@ repos:
          - -c
          - |
            # Skip in worktrees if node_modules doesn't exist (dependencies not installed)
-            if [ -f ".git" ] && [ ! -d "apps/desktop/node_modules" ]; then
+            if [ -f ".git" ] && [ ! -d "apps/frontend/node_modules" ]; then
              echo "Skipping TypeScript check in worktree (node_modules not found)"
              exit 0
            fi
-            cd apps/desktop && npm run typecheck
+            cd apps/frontend && npm run typecheck
        language: system
-        files: ^apps/desktop/.*\.(ts|tsx)$
+        files: ^apps/frontend/.*\.(ts|tsx)$
        pass_filenames: false

  # General checks
@@ -0,0 +1,240 @@
+# GitHub Issues Documentation Design
+
+**Date:** 2025-02-16
+**Status:** Approved
+**Author:** Claude (Superpowers Brainstorming)
+
+## Overview
+
+Create comprehensive, user-friendly documentation for Auto Claude's GitHub Issues integration. The documentation will serve a mixed audience (new and existing users) with progressive depth across three documents.
+
+## Target Audience
+
+**Mixed audience with progressive depth:**
+- **End users** - New to Auto Claude, want to understand features and get started quickly
+- **Technical users** - Existing users wanting to optimize AI configuration and manage costs
+- **Pro users/developers** - Want to customize prompts, context injection, and extend the system
+
+## Document Structure
+
+### New Directory: `guides/github-issues/`
+
+```
+guides/
+└── github-issues/
+    ├── README.md                                    # Navigation index
+    ├── github-issues-user-guide.md                  # Document 1
+    ├── github-issues-advanced-ai-configuration.md   # Document 2
+    └── github-issues-customization-guide.md         # Document 3
+```
+
+### Document 1: User Guide
+
+**Audience:** End users (non-technical)
+**Focus:** Features, workflows, getting started
+
+**Sections:**
+1. **Overview** - What is GitHub Issues integration? Why use it?
+2. **Quick Start (5 minutes)** - Get your first issue investigated end-to-end
+3. **Key Features** - Bullet points of main capabilities
+4. **Integration Workflow** - Import → Investigate → Create Task → Implement (EMPHASIS)
+5. **Setup & Configuration** - GitHub authentication, connecting repos
+6. **Using the Features**
+   - Importing & browsing issues
+   - Running AI investigations
+   - Creating tasks from results
+   - Posting findings back to GitHub
+7. **FAQ** - Common questions
+
+**Tone:** Friendly, encouraging, approachable. Plain English with minimal jargon.
+
+### Document 2: Advanced AI Configuration
+
+**Audience:** Technical users, team leads
+**Focus:** AI tuning, performance, cost management
+
+**Sections:**
+1. **Overview** - Who this guide is for
+2. **Opus 4.6 Features** - Fast Mode, 128K tokens, adaptive thinking
+3. **The 4 Specialist Agents** - Deep dive into each agent's role
+4. **Pricing & Cost Management** - Token limits, cost optimization strategies
+5. **Advanced Configuration** - Per-specialist settings, performance tuning
+6. **Technical Architecture** - How investigation works under the hood
+
+**Tone:** Professional, technical but clear. Explains technical concepts in context.
+
+### Document 3: Customization Guide
+
+**Audience:** Developers, pro users extending Auto Claude
+**Focus:** System customization, prompts, context injection
+
+**Sections:**
+1. **Overview** - Who this is for (developers extending Auto Claude)
+2. **Prompt System Architecture** - How agent prompts are structured
+3. **Context Injection System** - How context is built and passed to agents
+4. **Customizing Agent Prompts** - Modifying XML prompt files
+   - Finding the prompt files
+   - Prompt structure and variables
+   - Creating custom investigation specialists
+5. **Context Configuration** - Customizing what data is included
+   - File selection patterns
+   - Context window management
+   - Repository context settings
+6. **Adding Custom Specialists** - Creating new investigation agents
+7. **Extending the Integration** - Hooks, custom providers
+8. **Examples & Recipes** - Common customizations
+
+**Tone:** Developer-to-developer, technical, code-heavy, minimal hand-holding.
+
+## Content Approach
+
+### Progressive Disclosure
+- Each document builds on the previous one
+- Clear navigation links between documents
+- Users can enter at their appropriate level
+
+### Writing Style Guidelines
+
+| Document | Tone | Language | Example Style |
+|----------|------|----------|---------------|
+| User Guide | Friendly, encouraging | Plain English | "Think of AI investigation as having a senior developer analyze the issue for you" |
+| Advanced Config | Professional, technical | Terms explained in context | "Fast Mode uses optimized token generation to reduce investigation time by 2.5x" |
+| Customization | Developer-to-developer | Technical, code-heavy | "Modify the `<system_context>` variable in `prompts/github/root_cause.xml`" |
+
+### Code & Configuration Examples
+- **Doc 1:** Simple copy-paste examples, UI navigation
+- **Doc 2:** Configuration snippets, environment variables
+- **Doc 3:** Full XML/Python examples, file paths, code blocks
+
+## Visual Elements
+
+### Screenshots (Hybrid Approach)
+
+**User Guide:**
+1. GitHub Issues main UI (issue list, filters, actions)
+2. Issue detail view (comments, labels, "Investigate" button)
+3. Investigation in progress (4 parallel specialists)
+4. Investigation results (completed report)
+5. Settings screen (GitHub auth, repo connection, Fast Mode toggle)
+
+**Advanced AI Config:**
+1. Settings → AI Investigation panel (configurable options)
+2. Token usage display (cost visibility)
+3. Investigation pipeline flowchart (issue → 4 specialists → report)
+
+**Customization Guide:**
+1. Directory structure diagram (prompt file locations)
+2. Annotated XML prompt file (variables and structure)
+3. Context injection flow diagram
+4. Code snippets throughout
+
+### Diagram Style
+- Clean, simple flowcharts
+- Consistent color scheme: Blue (user actions), Green (AI agents), Orange (data flow)
+- Minimal text, focus on flow and relationships
+
+## Navigation & Cross-References
+
+### Linking Strategy
+
+**User Guide → Deeper:**
+- "For detailed configuration, see [Advanced AI Configuration]"
+- "Learn how the 4 specialists work in [Advanced AI Configuration]"
+- "Want to customize agent behavior? See [Customization Guide]"
+
+**Advanced AI Config → Both Directions:**
+- "New to GitHub Issues? Start with the [User Guide]"
+- "Want to modify agent prompts? See [Customization Guide]"
+
+**Customization Guide → Reference:**
+- "Assumes familiarity with concepts from [User Guide] and [Advanced AI Config]"
+
+### Navigation Aids
+- Table of Contents at the top of each document
+- "In this section" callouts at the start of major sections
+- "Next steps" boxes at the end of each workflow section
+
+### External References
+- `guides/opus-4.6-features.md` - Opus 4.6 details (don't duplicate)
+- `ARCHITECTURE.md` - System architecture
+- GitHub CLI docs - Authentication setup
+
+## Key Emphasis
+
+**Integration Workflow** is the primary focus across all documents:
+
+> **Import Issues** → **AI Investigation** → **Create Task** → **Implement** → **Merge**
+
+This pipeline demonstrates the core value of Auto Claude - connecting GitHub issues to autonomous development.
+
+## File Organization
+
+### Naming Convention
+- Lowercase with hyphens (matches existing documentation style)
+- Descriptive names indicating audience and content
+- Keep under 80 characters for GitHub rendering
+
+### README.md (Index Page)
+
+```markdown
+# GitHub Issues Documentation
+
+Choose the guide that matches your needs:
+
+📖 **[User Guide](github-issues-user-guide.md)** - Get started with GitHub Issues integration
+  For: All users | Focus: Using the features
+
+⚙️ **[Advanced AI Configuration](github-issues-advanced-ai-configuration.md)** - Optimize AI investigations
+  For: Technical users | Focus: Performance, costs, Opus 4.6
+
+🔧 **[Customization Guide](github-issues-customization-guide.md)** - Extend and customize the system
+  For: Developers | Focus: Prompts, context, customization
+```
+
+## Markdown Format Guidelines
+
+- Standard GitHub-flavored markdown
+- H1 for title, H2 for main sections, H3 for subsections
+- Code blocks with language specification (`xml`, `python`, `bash`)
+- Callout boxes using `> **Note:**` format
+- Internal links use relative paths
+- Front matter with title and description metadata
+
+## Implementation Notes
+
+### Content Sources
+- `apps/frontend/src/renderer/components/github-issues/` - UI components
+- `apps/frontend/src/main/ipc-handlers/github/` - IPC handlers
+- `apps/backend/runners/github/` - Backend services
+- `guides/opus-4.6-features.md` - Reference for Opus 4.6 details
+- Existing code comments and docstrings
+
+### Screenshots to Capture
+- Need to run the application in development mode (`npm run dev`)
+- Capture each UI state listed in Visual Elements section
+- Save to `guides/github-issues/images/` with descriptive names
+
+### Diagram Creation
+- Use Mermaid syntax for flowcharts (if supported by docs build)
+- Alternatively, describe for manual creation with diagram tools
+
+## Success Criteria
+
+1. ✅ All three documents created in `guides/github-issues/`
+2. ✅ README.md index page created
+3. ✅ Progressive structure allows users to enter at appropriate level
+4. ✅ Integration workflow is clear and emphasized
+5. ✅ Screenshots included for key UI states
+6. ✅ Cross-references enable navigation between documents
+7. ✅ Writing style matches target audience for each document
+8. ✅ Content is accurate based on actual codebase features
+
+## Next Steps
+
+1. Create implementation plan using writing-plans skill
+2. Set up directory structure
+3. Draft content for each document
+4. Capture screenshots
+5. Create diagrams
+6. Review and refine
+7. Commit to repository
@@ -1,234 +1,3 @@
-## 2.7.6 - Stability & Feature Enhancements
-
-### ✨ New Features
-
- **Multi-profile account management** — Unified profile swapping with automatic token refresh and rate limit recovery for both OAuth and API-compatible providers
-
- **Enhanced terminal experience** — Customizable terminal fonts with OS-specific defaults, Claude Code CLI settings injection, and improved worktree integration
-
- **Advanced roadmap management** — Expand/collapse functionality for phase features and real-time sync with task lifecycle
-
- **Queue System v2** — Smart task prioritization with auto-promotion and intelligent rate limit recovery
-
- **GitHub integration enhancements** — AI-powered PR template generation, user-friendly API error handling, and improved review visibility
-
- **UI/UX improvements** — Spell check support for text inputs, collapsible sidebar toggle, task screenshot capture, expandable task descriptions, and bulk worktree operations
-
- **Evidence-based PR validation** — Advanced review system with trigger-driven exploration and enhanced recovery mechanisms
-
-### 🛠️ Improvements
-
- **Performance optimizations** — Async parallel worktree listing prevents UI freezes and improves responsiveness
-
- **Robustness enhancements** — Atomic file writes, better error detection in AI responses, and improved OOM/orphaned agent management for overnight builds
-
- **Terminal stability** — Fixed GPU context exhaustion from large pastes, SIGABRT crashes on macOS shutdown, and session restoration on app restart
-
- **Build & packaging** — XState bundling for packaged apps, aligned Linux package builds, and improved auto-updater for beta releases and DMG installations
-
- **Diagnostic improvements** — Sentry instrumentation for Python subprocesses and better error tracking across the system
-
-### 🐛 Bug Fixes
-
- **Terminal & PTY** — Fixed paste size limits, race conditions, rendering issues, text alignment, worktree crashes, and terminal content resizing on expansion
-
- **PR review system** — Resolved error visibility in bundled apps, improved structured output validation with three-tier recovery, preserved findings during crashes, and fixed UTC timestamp detection for comment tracking
-
- **Planning & task execution** — Fixed handling of empty/greenfield projects, atomic writes to prevent 0-byte file corruption, planning phase crashes, and implementation plan file watching
-
- **Authentication & profiles** — Resolved OAuth token revocation loops, API profile mode support without OAuth requirement, subscription type preservation during token refresh, and Linux credential file detection
-
- **Windows/cross-platform** — Complete System32 executable path fixes for where.exe and taskkill.exe, Windows credential normalization, and proper shell detection for Windows terminals
-
- **Agent management** — Fixed infinite retry loops for tool concurrency errors, auth error detection, and title generator production path resolution
-
- **UI/UX fixes** — Resolved Insights scroll-to-blank-space issues, infinite re-render loops in terminal font settings, kanban board scaling collisions, ideation stuck states, and panel constraint errors during terminal exit
-
- **Worktree & Git** — Improved branch pattern validation, removed auto-commit on deletion, support for detached HEAD state during PR creation, and better merge conflict resolution with progress tracking
-
- **Integrations** — Fixed Ollama infinite subprocess spawning, Graphiti import paths, OpenRouter API URL suffix, and GitLab authentication bugs
-
- **Settings & configuration** — Corrected .auto-claude path discovery timeout, z.AI China preset URL, log order sorting, and onboarding completion state persistence
-
-### 📚 Documentation
-
- Added Awesome Claude Code badge to README
-
- Added instructions for resetting PR review state in CLAUDE.md
-
---
-
-## What's Changed
-
- fix: handle unknown SDK message types (rate_limit_event) to prevent session crashes by @AndyMik90 in 4a75ea9f9
- fix: PR review error visibility and gh CLI resolution in bundled apps by @AndyMik90 in 732fc1cd3
- fix: handle empty/greenfield projects in spec creation (#1426) (#1841) by @Andy in 819f98d9f
- fix: clear terminalEventSeen on task restart to prevent stuck-after-planning (#1828) (#1840) by @Andy in 28a620079
- fix: watch worktree path for implementation_plan.json changes (#1805) (#1842) by @Andy in fb3a3fbda
- fix: resolve Claude CLI not found on Windows - PATH, prompt size, cwd (#1661) (#1843) by @Andy in 76d1d3b03
- fix: handle planning phase crash and resume recovery (#1562) (#1844) by @Andy in 3cb05781f
- fix: show dismissed PR review findings in UI instead of silently dropping them (#1852) by @Andy in d98ff7d19
- fix: preserve file/line info in PR review extraction recovery (#1857) by @Andy in 635b53eea
- docs: add Awesome Claude Code badge to README (#1838) by @Andy in 2e4b5ac65
- test: achieve 100% test coverage for backend CLI commands (#1772) by @StillKnotKnown in 385f04414
- fix: cap terminal paste size to 1MB to prevent GPU context exhaustion by @AndyMik90 in 7b0f3a2c0
- fix: prevent OOM, orphaned agents, and unbounded growth during overnight builds (#1813) by @Andy in 4091d1d4b
- docs: add instructions for resetting PR review state in CLAUDE.md by @AndyMik90 in ecb615802
- auto-claude: 217-investigate-symlink-issues-in-work-tree-creation-f (#1808) by @Andy in ae13ce14c
- auto-claude: 218-enable-claude-code-features-in-worktree-terminals (#1809) by @Andy in e3b219288
- auto-claude: 219-investigate-and-fix-authentication-subscription-sy (#1810) by @Andy in 6204d5fc2
- feat(roadmap): add expand/collapse functionality for phase features (#1796) by @Burak in f735f0b49
- auto-claude: 216-display-ongoing-pr-review-logs-in-progress (#1807) by @Andy in a4870fa0c
- fix(pr-review): reduce structured output failures and preserve findings in recovery (#1806) by @Andy in f1b8cd3a7
- fix(sentry): enable Sentry for Python subprocesses and add diagnostic instrumentation (#1804) by @Andy in 4d4234378
- fix(pr-review): add three-tier recovery for structured output validation failure (#1797) by @Andy in d1fbccde3
- test: improve backend agent test coverage to 94% (#1779) by @StillKnotKnown in ed93df698
- fix(github): use UTC timestamps for reviewed_at to fix comment detection (#1795) by @Andy in 8872d33e3
- feat: add user-friendly GitHub API error handling (#1790) by @StillKnotKnown in 8ece0009e
- fix(roadmap): sync roadmap features with task lifecycle (#1791) by @Andy in 115576e85
- fix(github): resolve PR review hanging in bundled app (#1793) by @Andy in 3791b37bb
- feat(profiles): implement unified profile swapping across OAuth and API accounts (#1794) by @StillKnotKnown in 282387356
- test: improve backend memory system test coverage to 100% (#1780) by @StillKnotKnown in 4f1b7b2a9
- fix(ideation): guard against non-string properties in IdeaCard badges by @AndyMik90 in 5e78d748e
- fix(updater): convert HTML release notes to markdown before rendering by @AndyMik90 in aa5fc7f95
- fix(pr-review): simplify structured output schema to reduce validation failures (#1787) by @Andy in cd8914700
- fix(qa): enforce visual verification for UI changes and inject startup commands (#1784) by @Andy in f149a7fbd
- fix(plan-files): use atomic writes to prevent 0-byte corruption (#1785) by @Andy in c2245b812
- fix(terminal): make worktree dropdown scrollable and show all items by @AndyMik90 in 950da45e4
- auto-claude: subtask-1-1 - Add adaptive thinking badge to thinking level label (#1782) by @Andy in 25acf2826
- auto-claude: subtask-1-1 - Add overflow-hidden and break-words to subtask cards by @AndyMik90 in 39aa08872
- refactor(app-updater): disable automatic downloads and allow intentional downgrades by @AndyMik90 in 8de8039db
- fix(auth): detect auth errors in AI response text and prevent retry loops (#1776) by @Andy in f4788e4af
- test: achieve 100% coverage for backend core workspace module (#1774) by @StillKnotKnown in 3f95765cf
- fix(title-generator): add production path resolution for backend source (#1778) by @Andy in 923880f5b
- fix(fast-mode): use setting_sources instead of env var for CLI fast mode (#1771) by @Andy in 390ba6a58
- fix(windows): complete System32 executable path fixes for where.exe and taskkill.exe (#1715) by @VDT-91 in aa7f56e5d
- fix(worktree): remove auto-commit on deletion and add uncommitted changes warning by @AndyMik90 in cec8e65ee
- Smart PR Status Polling System (#1766) by @Andy in 48d5f7a32
- feat: simplify thinking system and remove opus-1m model variant (#1760) by @Andy in bb7e18937
- auto-claude: 203-fix-pr-review-ui-update-issue (#1732) by @Andy in 7589f8e4f
- auto-claude: subtask-2-1 - Create isAPIProfileAuthenticated() function to val (#1745) by @Andy in 57e38a692
- auto-claude: 202-fix-kanban-board-scaling-collisions (#1731) by @Andy in d09ebb850
- auto-claude: 204-fix-pr-review-ui-not-updating-without-manual-navig (#1734) by @Andy in 087091cef
- auto-claude: 203-fix-ui-not-updating-during-pr-review-operations (#1733) by @Andy in f085c08bd
- auto-claude: 205-fix-insights-chat-only-shows-last-task-suggestion- (#1735) by @Andy in f121f9cdd
- auto-claude: 197-roadmap-generation-stuck-at-50-file-locking-race-c (#1746) by @Andy in f41f15e59
- auto-claude: 193-fix-update-context7-mcp-tool-name-from-get-library (#1744) by @Andy in bdff9141a
- auto-claude: 192-changelog-generation-multiple-critical-bugs-tasks- (#1725) by @Andy in 8c9a504df
- auto-claude: 194-bug-rate-limit-during-task-execution-causes-subtas (#1726) by @Andy in 8a7443d24
- auto-claude: 201-bug-pr-review-logs-and-analysis (#1730) by @Andy in e0d53adb4
- auto-claude: 196-fix-worktrees-dialog-auto-close-race-condition-and (#1727) by @Andy in 323b0d3be
- auto-claude: 199-bug-logs-disappear-after-restart (#1728) by @Andy in d639f6ef8
- auto-claude: 198-critical-oauth-token-revocation-causes-infinite-40 (#1747) by @Andy in 4438c0b10
- Fix Panel Constraints Error During Terminal Exit (#1757) by @Andy in 32bf353da
- auto-claude: 190-bug-context-page-crash-multiple-root-causes-when-v (#1724) by @Andy in 2db36982f
- feat: add search/filter to WorktreeSelector dropdown (#1754) by @Andy in 09f059ca3
- fix(terminal): push worktree branch to remote with tracking on creation (#1753) by @Andy in b5de0d9ff
- auto-claude: 189-subtask-execution-stuck-in-infinite-retry-loop-whe (#1723) by @Andy in 445da186c
- auto-claude: 188-terminal-claude-sessions-require-manual-click-to-r (#1743) by @Andy in f8499e965
- auto-claude: 200-bug-changelog-and-release-generation (#1729) by @Andy in 826583b82
- fix(terminal): use each terminal's cwd for invoke Claude all button (#1756) by @Andy in ac4fe4f42
- feat(terminal): read Claude Code CLI settings and inject env vars into PTY sessions (#1750) by @Andy in 152e54093
- fix: correct .auto-claude path mismatch causing discovery phase timeout (#1748) by @VDT-91 in 2c2a8a754
- fix: remove incorrect /v1 suffix from OpenRouter API URL (#1749) by @StillKnotKnown in 7e799ee57
- fix: prevent terminal worktree crash with race condition fixes (#1586) (#1658) by @VDT-91 in 216b58bcf
- fix: correct log order sorting and add configurable log order setting (#1720) by @Burak in 2e2b82365
- fix(ollama): stop infinite subprocess spawning from useEffect re-render loop (#1716) by @Quentin Veys in acb131b72
- fix(graphiti): migrate graphiti_memory imports to canonical paths (#1714) by @Quentin Veys in df528f065
- fix: improve auto-updater for beta releases and DMG installs (#1681) by @Andy in ff91a1af0
- feat: unified operation registry for intelligent auth/rate limit recovery (#1698) by @Andy in 6d0222fa9
- fix: Prevent stale worktree data from overriding correct task status (#1710) by @Burak in fe08c644c
- feat: add subscriptionType and rateLimitTier to ClaudeProfile (#1688) by @Andy in a5e3cc9a2
- auto-claude: subtask-1-1 - Add useTaskStore import and update task state after successful PR creation (#1683) by @Andy in 4587162e4
- auto-claude: 182-implement-pagination-and-filtering-for-github-pr-l (#1654) by @Andy in b4e6b2fe4
- auto-claude: 181-add-expand-button-for-long-task-descriptions (#1653) by @Andy in d9cd300fe
- fix(terminal): resolve text alignment issues on expand/minimize (#1650) by @VDT-91 in f5a7e26d9
- fix(windows): use full path to where.exe for reliable executable lookup (#1659) by @VDT-91 in 5f63daa3c
- fix: resolve ideation stuck at 3/6 types bug (#1660) by @VDT-91 in e6e8da17c
- Clarify Local and Origin Branch Distinction (#1652) by @Andy in 9317148b6
- auto-claude: 186-set-default-dark-mode-on-startup (#1656) by @Andy in 473020621
- auto-claude: subtask-1-1 - Add min-h-0 to enable scrolling in Roadmap tabs (#1655) by @Andy in ae703be9f
- fix: XState status lifecycle & cross-project contamination fixes (#1647) by @kaigler in 5293fb399
- refactor(frontend): complete XState task state machine migration (#1338) (#1575) by @kaigler in e2f9abadb
- Merge conflict resolution progress bar and log viewer (#1620) by @Andy in d16be3077
- fix: align Linux package builds (AppImage/deb/Flatpak) with target-specific extraResources (#1623) by @StillKnotKnown in bad1a9b2c
- Fix/gitlab bugs (#1519 and #1521) (#1544) by @bu5hm4nn in cd423c65c
- feat(kanban): add bulk task delete and worktree cleanup improvements (#1588) by @kaigler in 02ed91c91
- fix: add worktree isolation warning to prevent agent escape (#1528) by @kaigler in fe5cc582b
- feat(ui): add spell check support for text inputs (#1304) by @kaigler in 8f02a5129
- fix(windows): complete Windows credential fixes with path normalization (#1585) by @kaigler in 1e1997167
- AI-Powered GitHub PR Template Generation (#1618) by @Andy in 900dd4360
- Fix pty.node SIGABRT crash on macOS shutdown (#1619) by @Andy in f355e09d7
- fix(merge): use git merge for diverged branches with progress tracking (#1605) by @Andy in bde2ca4b2
- Surface Billing/Credit Exhaustion Errors to UI (Issue #1580) (#1617) by @Andy in 7bf12e856
- auto-claude: subtask-1-1 - Change $teamId type from ID! to String! in the team query (#1627) by @Andy in 54d0cd2f4
- fix(auth): support API profile mode without OAuth requirement (#1616) by @StillKnotKnown in f8cc63af4
- fix: agent retry loop for tool concurrency errors (#1546) [v3] (#1606) by @Michael Ludlow in 0aea4fb5e
- fix(queue): enforce max parallel tasks and auto-refresh UI (#1594) by @Andy in 4070a4c29
- Persist Kanban column collapse state per project via main process (#1579) by @Andy in a1114664e
- feat(pr-review): evidence-based validation and trigger-driven exploration (#1593) by @Andy in bfc232825
- fix(ui): smart auto-scroll for Insights streaming responses (#1591) by @kaigler in eee97e7ea
- fix(changelog): validate Claude CLI exists before generation (#1305) by @kaigler in c1f24c07f
- auto-claude: subtask-1-1 - Add min-w-0 class to subtask title row flex container (#1578) by @Andy in 286591c02
- auto-claude: subtask-1-1 - Remove Popover wrapper and related functionality from ClaudeCodeStatusBadge (#1566) by @Andy in 8d18cc81a
- fix(claude-profile): preserve subscriptionType and rateLimitTier during token refresh (#1556) by @Andy in 52e426a48
- auto-claude: subtask-1-1 - Update cancelReview callback to handle both success and failure cases (#1551) by @Andy in d8f00fe5a
- fix(backend): prioritize git remote detection over env var for repo (#1555) by @Andy in 9b07ed464
- fix(backend): handle detached HEAD state when pushing branch for PR creation (#1560) by @Andy in 2b72694d0
- fix: add explicit UTF-8 encoding across all Electron main process I/O (#1554) by @Andy in 4243530e9
- fix(backend): pass OAuth token to Python subprocess for authentication by @AndyMik90 in 6f1002dd7
- perf(frontend): async parallel worktree listing to prevent UI freezes (#1553) by @Andy in 399a7e736
- auto-claude: subtask-1-1 - Remove amber lock indicator line from kanban resize handle (#1557) by @Andy in 83a64b88e
- fix(frontend): resolve TerminalFontSettings infinite re-render loop (#1536) by @StillKnotKnown in 1c6266025
- fix(frontend): respect hasCompletedOnboarding from ~/.claude.json (#1537) by @StillKnotKnown in 1860c2c43
- fix: prevent planner from generating invalid verification types (#1388) (#1529) by @kaigler in 94d941333
- fix(frontend): resolve Insights scroll-to-blank-space issue on macOS (ACS-382) (#1535) by @StillKnotKnown in 496b2b96a
- feat: add customizable terminal fonts with OS-specific defaults (#1412) by @StillKnotKnown in f289107b8
- Add dev mode screenshot capture warning (#1516) by @Andy in 16eeb301a
- fix: add worktree isolation warnings to prevent agent escape (ACS-394) (#1495) by @StillKnotKnown in 1e453653b
- fix: resolve flaky subprocess-spawn test on Windows CI (ACS-392) (#1494) by @StillKnotKnown in f6b264d56
- feat(task-logger): strip ANSI escape codes from logs and extend coverage (#1411) by @StillKnotKnown in 988ec0c25
- fix(frontend): use spawn() instead of exec() for Windows terminal launching (#1498) by @StillKnotKnown in 26c9083d3
- fix(api-profiles): correct z.AI China preset URL and rename provider presets (#1500) by @StillKnotKnown in 05cf0a516
- fix: validate branch pattern before worktree cleanup to prevent deleting wrong branch (#1493) by @StillKnotKnown in 8576754a1
- Real-Time Updates for Insights Chat (#1511) by @Andy in d940b6ade
- Fix Terminal UI Rendering Issues (#1514) by @Andy in 8d8306b8e
- Fix terminal content resizing on expansion (#1512) by @Andy in 9f6c0026b
- Restore Terminal Session History on App Restart (#1515) by @Andy in 63e2847fc
- Move Reference Images Above Task Title & Fix Image Display Issues (#1513) by @Andy in b269ac305
- auto-claude: 143-fix-github-integration-ui-refresh-issues (#1467) by @Andy in aa2cb4fa6
- feat: Multi-profile account swapping with token refresh and queue routing (#1496) by @Andy in 1e72c8d77
- Simplified Testing Strategy for Regression Prevention (#1379) by @Andy in ae4e48e8b
- auto-claude: 152-persist-tasks-during-roadmap-regeneration (#1463) by @Andy in 9bd3d7e3b
- Debug Kanban Memory & Add Sentry Monitoring (#1380) by @Andy in bc5f550ee
- auto-claude: 147-remove-outdated-compatibility-shims (#1465) by @Andy in 53111dbb9
- auto-claude: 162-fix-worktree-error-on-repeated-task-starts (#1453) by @Andy in b955badf7
- auto-claude: 155-fix-pr-list-diff-display-metrics (#1458) by @Andy in 31f116db5
- auto-claude: 151-fix-pr-review-agent-token-refresh-on-account-swap (#1456) by @Andy in d081af042
- auto-claude: 148-add-progress-persistence-and-status-indicators (#1464) by @Andy in 4937d5745
- auto-claude: 154-fix-task-modal-conflict-check-status-refresh (#1462) by @Andy in 0299009df
- auto-claude: 153-widen-kanban-columns-and-add-collapse-feature (#1457) by @Andy in d65973075
- auto-claude: subtask-1-1 - Add filter after map operation to remove empty str (#1466) by @Andy in 783f0fe0e
- fix: add formatReleaseNotes helper for markdown changelog rendering (#1468) by @Andy in 43a97e1b3
- feat(sidebar): add collapsible sidebar toggle (#1501) by @Michael Ludlow in d17c17887
- fix(auth): check .credentials.json for Linux profile authentication (#1492) by @StillKnotKnown in 8d2f66291
- auto-claude: subtask-1-1 - Replace ReleaseNotesRenderer with ReactMarkdown (#1454) by @Andy in 1185a558c
- auto-claude: 156-fix-electron-app-version-detection-bug (#1459) by @Andy in 9a3b48c25
- auto-claude: subtask-1-1 - Add --no-track flag to git worktree add command (#1455) by @Andy in 0c2990815
- auto-claude: subtask-1-1 - Change task.specId to taskId in 3 startSpecCreation calls (#1461) by @Andy in 91edc0e14
- fix(onboarding): align MemoryStep layout with Settings MemoryBackendSection (#1445) by @Michael Ludlow in e9de26d59
- auto-claude: subtask-1-1 - Add metadata?.requireReviewBeforeCoding check (#1460) by @Andy in 426d56571
- fix: use API profile environment variables for task title generation (#1471) by @JoshuaRileyDev in c5a0f042d
- fix(auth): Long-lived OAuth authentication with multi-profile usage display (#1443) by @Andy in 12e788417
- feat: Add screenshot capture to task creation modal (#1429) by @JoshuaRileyDev in 1a2a1b1fc
- fix: prevent queue settings modal from disappearing when tasks change (#1430) by @JoshuaRileyDev in 33acc1430
- feat: Queue System v2 with Auto-Promotion and Smart Task Management (#1203) by @JoshuaRileyDev in 3b87e24d7
- feat: Add API profile providers usage endpoints support (#1279) by @StillKnotKnown in cfe7dedd0
-
-## Thanks to all contributors
-
-@AndyMik90, @Andy, @Burak, @StillKnotKnown, @VDT-91, @kaigler, @Michael Ludlow, @JoshuaRileyDev, @Quentin Veys, @bu5hm4nn
-
 ## 2.7.5 - Security & Platform Improvements

 ### ✨ New Features
@@ -1246,17 +1015,17 @@
 - feat(python): bundle Python 3.12 with packaged Electron app (#284) by @Andy in 7f19c2e1
 - fix: resolve spawn python ENOENT error on Linux by using getAugmentedEnv() (#281) by @Todd W. Bucy in d98e2830
 - fix(ci): add write permissions to beta-release update-version job by @AndyMik90 in 0b874d4b
- chore(deps): bump @xterm/xterm from 5.5.0 to 6.0.0 in /apps/desktop (#270) by @dependabot[bot] in 50dd1078
+- chore(deps): bump @xterm/xterm from 5.5.0 to 6.0.0 in /apps/frontend (#270) by @dependabot[bot] in 50dd1078
 - fix(github): resolve follow-up review API issues by @AndyMik90 in f1cc5a09
 - fix(security): resolve CodeQL file system race conditions and unused variables (#277) by @Andy in b005fa5c
 - fix(ci): use correct electron-builder arch flags (#278) by @Andy in d79f2da4
- chore(deps): bump jsdom from 26.1.0 to 27.3.0 in /apps/desktop (#268) by @dependabot[bot] in 5ac566e2
- chore(deps): bump typescript-eslint in /apps/desktop (#269) by @dependabot[bot] in f49d4817
+- chore(deps): bump jsdom from 26.1.0 to 27.3.0 in /apps/frontend (#268) by @dependabot[bot] in 5ac566e2
+- chore(deps): bump typescript-eslint in /apps/frontend (#269) by @dependabot[bot] in f49d4817
 - fix(ci): use develop branch for dry-run builds in beta-release workflow (#276) by @Andy in 1e1d7d9b
 - fix: accept bug_fix workflow_type alias during planning (#240) by @Daniel Frey in e74a3dff
 - fix(paths): normalize relative paths to posix (#239) by @Daniel Frey in 6ac8250b
- chore(deps): bump @electron/rebuild in /apps/desktop (#271) by @dependabot[bot] in a2cee694
- chore(deps): bump vitest from 4.0.15 to 4.0.16 in /apps/desktop (#272) by @dependabot[bot] in d4cad80a
+- chore(deps): bump @electron/rebuild in /apps/frontend (#271) by @dependabot[bot] in a2cee694
+- chore(deps): bump vitest from 4.0.15 to 4.0.16 in /apps/frontend (#272) by @dependabot[bot] in d4cad80a
 - feat(github): add automated PR review with follow-up support (#252) by @Andy in 596e9513
 - ci: implement enterprise-grade PR quality gates and security scanning (#266) by @Alex in d42041c5
 - fix: update path resolution for ollama_model_detector.py in memory handlers (#263) by @delyethan in a3f87540
@@ -1526,17 +1295,17 @@
 - feat(python): bundle Python 3.12 with packaged Electron app (#284) by @Andy in 7f19c2e1
 - fix: resolve spawn python ENOENT error on Linux by using getAugmentedEnv() (#281) by @Todd W. Bucy in d98e2830
 - fix(ci): add write permissions to beta-release update-version job by @AndyMik90 in 0b874d4b
- chore(deps): bump @xterm/xterm from 5.5.0 to 6.0.0 in /apps/desktop (#270) by @dependabot[bot] in 50dd1078
+- chore(deps): bump @xterm/xterm from 5.5.0 to 6.0.0 in /apps/frontend (#270) by @dependabot[bot] in 50dd1078
 - fix(github): resolve follow-up review API issues by @AndyMik90 in f1cc5a09
 - fix(security): resolve CodeQL file system race conditions and unused variables (#277) by @Andy in b005fa5c
 - fix(ci): use correct electron-builder arch flags (#278) by @Andy in d79f2da4
- chore(deps): bump jsdom from 26.1.0 to 27.3.0 in /apps/desktop (#268) by @dependabot[bot] in 5ac566e2
- chore(deps): bump typescript-eslint in /apps/desktop (#269) by @dependabot[bot] in f49d4817
+- chore(deps): bump jsdom from 26.1.0 to 27.3.0 in /apps/frontend (#268) by @dependabot[bot] in 5ac566e2
+- chore(deps): bump typescript-eslint in /apps/frontend (#269) by @dependabot[bot] in f49d4817
 - fix(ci): use develop branch for dry-run builds in beta-release workflow (#276) by @Andy in 1e1d7d9b
 - fix: accept bug_fix workflow_type alias during planning (#240) by @Daniel Frey in e74a3dff
 - fix(paths): normalize relative paths to posix (#239) by @Daniel Frey in 6ac8250b
- chore(deps): bump @electron/rebuild in /apps/desktop (#271) by @dependabot[bot] in a2cee694
- chore(deps): bump vitest from 4.0.15 to 4.0.16 in /apps/desktop (#272) by @dependabot[bot] in d4cad80a
+- chore(deps): bump @electron/rebuild in /apps/frontend (#271) by @dependabot[bot] in a2cee694
+- chore(deps): bump vitest from 4.0.15 to 4.0.16 in /apps/frontend (#272) by @dependabot[bot] in d4cad80a
 - feat(github): add automated PR review with follow-up support (#252) by @Andy in 596e9513
 - ci: implement enterprise-grade PR quality gates and security scanning (#266) by @Alex in d42041c5
 - fix: update path resolution for ollama_model_detector.py in memory handlers (#263) by @delyethan in a3f87540
@@ -2,9 +2,9 @@

 This file provides guidance to Claude Code when working with this repository.

-Auto Claude is an autonomous multi-agent coding framework that plans, builds, and validates software for you. It's a TypeScript-first Electron desktop application with a self-contained AI agent layer (Vercel AI SDK v6). A lightweight Python sidecar provides the optional Graphiti memory system.
+Auto Claude is an autonomous multi-agent coding framework that plans, builds, and validates software for you. It's a monorepo with a Python backend (CLI + agent logic) and an Electron/React frontend (desktop UI).

-> **Deep-dive reference:** [ARCHITECTURE.md](shared_docs/ARCHITECTURE.md) | **Frontend contributing:** [apps/desktop/CONTRIBUTING.md](apps/desktop/CONTRIBUTING.md)
+> **Deep-dive reference:** [ARCHITECTURE.md](shared_docs/ARCHITECTURE.md) | **Frontend contributing:** [apps/frontend/CONTRIBUTING.md](apps/frontend/CONTRIBUTING.md)

 ## Product Overview

@@ -30,55 +30,27 @@ Auto Claude is a desktop application (+ CLI) where users describe a goal and AI

 ## Critical Rules

-**Vercel AI SDK only** — All AI interactions use the Vercel AI SDK v6 (`ai` package) via the TypeScript agent layer in `apps/desktop/src/main/ai/`. NEVER use `@anthropic-ai/sdk` or `anthropic.Anthropic()` directly. Use `createProvider()` from `ai/providers/factory.ts` and `streamText()`/`generateText()` from the `ai` package. Provider-specific adapters (e.g., `@ai-sdk/anthropic`, `@ai-sdk/openai`) are managed through the provider registry.
+**Claude Agent SDK only** — All AI interactions use `claude-agent-sdk`. NEVER use `anthropic.Anthropic()` directly. Always use `create_client()` from `core.client`.

-**i18n required** — All frontend user-facing text uses `react-i18next` translation keys. Hardcoded strings in JSX/TSX break localization for non-English users. Add keys to both `en/*.json` and `fr/*.json`.
+**i18n required** — All frontend user-facing text MUST use `react-i18next` translation keys. Never hardcode strings in JSX/TSX. Add keys to both `en/*.json` and `fr/*.json`.

-**Platform abstraction** — Never use `process.platform` directly. Import from `apps/desktop/src/main/platform/`. CI tests all three platforms.
+**Platform abstraction** — Never use `process.platform` directly. Import from `apps/frontend/src/main/platform/` or `apps/backend/core/platform/`. CI tests all three platforms.

-**No time estimates** — Provide priority-based ordering instead of duration predictions.
+**No time estimates** — Never provide duration predictions. Use priority-based ordering instead.

-**PR target** — Always target the `develop` branch for PRs, not `main`. Main is reserved for releases.
+**PR target** — Always target the `develop` branch for PRs to AndyMik90/Auto-Claude, NOT `main`.

-**No console.log in production code** — `console.log` output is invisible in bundled Electron apps. Use Sentry for error tracking in production; reserve `console.log` for development only.
+## Work Approach

-## Work Approach: Orchestrator-First
+**Investigate before speculating** — Always read the actual code before proposing root causes. Spawn agents to grep and read relevant source files before forming any hypothesis. Never guess at causes without evidence from the codebase.

-You are an orchestrator. Your primary role is to understand what needs to be done, break it into workstreams, and delegate execution to agent teams. This keeps your context window focused on coordination and decision-making rather than filling up with implementation details.
+**Spawn agents for complex tasks** — When tackling complex tasks, spawn sub-agents/agent teams immediately rather than trying to handle everything in a single context window. Never attempt to analyze large codebases or multiple features monolithically.

-<orchestrator_pattern>
-When given a task, follow this pattern:
-
-1. **Investigate first** — Read the actual code before forming any hypothesis. Use targeted searches (Glob, Grep, Read) for simple lookups. For broader exploration, spawn an Explore agent.
-
-2. **Plan the approach** — Identify what needs to change, which files are involved, and whether work can be parallelized. For multi-step tasks, create a task list to track workstreams.
-
-3. **Delegate execution** — Spawn agent teams to do the implementation work. Each agent gets a clear, self-contained assignment with all the context it needs: relevant file paths, the specific change to make, and acceptance criteria. Run independent workstreams in parallel.
-
-4. **Verify and integrate** — Review agent outputs, run tests, and ensure changes work together. Fix integration issues or spawn follow-up agents as needed.
-</orchestrator_pattern>
-
-**When to delegate vs. do directly:**
- Delegate: multi-file changes, research across the codebase, independent parallel workstreams, tasks that would consume significant context
- Do directly: single-file edits, simple bug fixes, quick lookups, tasks where you already have the context
-
-**Giving agents good assignments** — Each agent works with a fresh context. Include: the specific goal, relevant file paths, code patterns to follow, and what "done" looks like. Agents perform better with explicit, complete instructions than with vague references to "the current task."
-
-**Minimal changes only** — Prefer the simplest approach (e.g., prompt-only changes, single guard clause) before suggesting multi-component solutions. If the user asks for X, implement X — don't bundle additional fixes they didn't request.
-
-**Default to action** — When the user's intent implies making changes, implement them rather than only suggesting. If something is unclear, read the relevant code to fill in the gaps rather than asking. Only ask when genuine ambiguity remains about what the user wants.
-
-## Context Management
-
-Your context window will be automatically compacted as it approaches its limit, allowing you to continue working indefinitely. Do not stop tasks early due to context concerns — instead, persist progress and keep going.
-
-**For long-running tasks:** Use git commits, task lists, and structured notes to track state. When context compacts, review git log and any progress files to re-orient. Focus on incremental progress — complete one component before moving to the next, and commit working states along the way.
-
-**Parallel tool calls** — When reading multiple files, running independent searches, or executing unrelated commands, make all calls in parallel rather than sequentially. This significantly speeds up investigation and implementation.
+**Minimal fixes only** — Prefer the simplest approach (e.g., prompt-only changes, single guard clause) before suggesting multi-component solutions. If the user asks for X, implement X — don't bundle additional fixes they didn't request.

 ## Known Gotchas

-**Electron path resolution** — For bug fixes in the Electron app, check path resolution differences between dev and production builds (`app.isPackaged`, `process.resourcesPath`). Paths that work in dev often break when Electron is bundled for production — verify both contexts.
+**Electron path resolution** — For bug fixes in the Electron app, always check path resolution differences between dev and production builds (`app.isPackaged`, `process.resourcesPath`). Paths that work in dev often break when Electron is bundled for production — verify both contexts.

 ### Resetting PR Review State

@@ -88,34 +60,34 @@ To fully clear all PR review data so reviews run fresh, delete/reset these three
 2. `rm .auto-claude/github/pr/review_*.json` — review result files
 3. Reset `pr/index.json` to `{"reviews": [], "last_updated": null}`
 4. Reset `bot_detection_state.json` to `{"reviewed_commits": {}}` — this is the gatekeeper; without clearing it, the bot detector skips already-seen commits
-
 ## Project Structure

 ```
 autonomous-coding/
 ├── apps/
-│   └── desktop/                 # Electron desktop application (sole app)
-│       ├── prompts/             # Agent system prompts (.md)
+│   ├── backend/                 # Python backend/CLI — ALL agent logic
+│   │   ├── core/                # client.py, auth.py, worktree.py, platform/
+│   │   ├── security/            # Command allowlisting, validators, hooks
+│   │   ├── agents/              # planner, coder, session management
+│   │   ├── qa/                  # reviewer, fixer, loop, criteria
+│   │   ├── spec/                # Spec creation pipeline
+│   │   ├── cli/                 # CLI commands (spec, build, workspace, QA)
+│   │   ├── context/             # Task context building, semantic search
+│   │   ├── runners/             # Standalone runners (spec, roadmap, insights, github)
+│   │   ├── services/            # Background services, recovery orchestration
+│   │   ├── integrations/        # graphiti/, linear, github
+│   │   ├── project/             # Project analysis, security profiles
+│   │   ├── merge/               # Intent-aware semantic merge for parallel agents
+│   │   └── prompts/             # Agent system prompts (.md)
+│   └── frontend/                # Electron desktop UI
 │       └── src/
 │           ├── main/            # Electron main process
-│           │   ├── ai/          # TypeScript AI agent layer (Vercel AI SDK v6)
-│           │   │   ├── providers/   # Multi-provider registry + factory (9+ providers)
-│           │   │   ├── tools/       # Builtin tools (Read, Write, Edit, Bash, Glob, Grep, etc.)
-│           │   │   ├── security/    # Bash validator, command parser, path containment
-│           │   │   ├── config/      # Agent configs (25+ types), phase config, model resolution
-│           │   │   ├── session/     # streamText() agent loop, error classification, progress
-│           │   │   ├── agent/       # Worker thread executor + bridge
-│           │   │   ├── orchestration/ # Build pipeline (planner → coder → QA)
-│           │   │   ├── runners/     # Utility runners (insights, roadmap, PR review, etc.)
-│           │   │   ├── mcp/         # MCP client integration
-│           │   │   ├── client/      # Client factory convenience constructors
-│           │   │   └── auth/        # Token resolution (reuses claude-profile/)
 │           │   ├── agent/       # Agent queue, process, state, events
 │           │   ├── claude-profile/ # Multi-profile credentials, token refresh, usage
 │           │   ├── terminal/    # PTY daemon, lifecycle, Claude integration
 │           │   ├── platform/    # Cross-platform abstraction
 │           │   ├── ipc-handlers/# 40+ handler modules by domain
-│           │   ├── services/    # Session recovery, profile service
+│           │   ├── services/    # SDK session recovery, profile service
 │           │   └── changelog/   # Changelog generation and formatting
 │           ├── preload/         # Electron preload scripts (electronAPI bridge)
 │           ├── renderer/        # React UI
@@ -132,6 +104,7 @@ autonomous-coding/
 │           │   └── utils/       # ANSI sanitizer, shell escape, provider detection
 │           └── types/           # TypeScript type definitions
 ├── guides/                      # Documentation
+├── tests/                       # Backend test suite
 └── scripts/                     # Build and utility scripts
 ```

@@ -141,15 +114,18 @@ autonomous-coding/
 ```bash
 npm run install:all              # Install all dependencies from root
 # Or separately:
-cd apps/desktop && npm install
+cd apps/backend && uv venv && uv pip install -r requirements.txt
+cd apps/frontend && npm install
 ```

 ### Testing

 | Stack | Command | Tool |
 |-------|---------|------|
-| Frontend unit | `cd apps/desktop && npm test` | Vitest |
-| Frontend E2E | `cd apps/desktop && npm run test:e2e` | Playwright |
+| Backend | `apps/backend/.venv/bin/pytest tests/ -v` | pytest |
+| Frontend unit | `cd apps/frontend && npm test` | Vitest |
+| Frontend E2E | `cd apps/frontend && npm run test:e2e` | Playwright |
+| All backend | `npm run test:backend` (from root) | pytest |

 ### Releases
 ```bash
@@ -159,53 +135,15 @@ git push && gh pr create --base main             # PR to main triggers release

 See [RELEASE.md](RELEASE.md) for full release process.

-## AI Agent Layer (`apps/desktop/src/main/ai/`)
+## Backend Development

-All AI agent logic lives in TypeScript using the Vercel AI SDK v6. This replaces the previous Python `claude-agent-sdk` integration.
+### Claude Agent SDK Usage

-### Architecture Overview
+Client: `apps/backend/core/client.py` — `create_client()` returns a configured `ClaudeSDKClient` with security hooks, tool permissions, and MCP server integration.

- **Provider Layer** (`providers/`) — Multi-provider support via `createProviderRegistry()`. Supports Anthropic, OpenAI, Google, Bedrock, Azure, Mistral, Groq, xAI, and Ollama. Provider-specific transforms handle thinking token normalization and prompt caching.
- **Session Runtime** (`session/`) — `runAgentSession()` uses `streamText()` with `stopWhen: stepCountIs(N)` for agentic tool-use loops. Includes error classification (429/401/400) and progress tracking.
- **Worker Threads** (`agent/`) — Agent sessions run in `worker_threads` to avoid blocking the Electron main process. The `WorkerBridge` relays `postMessage()` events to the existing `AgentManagerEvents` interface.
- **Build Orchestration** (`orchestration/`) — Full planner → coder → QA pipeline. Parallel subagent execution via `Promise.allSettled()`.
- **Tools** (`tools/`) — 8 builtin tools (Read, Write, Edit, Bash, Glob, Grep, WebFetch, WebSearch) defined with Zod schemas via AI SDK `tool()`.
- **Security** (`security/`) — Bash validator, command parser, and path containment ported from Python with identical allowlist behavior.
- **Config** (`config/`) — `AGENT_CONFIGS` registry (25+ agent types), phase-aware model resolution, thinking budgets.
+Model and thinking level are user-configurable (via the Electron UI settings or CLI override). Use `phase_config.py` helpers to resolve the correct values

-### Key Patterns
-
-```typescript
-// Agent session using streamText()
-import { streamText, stepCountIs } from 'ai';
-
-const result = streamText({
-  model: provider,
-  system: systemPrompt,
-  messages: conversationHistory,
-  tools: toolRegistry.getToolsForAgent(agentType),
-  stopWhen: stepCountIs(1000),
-  onStepFinish: ({ toolCalls, text, usage }) => {
-    progressTracker.update(toolCalls, text);
-  },
-});
-
-// Tool definition with Zod schema
-import { tool } from 'ai';
-import { z } from 'zod';
-
-const readTool = tool({
-  description: 'Read a file from the filesystem',
-  inputSchema: z.object({
-    file_path: z.string(),
-    offset: z.number().optional(),
-    limit: z.number().optional(),
-  }),
-  execute: async ({ file_path, offset, limit }) => { /* ... */ },
-});
-```
-
-### Agent Prompts (`apps/desktop/prompts/`)
+### Agent Prompts (`apps/backend/prompts/`)

 | Prompt | Purpose |
 |--------|---------|
@@ -221,13 +159,24 @@ Each spec in `.auto-claude/specs/XXX-name/` contains: `spec.md`, `requirements.j

 ### Memory System (Graphiti)

-Graph-based semantic memory accessed via a Python MCP sidecar (lives outside `apps/desktop/`). The AI layer connects to it via `createMCPClient` from `@ai-sdk/mcp`. Configured through the Electron app's onboarding/settings UI. See [ARCHITECTURE.md](shared_docs/ARCHITECTURE.md#memory-system) for details.
+Graph-based semantic memory in `integrations/graphiti/`. Configured through the Electron app's onboarding/settings UI (CLI users can alternatively set `GRAPHITI_ENABLED=true` in `.env`). See [ARCHITECTURE.md](shared_docs/ARCHITECTURE.md#memory-system) for details.
+
+### Opus 4.6 Features
+
+Auto Claude leverages Opus 4.6's advanced capabilities for GitHub issue investigations:
+
+- **Fast Mode:** 2.5x faster investigations (toggle in Settings > GitHub > AI Investigation)
+- **128K Output Tokens:** Root cause specialist gets max tokens for deep analysis
+- **Per-Specialist Limits:** Different token limits per investigation specialist
+- **Adaptive Thinking:** High-effort mode for thorough investigations
+
+See [guides/opus-4.6-features.md](guides/opus-4.6-features.md) for detailed documentation on Opus 4.6 features, pricing, and usage.

 ## Frontend Development

 ### Tech Stack

-React 19, TypeScript (strict), Electron 39, Vercel AI SDK v6, Zustand 5, Tailwind CSS v4, Radix UI, xterm.js 6, Vite 7, Vitest 4, Biome 2, Motion (Framer Motion)
+React 19, TypeScript (strict), Electron 39, Zustand 5, Tailwind CSS v4, Radix UI, xterm.js 6, Vite 7, Vitest 4, Biome 2, Motion (Framer Motion)

 ### Path Aliases (tsconfig.json)

@@ -273,9 +222,9 @@ Main ↔ Renderer communication via Electron IPC:

 The frontend manages agent lifecycle end-to-end:
 - **`agent-queue.ts`** — Queue routing, prioritization, spec number locking
- **`agent-process.ts`** — Spawns worker threads via `WorkerBridge` for agent execution
+- **`agent-process.ts`** — Spawns and manages agent subprocess communication
 - **`agent-state.ts`** — Tracks running agent state and status
- **`agent-events.ts`** — Agent lifecycle events and state transitions (structured events from worker threads)
+- **`agent-events.ts`** — Agent lifecycle events and state transitions

 ### Claude Profile System (`src/main/claude-profile/`)

@@ -301,10 +250,13 @@ Full PTY-based terminal integration:
 - **Pre-commit:** Husky + lint-staged runs Biome on staged `.ts/.tsx/.js/.jsx/.json`
 - **Testing:** Vitest + React Testing Library + jsdom

+### Backend
+- **Linting:** Ruff
+- **Testing:** pytest (`apps/backend/.venv/bin/pytest tests/ -v`)

 ## i18n Guidelines

-All frontend UI text uses `react-i18next`. Translation files: `apps/desktop/src/shared/i18n/locales/{en,fr}/*.json`
+All frontend UI text uses `react-i18next`. Translation files: `apps/frontend/src/shared/i18n/locales/{en,fr}/*.json`

 **Namespaces:** `common`, `navigation`, `settings`, `dialogs`, `tasks`, `errors`, `onboarding`, `welcome`

@@ -325,7 +277,7 @@ When adding new UI text: add keys to ALL language files, use `namespace:section.

 Supports Windows, macOS, Linux. CI tests all three.

-**Platform modules:** `apps/desktop/src/main/platform/`
+**Platform modules:** `apps/frontend/src/main/platform/` and `apps/backend/core/platform/`

 | Function | Purpose |
 |----------|---------|
@@ -334,21 +286,24 @@ Supports Windows, macOS, Linux. CI tests all three.
 | `findExecutable(name)` | Cross-platform executable lookup |
 | `requiresShell(command)` | `.cmd/.bat` shell detection (Win) |

-Use `findExecutable()` and `joinPaths()` instead of hardcoded paths. See [ARCHITECTURE.md](shared_docs/ARCHITECTURE.md#cross-platform-development) for extended guide.
+Never hardcode paths. Use `findExecutable()` and `joinPaths()`. See [ARCHITECTURE.md](shared_docs/ARCHITECTURE.md#cross-platform-development) for extended guide.

 ## E2E Testing (Electron MCP)

 QA agents can interact with the running Electron app via Chrome DevTools Protocol:

 1. Start app: `npm run dev:debug` (debug mode for AI self-validation via Electron MCP)
-2. Enable Electron MCP in settings
-3. QA runs automatically through the TypeScript agent pipeline
+2. Set `ELECTRON_MCP_ENABLED=true` in `apps/backend/.env`
+3. Run QA: `python run.py --spec 001 --qa`

 Tools: `take_screenshot`, `click_by_text`, `fill_input`, `get_page_structure`, `send_keyboard_shortcut`, `eval`. See [ARCHITECTURE.md](shared_docs/ARCHITECTURE.md#end-to-end-testing) for full capabilities.

 ## Running the Application

 ```bash
+# CLI only
+cd apps/backend && python run.py --spec 001
+
 # Desktop app
 npm start          # Production build + run
 npm run dev        # Development mode with HMR
@@ -1,348 +0,0 @@
-# Codex Rate Limit Monitoring — Full System Research
-
-> Temporary research file. Delete after implementation.
-
-## Table of Contents
-
-1. [Codex Usage API](#1-codex-usage-api)
-2. [Current System Architecture](#2-current-system-architecture)
-3. [Anthropic-Hardcoded Locations](#3-anthropic-hardcoded-locations)
-4. [Provider-Agnostic Parts (No Changes Needed)](#4-provider-agnostic-parts)
-5. [Implementation Plan](#5-implementation-plan)
-
---
-
-## 1. Codex Usage API
-
-**Sources:** OpenAI Codex source code (`github.com/openai/codex`, Rust codebase), CodexBar macOS app (`github.com/steipete/CodexBar`), Context7 Codex developer docs.
-
-### 1.1 Active Polling Endpoint
-
-```
-GET https://chatgpt.com/backend-api/wham/usage
-```
-
-Fallback (when base URL doesn't contain `/backend-api`):
-```
-GET {base_url}/api/codex/usage
-```
-
-**Required Headers:**
-```http
-Authorization: Bearer <access_token>
-ChatGPT-Account-Id: <account_id>
-Content-Type: application/json
-Accept: application/json
-```
-
- `access_token` — The OAuth access token from `auth.openai.com` (same token our `codex-oauth.ts` already obtains)
- `account_id` — Account UUID from OAuth token data. Stored in `~/.codex/auth.json` under `tokens.account_id`. Optional per CodexBar ("when available") but may be required.
-
-### 1.2 Response Schema
-
-From `codex-rs/codex-backend-openapi-models/src/models/rate_limit_status_payload.rs`:
-
-```json
-{
-  "plan_type": "plus",
-  "rate_limit": {
-    "allowed": true,
-    "limit_reached": false,
-    "primary_window": {
-      "used_percent": 96,
-      "limit_window_seconds": 18000,
-      "reset_after_seconds": 673,
-      "reset_at": 1730947200
-    },
-    "secondary_window": {
-      "used_percent": 70,
-      "limit_window_seconds": 604800,
-      "reset_after_seconds": 43200,
-      "reset_at": 1730980800
-    }
-  },
-  "credits": {
-    "has_credits": false,
-    "unlimited": true,
-    "balance": null
-  },
-  "additional_rate_limits": [
-    {
-      "limit_name": "codex_other",
-      "metered_feature": "codex_other",
-      "rate_limit": {
-        "allowed": true,
-        "limit_reached": false,
-        "primary_window": {
-          "used_percent": 70,
-          "limit_window_seconds": 3600,
-          "reset_after_seconds": 1800,
-          "reset_at": 1730947200
-        }
-      }
-    }
-  ]
-}
-```
-
- `primary_window` = 5h session (18000s). Maps to our `sessionPercent`.
- `secondary_window` = Weekly (604800s = 7d). Maps to our `weeklyPercent`.
- `reset_at` = Unix timestamp (seconds). Convert to ms for our `sessionResetTimestamp`/`weeklyResetTimestamp`.
- `plan_type` values: `guest`, `free`, `go`, `plus`, `pro`, `free_workspace`, `team`, `business`, `education`, `quorum`, `k12`, `enterprise`, `edu`
-
-### 1.3 Passive Headers (From API Responses)
-
-Rate limit data is also returned in HTTP response headers on every `/v1/responses` call:
-
-```
-x-codex-primary-used-percent         → float (e.g., "25.0")
-x-codex-primary-window-minutes       → integer (e.g., "300" for 5h)
-x-codex-primary-reset-at             → unix timestamp seconds
-x-codex-secondary-used-percent       → float (weekly)
-x-codex-secondary-window-minutes     → integer
-x-codex-secondary-reset-at           → unix timestamp seconds
-x-codex-credits-has-credits          → "true" or "false"
-x-codex-credits-unlimited            → "true" or "false"
-x-codex-credits-balance              → decimal string e.g. "9.99"
-```
-
-SSE event type `codex.rate_limits` also carries this data inline in streaming responses.
-
-### 1.4 Token Details
-
-Our `codex-oauth.ts` already uses the correct flow:
- **Client ID:** `app_EMoamEEZ73f0CkXaXp7hrann` (same as Codex CLI)
- **Auth endpoint:** `https://auth.openai.com/oauth/authorize`
- **Token endpoint:** `https://auth.openai.com/oauth/token`
- **Scopes:** `openid profile email offline_access`
- **Refresh:** `POST https://auth.openai.com/oauth/token` with `grant_type=refresh_token`
-
-**Missing:** `account_id` for the `ChatGPT-Account-Id` header. Options:
-1. Decode from the JWT access token
-2. Read from `~/.codex/auth.json` (`tokens.account_id`)
-3. Extract during OAuth token exchange (may be in response)
-4. Try without it first (optional per CodexBar docs)
-
---
-
-## 2. Current System Architecture
-
-### 2.1 Two Parallel Account Systems
-
-The app has TWO account management systems that don't fully integrate:
-
-**System A: Legacy Claude Profile Manager (Main Process)**
- `claude-profile-manager.ts` — Manages OAuth profiles, rate limits, usage, auto-swap
- `claude-profiles.json` — Stores profiles with `activeProfileId`, `accountPriorityOrder`
- `usage-monitor.ts` — Polls Anthropic's `/api/oauth/usage` endpoint every 30s
- `token-refresh.ts` — Refreshes tokens via `console.anthropic.com/v1/oauth/token`
- `rate-limit-detector.ts` — Detects rate limits, triggers auto-swap
- `profile-scorer.ts` — Scores profiles by availability for auto-swap
- **100% Anthropic-specific.** Only knows about Anthropic OAuth tokens, Anthropic endpoints, Anthropic keychain format.
-
-**System B: Multi-Provider Accounts (Renderer + Settings)**
- `ProviderAccount[]` in `settings-store.ts` — All connected accounts (any provider)
- `globalPriorityOrder: string[]` in AppSettings — Manual priority queue
- `useActiveProvider()` hook — First account in priority order = active
- **Provider-agnostic.** Works for all 10 providers. But has NO usage monitoring, NO auto-swap.
-
-**The gap:** System A handles usage monitoring + auto-swap but only for Anthropic. System B handles multi-provider accounts but has no usage awareness.
-
-### 2.2 Data Flow: Usage Polling
-
-```
-UsageMonitor.start() → 30s interval
-  ↓
-checkUsageAndSwap()
-  ├─ determineActiveProfile()           ← Hardcoded: defaults to anthropic baseUrl
-  ├─ getCredential()                    ← Hardcoded: reads from Anthropic keychain
-  │   └─ ensureValidToken(configDir)    ← Hardcoded: refreshes via Anthropic endpoint
-  ├─ fetchUsageViaAPI()                 ← Hardcoded: only allows anthropic/zai/zhipu domains
-  │   ├─ getUsageEndpoint(provider)     ← Only 3 providers configured
-  │   ├─ Add anthropic-specific headers ← if (provider === 'anthropic') add beta headers
-  │   └─ Parse response                ← Provider-specific normalization
-  ├─ emit('usage-updated')             → IPC 'claude:usageUpdated' → renderer
-  ├─ emit('all-profiles-usage-updated') → IPC 'claude:allProfilesUsageUpdated' → renderer
-  └─ checkThresholdsExceeded()
-     └─ performProactiveSwap()          ← Only swaps Anthropic profiles
-```
-
-### 2.3 Data Flow: Account Swapping
-
-**Manual swap (UI):**
-```
-User clicks account in UsageIndicator popover
-  → handleSwapAccount(accountId)
-  → setQueueOrder([accountId, ...rest])    ← Reorders globalPriorityOrder
-  → requestUsageUpdate()                   ← Refreshes usage display
-```
-
-**Automatic swap (rate limit hit):**
-```
-SDK operation fails with 429
-  → detectRateLimit(output)                ← Pattern: "Limit reached · resets..."
-  → recordRateLimitEvent(profileId)
-  → getBestAvailableProfileEnv()
-  → profileManager.setActiveProfile()      ← Only updates claude-profiles.json
-  → usageMonitor.getAllProfilesUsage()     ← Refreshes UI
-  ← Returns new profile env vars
-```
-
-**Problem:** Auto-swap updates `claude-profiles.json` but NOT `globalPriorityOrder`. The renderer's priority queue may be out of sync.
-
-### 2.4 UI Components
-
-| Component | What it shows | Provider-specific? |
-|---|---|---|
-| `AuthStatusIndicator` | Provider badge (OpenAI/Anthropic) + auth type label | Codex = green "Codex", Anthropic = orange "OAuth" |
-| `UsageIndicator` | Usage bars OR "Subscription" OR "Unlimited" | Anthropic OAuth = bars, Codex OAuth = "Subscription", API = "Unlimited" |
-| `ProviderAccountCard` | Account card in settings with usage bars | Shows usage bars only when `account.usage` populated (Anthropic only) |
-| `ProviderAccountsList` | All accounts grouped by provider | Generic, but re-auth routes differ per provider |
-| `AddAccountDialog` | OAuth flow + account creation | Different flows: Codex → `codexAuthLogin()`, Anthropic → `claudeAuthLoginSubprocess()` |
-| `ProviderSection` | Provider group with "Add" buttons | Button label: "Add Codex Subscription" vs "Add OAuth" |
-
-### 2.5 Type Naming
-
-Types use "Claude" prefix but are structurally generic:
-```typescript
-ClaudeUsageSnapshot    → { sessionPercent, weeklyPercent, resetTimestamps, profileId, ... }
-ClaudeUsageData        → { sessionUsagePercent, weeklyUsagePercent }
-ClaudeRateLimitEvent   → { type, hitAt, resetAt }
-ProfileUsageSummary    → { sessionPercent, weeklyPercent, availabilityScore, ... }
-AllProfilesUsage       → { activeProfile, allProfiles[], fetchedAt }
-```
-
-These types work perfectly for Codex data — same session/weekly model. No structural changes needed, just need to populate them.
-
---
-
-## 3. Anthropic-Hardcoded Locations
-
-### 3.1 CRITICAL — Must Change
-
-| File | Line(s) | What's hardcoded | What to do |
-|---|---|---|---|
-| `usage-monitor.ts:45-49` | `ALLOWED_USAGE_API_DOMAINS` | Only `api.anthropic.com`, `api.z.ai`, `open.bigmodel.cn` | Add `chatgpt.com` |
-| `usage-monitor.ts:60-73` | `PROVIDER_USAGE_ENDPOINTS` | Only anthropic/zai/zhipu paths | Add `{ provider: 'openai', usagePath: '/wham/usage' }` |
-| `usage-monitor.ts:662,1069,1346,1359` | `baseUrl: 'https://api.anthropic.com'` | Hardcoded fallback for all OAuth profiles | Detect provider from account, use `chatgpt.com/backend-api` for Codex |
-| `usage-monitor.ts:1424` | `if (provider === 'anthropic')` adds beta headers | Anthropic-specific `anthropic-beta` header | Add `else if (provider === 'openai')` to add `ChatGPT-Account-Id` header |
-| `token-refresh.ts:31` | `ANTHROPIC_TOKEN_ENDPOINT = 'https://console.anthropic.com/v1/oauth/token'` | Only Anthropic refresh endpoint | Route to `auth.openai.com/oauth/token` for Codex |
-| `token-refresh.ts:37` | `CLAUDE_CODE_CLIENT_ID = '9d1c250a-...'` | Only Anthropic client ID | Use `app_EMoamEEZ73f0CkXaXp7hrann` for Codex |
-| `UsageIndicator.tsx:118` | `provider === 'anthropic' && authType === 'oauth'` | Only Anthropic gets usage bars | Add `\|\| provider === 'openai'` |
-
-### 3.2 MODERATE — Should Change
-
-| File | Line(s) | What's hardcoded | What to do |
-|---|---|---|---|
-| `usage-monitor.ts:1040-1072` | `determineActiveProfile()` | Returns `baseUrl: 'https://api.anthropic.com'` for all OAuth | Detect provider, return `chatgpt.com/backend-api` for Codex |
-| `credential-utils.ts` | Keychain service names | `"Claude Code-credentials"` | Codex tokens stored differently (file-based, not keychain) |
-| `usage-monitor.ts:1513` | `if (provider === 'zai' \|\| provider === 'zhipu')` | Provider-specific response unwrapping | Add Codex response parsing (different JSON structure) |
-| `rate-limit-detector.ts:14` | `RATE_LIMIT_PATTERN` | Claude-specific: `"Limit reached · resets..."` | Add Codex-specific patterns |
-| IPC channel names | `'claude:usageUpdated'`, `'claude:allProfilesUsageUpdated'` | "claude" prefix | Cosmetic — rename to `'usage:updated'` etc. (optional, low priority) |
-
-### 3.3 LOW PRIORITY — Nice to Have
-
-| Item | What | Why low priority |
-|---|---|---|
-| Type naming | `ClaudeUsageSnapshot` → `UsageSnapshot` | Structural refactor, types work as-is for Codex |
-| IPC method names | `requestUsageUpdate` returns `ClaudeUsageSnapshot` | Works fine, just naming |
-| `claudeProfileId` on `ProviderAccount` | Only used for Anthropic OAuth | Codex doesn't need it |
-
---
-
-## 4. Provider-Agnostic Parts
-
-These components already work for any provider and need NO changes:
-
-| Component/Module | Why it's already generic |
-|---|---|
-| `profile-scorer.ts` | Scores by `billingModel`, usage thresholds, rate limit events — no provider checks |
-| `rate-limit-manager.ts` | Stores/checks rate limit events — pure data, no provider logic |
-| `operation-registry.ts` | Tracks running operations — no provider awareness |
-| `ProviderAccount` type | Has `provider` field, `billingModel`, `usage` — works for any provider |
-| `globalPriorityOrder` | Array of account IDs — provider-agnostic ordering |
-| `useActiveProvider()` hook | Returns first account in priority order — generic |
-| `ProviderAccountCard` | Shows usage bars when `account.usage` is populated — will work for Codex once data flows |
-| `AddAccountDialog` | Already has separate Codex OAuth flow |
-| `AuthStatusIndicator` | Already shows Codex-specific green badge |
-| All i18n keys | Codex-specific labels already exist |
-
---
-
-## 5. Implementation Plan
-
-### Phase 1: Codex Usage Fetcher (Core)
-
-Create `apps/desktop/src/main/claude-profile/codex-usage-fetcher.ts`:
-
-```typescript
-// Responsibilities:
-// 1. Read Codex OAuth token (from our codex-auth.json)
-// 2. Read account_id (from ~/.codex/auth.json or JWT decode)
-// 3. Call GET https://chatgpt.com/backend-api/wham/usage
-// 4. Parse response into ClaudeUsageSnapshot format
-// 5. Handle 401 → refresh token via codex-oauth.ts
-// 6. Handle 403 → mark as needsReauthentication
-```
-
-**Key function:**
-```typescript
-async function fetchCodexUsage(accessToken: string, accountId?: string): Promise<ClaudeUsageSnapshot>
-```
-
-### Phase 2: Wire into Usage Monitor
-
-Modify `usage-monitor.ts`:
-
-1. Add `chatgpt.com` to `ALLOWED_USAGE_API_DOMAINS`
-2. Add Codex to `PROVIDER_USAGE_ENDPOINTS`
-3. Update `determineActiveProfile()` to detect Codex accounts from `globalPriorityOrder`
-4. Update `getCredential()` to read Codex OAuth token (from `codex-auth.json`)
-5. Update `fetchUsageViaAPI()` to handle Codex response format
-6. Add Codex-specific headers (`ChatGPT-Account-Id`)
-7. Add Codex response parsing (different JSON structure than Anthropic)
-
-### Phase 3: Token Refresh Routing
-
-Modify `token-refresh.ts` or create parallel Codex path:
-
- When refreshing a Codex token, use `auth.openai.com/oauth/token` with Codex client ID
- When refreshing an Anthropic token, use `console.anthropic.com/v1/oauth/token` with Claude client ID
- Provider detection: check the account's `provider` field, or detect from token prefix
-
-### Phase 4: UI Updates
-
-1. `UsageIndicator.tsx:118` — Add `|| provider === 'openai'` to `hasUsageMonitoring`
-2. That's it — the rest of the UI already handles usage bars, reset times, multi-profile display generically
-
-### Phase 5: Auto-Swap for Codex
-
-1. Add Codex-specific rate limit patterns to `rate-limit-detector.ts`
-2. Codex returns `"codexErrorInfo": "UsageLimitExceeded"` on limit hit
-3. Auto-swap logic in `profile-scorer.ts` already works — it just needs usage data populated
-
---
-
-## Appendix: Comparison Table
-
-| Aspect | Anthropic (Claude Code) | OpenAI (Codex) |
-|---|---|---|
-| **Usage endpoint** | `api.anthropic.com/api/oauth/usage` | `chatgpt.com/backend-api/wham/usage` |
-| **Auth header** | `Bearer <oauth_token>` | `Bearer <access_token>` + `ChatGPT-Account-Id` |
-| **Session window** | ~5h | Configurable (`limit_window_seconds`) |
-| **Weekly window** | 7 days | Configurable (`limit_window_seconds`) |
-| **Token source** | Keychain (`Claude Code-credentials`) | File (`codex-auth.json`) |
-| **Token refresh** | `console.anthropic.com/v1/oauth/token` | `auth.openai.com/oauth/token` |
-| **Client ID** | `9d1c250a-e61b-44d9-88ed-5944d1962f5e` | `app_EMoamEEZ73f0CkXaXp7hrann` |
-| **Passive tracking** | Not available | `x-codex-*` response headers |
-| **Rate limit error** | `"Limit reached · resets Dec 17..."` | `"codexErrorInfo": "UsageLimitExceeded"` |
-| **Profile isolation** | `~/.claude-profiles/{name}/` dirs | Single `codex-auth.json` file |
-| **Multi-account** | Multiple config dirs in keychain | Single file (no multi-account yet) |
-
-## Appendix: Caveats
-
-1. **Undocumented API** — `chatgpt.com/backend-api/wham/usage` is internal. The Codex CLI depends on it, so it's unlikely to break silently.
-2. **Account ID** — May be required. Test without it first. If needed, decode from JWT or read `~/.codex/auth.json`.
-3. **CORS** — Not an issue (Electron main process = Node.js).
-4. **Polling rate** — Unknown if OpenAI rate-limits `wham/usage`. Start conservatively (every 30-60s).
-5. **Multi-account Codex** — Codex CLI doesn't support multiple accounts. We store one token file. If user has multiple Codex accounts, they'd need to re-auth each time (unlike Anthropic which supports multiple config dirs).
@@ -73,11 +73,35 @@ Read the full CLA here: [CLA.md](CLA.md)

 Before contributing, ensure you have the following installed:

- **Node.js 24+** - For the Electron desktop app
- **npm 10+** - Package manager (comes with Node.js)
- **CMake** - Required for building native dependencies (e.g., node-pty)
+- **Python 3.12+** - For the backend framework
+- **Node.js 24+** - For the Electron frontend
+- **npm 10+** - Package manager for the frontend (comes with Node.js)
+- **uv** (recommended) or **pip** - Python package manager
+- **CMake** - Required for building native dependencies (e.g., LadybugDB)
 - **Git** - Version control

+### Installing Python 3.12
+
+**Windows:**
+```bash
+winget install Python.Python.3.12
+```
+
+**macOS:**
+```bash
+brew install python@3.12
+```
+
+**Linux (Ubuntu/Debian):**
+```bash
+sudo apt install python3.12 python3.12-venv
+```
+
+**Linux (Fedora):**
+```bash
+sudo dnf install python3.12
+```
+
 ### Installing Node.js 24+

 **Windows:**
@@ -144,27 +168,43 @@ npm start

 ## Development Setup

-The project is a single Electron desktop application in `apps/desktop/`. All AI agent logic lives in TypeScript using the Vercel AI SDK v6.
+The project consists of two main components:

-From the repository root:
+1. **Python Backend** (`apps/backend/`) - The core autonomous coding framework
+2. **Electron Frontend** (`apps/frontend/`) - Desktop UI
+
+From the repository root, two commands handle everything:

 ```bash
-# Install all dependencies
+# Install all dependencies (Python backend + Electron frontend)
 npm run install:all

 # Start development mode (hot reload)
 npm run dev
 ```

-`npm run install:all` installs the npm dependencies for `apps/desktop/`.
+`npm run install:all` automatically:
+- Detects Python 3.12+ on your system
+- Creates a virtual environment (`apps/backend/.venv`)
+- Installs backend runtime and test dependencies
+- Copies `.env.example` to `.env` (if not already present)
+- Installs frontend npm dependencies
+
+After install, configure your credentials in `apps/backend/.env`:
+```bash
+# Get your Claude Code OAuth token
+claude setup-token
+
+# Then edit apps/backend/.env with your token and any other provider keys
+```

 ### Other Useful Commands

 ```bash
 npm start              # Build and run production
-npm run build          # Build for production
+npm run build          # Build frontend for production
 npm run package        # Package for distribution
-npm test               # Run frontend tests
+npm run test:backend   # Run Python tests
 ```

 <details>
@@ -183,20 +223,28 @@ Auto Claude automatically downloads prebuilt binaries for Windows. If prebuilts

 ## Pre-commit Hooks

-We use Husky + lint-staged to run Biome linting and formatting checks before each commit.
+We use [pre-commit](https://pre-commit.com/) to run linting and formatting checks before each commit. This ensures code quality and consistency across the project.

 ### Setup

-Husky is installed automatically when you run `npm install` inside `apps/desktop/`.
+```bash
+# Install pre-commit
+pip install pre-commit
+
+# Install the git hooks (run once after cloning)
+pre-commit install
+```

 ### What Runs on Commit

-When you commit, the following checks run automatically on staged files:
+When you commit, the following checks run automatically:

 | Check | Scope | Description |
 |-------|-------|-------------|
-| **Biome** | `apps/desktop/` | TypeScript/React linter + formatter |
-| **typecheck** | `apps/desktop/` | TypeScript type checking |
+| **ruff** | `apps/backend/` | Python linter with auto-fix |
+| **ruff-format** | `apps/backend/` | Python code formatter |
+| **eslint** | `apps/frontend/` | TypeScript/React linter |
+| **typecheck** | `apps/frontend/` | TypeScript type checking |
 | **trailing-whitespace** | All files | Removes trailing whitespace |
 | **end-of-file-fixer** | All files | Ensures files end with newline |
 | **check-yaml** | All files | Validates YAML syntax |
@@ -205,29 +253,55 @@ When you commit, the following checks run automatically on staged files:
 ### Running Manually

 ```bash
-cd apps/desktop
+# Run all checks on all files
+pre-commit run --all-files

-# Run linter (Biome)
-npm run lint
+# Run a specific hook
+pre-commit run ruff --all-files

-# Auto-fix lint issues
-npm run lint:fix
-
-# Run type checking
-npm run typecheck
+# Skip hooks temporarily (not recommended)
+git commit --no-verify -m "message"
 ```

 ### If a Check Fails

-1. **Biome auto-fixes**: Run `npm run lint:fix` in `apps/desktop/`. Stage the changes and commit again.
-2. **Type errors**: Resolve TypeScript type issues before committing.
+1. **Ruff auto-fixes**: Some issues are fixed automatically. Stage the changes and commit again.
+2. **ESLint errors**: Fix the reported issues in your code.
+3. **Type errors**: Resolve TypeScript type issues before committing.

 ## Code Style

+### Python
+
+- Follow PEP 8 style guidelines
+- Use type hints for function signatures
+- Use docstrings for public functions and classes
+- Keep functions focused and under 50 lines when possible
+- Use meaningful variable and function names
+
+```python
+# Good
+def get_next_chunk(spec_dir: Path) -> dict | None:
+    """
+    Find the next pending chunk in the implementation plan.
+
+    Args:
+        spec_dir: Path to the spec directory
+
+    Returns:
+        The next chunk dict or None if all chunks are complete
+    """
+    ...
+
+# Avoid
+def gnc(sd):
+    ...
+```
+
 ### TypeScript/React

 - Use TypeScript strict mode
- Follow the existing component patterns in `apps/desktop/src/`
+- Follow the existing component patterns in `apps/frontend/src/`
 - Use functional components with hooks
 - Prefer named exports over default exports
 - Use the UI components from `src/renderer/components/ui/`
@@ -252,12 +326,96 @@ export default function(props) {
 - End files with a newline
 - Keep line length under 100 characters when practical

+### File Encoding (Python)
+
+**Always specify `encoding="utf-8"` for text file operations** to ensure Windows compatibility.
+
+Windows Python defaults to `cp1252` encoding instead of UTF-8, causing errors with:
+- Emoji (🚀, ✅, ❌)
+- International characters (ñ, é, 中文, العربية)
+- Special symbols (™, ©, ®)
+
+**DO:**
+
+```python
+# Reading files
+with open(path, encoding="utf-8") as f:
+    content = f.read()
+
+# Writing files
+with open(path, "w", encoding="utf-8") as f:
+    f.write(content)
+
+# Path methods
+from pathlib import Path
+content = Path(file).read_text(encoding="utf-8")
+Path(file).write_text(content, encoding="utf-8")
+
+# JSON files - reading
+import json
+with open(path, encoding="utf-8") as f:
+    data = json.load(f)
+
+# JSON files - writing
+with open(path, "w", encoding="utf-8") as f:
+    json.dump(data, f, ensure_ascii=False, indent=2)
+```
+
+**DON'T:**
+
+```python
+# Wrong - platform-dependent encoding
+with open(path) as f:
+    content = f.read()
+
+# Wrong - Path methods without encoding
+content = Path(file).read_text()
+
+# Wrong - encoding on json.dump (not open!)
+json.dump(data, f, encoding="utf-8")  # ERROR
+```
+
+**Binary files - NO encoding:**
+
+```python
+with open(path, "rb") as f:  # Correct
+    data = f.read()
+```
+
+Our pre-commit hooks automatically check for missing encoding parameters. See [PR #782](https://github.com/AndyMik90/Auto-Claude/pull/782) for the comprehensive encoding fix and [guides/windows-development.md](guides/windows-development.md) for Windows-specific development guidance.
+
 ## Testing

+### Python Tests
+
+```bash
+# Run all tests (from repository root)
+npm run test:backend
+
+# Or manually with pytest
+cd apps/backend
+.venv/Scripts/pytest.exe ../tests -v          # Windows
+.venv/bin/pytest ../tests -v                   # macOS/Linux
+
+# Run a specific test file
+npm run test:backend -- tests/test_security.py -v
+
+# Run a specific test
+npm run test:backend -- tests/test_security.py::test_bash_command_validation -v
+
+# Skip slow tests
+npm run test:backend -- -m "not slow"
+
+# Run with coverage
+pytest tests/ --cov=apps/backend --cov-report=html
+```
+
+Test configuration is in `tests/pytest.ini`.
+
 ### Frontend Tests

 ```bash
-cd apps/desktop
+cd apps/frontend

 # Run unit tests
 npm test
@@ -296,22 +454,29 @@ All pull requests and pushes to `main` trigger automated CI checks via GitHub Ac

 | Workflow | Trigger | What it checks |
 |----------|---------|----------------|
-| **CI** | Push to `main`, PRs | Frontend tests (all 3 platforms), TypeScript type check, build |
-| **Lint** | Push to `main`, PRs | Biome (TypeScript/React) |
+| **CI** | Push to `main`, PRs | Python tests (3.11 & 3.12), Frontend tests |
+| **Lint** | Push to `main`, PRs | Ruff (Python), ESLint + TypeScript (Frontend) |

 ### PR Requirements

 Before a PR can be merged:

 1. All CI checks must pass (green checkmarks)
-2. Frontend tests pass on all three platforms (Ubuntu, Windows, macOS)
-3. Linting passes (no Biome errors)
-4. TypeScript type checking passes
+2. Python tests pass on both Python 3.11 and 3.12
+3. Frontend tests pass
+4. Linting passes (no ruff or eslint errors)
+5. TypeScript type checking passes

 ### Running CI Checks Locally

 ```bash
-cd apps/desktop
+# Python tests
+cd apps/backend
+source .venv/bin/activate
+pytest ../../tests/ -v
+
+# Frontend tests
+cd apps/frontend
 npm test
 npm run lint
 npm run typecheck
@@ -622,7 +787,8 @@ git rebase -i origin/develop
 git push --force-with-lease

 # Verify everything works
-cd apps/desktop && npm test && npm run lint && npm run typecheck
+npm run test:backend
+cd apps/frontend && npm test && npm run lint && npm run typecheck
 ```

 **PR size:**
@@ -643,7 +809,11 @@ cd apps/desktop && npm test && npm run lint && npm run typecheck

 3. **Test thoroughly**:
   ```bash
-   cd apps/desktop && npm test && npm run lint && npm run typecheck
+   # Python (from repository root)
+   npm run test:backend
+
+   # Frontend
+   cd apps/frontend && npm test && npm run lint && npm run typecheck
   ```

 4. **Update documentation** if your changes affect:
@@ -681,7 +851,8 @@ When reporting a bug, include:
 1. **Clear title** describing the issue
 2. **Environment details**:
   - OS and version
-   - Node.js version
+   - Python version
+   - Node.js version (for UI issues)
   - Auto Claude version
 3. **Steps to reproduce** the issue
 4. **Expected behavior** vs **actual behavior**
@@ -699,14 +870,25 @@ When requesting a feature:

 ## Architecture Overview

-Auto Claude is a single Electron desktop application in `apps/desktop/`.
+Auto Claude consists of two main parts:

-### Electron Desktop (`apps/desktop/`)
+### Python Backend (`apps/backend/`)

- **AI Agent Layer** (`src/main/ai/`) - Vercel AI SDK v6 agent runtime, providers, tools, security, orchestration
- **Main Process** (`src/main/`) - IPC handlers, agent queue, terminal management, claude-profile
- **Renderer** (`src/renderer/`) - React UI components and Zustand stores
- **Shared** (`src/shared/`) - Types, i18n locales, constants, utilities
+The core autonomous coding framework:
+
+- **Entry Points**: `run.py` (build runner), `spec_runner.py` (spec creator)
+- **Agent System**: `agent.py`, `client.py`, `prompts/`
+- **Execution**: `coordinator.py` (parallel), `worktree.py` (isolation)
+- **Memory**: `memory.py` (file-based), `graphiti_memory.py` (graph-based)
+- **QA**: `qa_loop.py`, `prompts/qa_*.md`
+
+### Electron Frontend (`apps/frontend/`)
+
+Desktop interface:
+
+- **Main Process**: `src/main/` - Electron main process, IPC handlers
+- **Renderer**: `src/renderer/` - React UI components
+- **Shared**: `src/shared/` - Types and utilities

 For detailed architecture information, see [CLAUDE.md](CLAUDE.md).

@@ -1,8 +1,8 @@
-# Aperant (formerly Auto Claude)
+# Auto Claude

 **Autonomous multi-agent coding framework that plans, builds, and validates software for you.**

-![Aperant Kanban Board](.github/assets/Auto-Claude-Kanban.png)
+![Auto Claude Kanban Board](.github/assets/Auto-Claude-Kanban.png)

 [![License](https://img.shields.io/badge/license-AGPL--3.0-green?style=flat-square)](./agpl-3.0.txt)
 [![Discord](https://img.shields.io/badge/Discord-Join%20Community-5865F2?style=flat-square&logo=discord&logoColor=white)](https://discord.gg/KCXaPBr4Dj)
@@ -17,18 +17,18 @@
 ### Stable Release

 <!-- STABLE_VERSION_BADGE -->
-[![Stable](https://img.shields.io/badge/stable-2.7.6-blue?style=flat-square)](https://github.com/AndyMik90/Auto-Claude/releases/tag/v2.7.6)
+[![Stable](https://img.shields.io/badge/stable-2.7.5-blue?style=flat-square)](https://github.com/AndyMik90/Auto-Claude/releases/tag/v2.7.5)
 <!-- STABLE_VERSION_BADGE_END -->

 <!-- STABLE_DOWNLOADS -->
 | Platform | Download |
 |----------|----------|
-| **Windows** | [Auto-Claude-2.7.6-win32-x64.exe](https://github.com/AndyMik90/Auto-Claude/releases/download/v2.7.6/Auto-Claude-2.7.6-win32-x64.exe) |
-| **macOS (Apple Silicon)** | [Auto-Claude-2.7.6-darwin-arm64.dmg](https://github.com/AndyMik90/Auto-Claude/releases/download/v2.7.6/Auto-Claude-2.7.6-darwin-arm64.dmg) |
-| **macOS (Intel)** | [Auto-Claude-2.7.6-darwin-x64.dmg](https://github.com/AndyMik90/Auto-Claude/releases/download/v2.7.6/Auto-Claude-2.7.6-darwin-x64.dmg) |
-| **Linux** | [Auto-Claude-2.7.6-linux-x86_64.AppImage](https://github.com/AndyMik90/Auto-Claude/releases/download/v2.7.6/Auto-Claude-2.7.6-linux-x86_64.AppImage) |
-| **Linux (Debian)** | [Auto-Claude-2.7.6-linux-amd64.deb](https://github.com/AndyMik90/Auto-Claude/releases/download/v2.7.6/Auto-Claude-2.7.6-linux-amd64.deb) |
-| **Linux (Flatpak)** | [Auto-Claude-2.7.6-linux-x86_64.flatpak](https://github.com/AndyMik90/Auto-Claude/releases/download/v2.7.6/Auto-Claude-2.7.6-linux-x86_64.flatpak) |
+| **Windows** | [Auto-Claude-2.7.5-win32-x64.exe](https://github.com/AndyMik90/Auto-Claude/releases/download/v2.7.5/Auto-Claude-2.7.5-win32-x64.exe) |
+| **macOS (Apple Silicon)** | [Auto-Claude-2.7.5-darwin-arm64.dmg](https://github.com/AndyMik90/Auto-Claude/releases/download/v2.7.5/Auto-Claude-2.7.5-darwin-arm64.dmg) |
+| **macOS (Intel)** | [Auto-Claude-2.7.5-darwin-x64.dmg](https://github.com/AndyMik90/Auto-Claude/releases/download/v2.7.5/Auto-Claude-2.7.5-darwin-x64.dmg) |
+| **Linux** | [Auto-Claude-2.7.5-linux-x86_64.AppImage](https://github.com/AndyMik90/Auto-Claude/releases/download/v2.7.5/Auto-Claude-2.7.5-linux-x86_64.AppImage) |
+| **Linux (Debian)** | [Auto-Claude-2.7.5-linux-amd64.deb](https://github.com/AndyMik90/Auto-Claude/releases/download/v2.7.5/Auto-Claude-2.7.5-linux-amd64.deb) |
+| **Linux (Flatpak)** | [Auto-Claude-2.7.5-linux-x86_64.flatpak](https://github.com/AndyMik90/Auto-Claude/releases/download/v2.7.5/Auto-Claude-2.7.5-linux-x86_64.flatpak) |
 <!-- STABLE_DOWNLOADS_END -->

 ### Beta Release
@@ -36,18 +36,18 @@
 > ⚠️ Beta releases may contain bugs and breaking changes. [View all releases](https://github.com/AndyMik90/Auto-Claude/releases)

 <!-- BETA_VERSION_BADGE -->
-[![Beta](https://img.shields.io/badge/beta-2.8.0--beta.1-orange?style=flat-square)](https://github.com/AndyMik90/Auto-Claude/releases/tag/v2.8.0-beta.1)
+[![Beta](https://img.shields.io/badge/beta-2.7.6--beta.5-orange?style=flat-square)](https://github.com/AndyMik90/Auto-Claude/releases/tag/v2.7.6-beta.5)
 <!-- BETA_VERSION_BADGE_END -->

 <!-- BETA_DOWNLOADS -->
 | Platform | Download |
 |----------|----------|
-| **Windows** | [Auto-Claude-2.8.0-beta.1-win32-x64.exe](https://github.com/AndyMik90/Auto-Claude/releases/download/v2.8.0-beta.1/Auto-Claude-2.8.0-beta.1-win32-x64.exe) |
-| **macOS (Apple Silicon)** | [Auto-Claude-2.8.0-beta.1-darwin-arm64.dmg](https://github.com/AndyMik90/Auto-Claude/releases/download/v2.8.0-beta.1/Auto-Claude-2.8.0-beta.1-darwin-arm64.dmg) |
-| **macOS (Intel)** | [Auto-Claude-2.8.0-beta.1-darwin-x64.dmg](https://github.com/AndyMik90/Auto-Claude/releases/download/v2.8.0-beta.1/Auto-Claude-2.8.0-beta.1-darwin-x64.dmg) |
-| **Linux** | [Auto-Claude-2.8.0-beta.1-linux-x86_64.AppImage](https://github.com/AndyMik90/Auto-Claude/releases/download/v2.8.0-beta.1/Auto-Claude-2.8.0-beta.1-linux-x86_64.AppImage) |
-| **Linux (Debian)** | [Auto-Claude-2.8.0-beta.1-linux-amd64.deb](https://github.com/AndyMik90/Auto-Claude/releases/download/v2.8.0-beta.1/Auto-Claude-2.8.0-beta.1-linux-amd64.deb) |
-| **Linux (Flatpak)** | [Auto-Claude-2.8.0-beta.1-linux-x86_64.flatpak](https://github.com/AndyMik90/Auto-Claude/releases/download/v2.8.0-beta.1/Auto-Claude-2.8.0-beta.1-linux-x86_64.flatpak) |
+| **Windows** | [Auto-Claude-2.7.6-beta.5-win32-x64.exe](https://github.com/AndyMik90/Auto-Claude/releases/download/v2.7.6-beta.5/Auto-Claude-2.7.6-beta.5-win32-x64.exe) |
+| **macOS (Apple Silicon)** | [Auto-Claude-2.7.6-beta.5-darwin-arm64.dmg](https://github.com/AndyMik90/Auto-Claude/releases/download/v2.7.6-beta.5/Auto-Claude-2.7.6-beta.5-darwin-arm64.dmg) |
+| **macOS (Intel)** | [Auto-Claude-2.7.6-beta.5-darwin-x64.dmg](https://github.com/AndyMik90/Auto-Claude/releases/download/v2.7.6-beta.5/Auto-Claude-2.7.6-beta.5-darwin-x64.dmg) |
+| **Linux** | [Auto-Claude-2.7.6-beta.5-linux-x86_64.AppImage](https://github.com/AndyMik90/Auto-Claude/releases/download/v2.7.6-beta.5/Auto-Claude-2.7.6-beta.5-linux-x86_64.AppImage) |
+| **Linux (Debian)** | [Auto-Claude-2.7.6-beta.5-linux-amd64.deb](https://github.com/AndyMik90/Auto-Claude/releases/download/v2.7.6-beta.5/Auto-Claude-2.7.6-beta.5-linux-amd64.deb) |
+| **Linux (Flatpak)** | [Auto-Claude-2.7.6-beta.5-linux-x86_64.flatpak](https://github.com/AndyMik90/Auto-Claude/releases/download/v2.7.6-beta.5/Auto-Claude-2.7.6-beta.5-linux-x86_64.flatpak) |
 <!-- BETA_DOWNLOADS_END -->

 > All releases include SHA256 checksums and VirusTotal scan results for security verification.
@@ -114,15 +114,39 @@ AI-assisted feature planning with competitor analysis and audience targeting.
 ## Project Structure

 ```
-Aperant/
+Auto-Claude/
 ├── apps/
-│   └── desktop/     # Electron desktop application (TypeScript AI agent layer + UI)
+│   ├── backend/     # Python agents, specs, QA pipeline
+│   └── frontend/    # Electron desktop application
 ├── guides/          # Additional documentation
+├── tests/           # Test suite
 └── scripts/         # Build utilities
 ```

 ---

+## CLI Usage
+
+For headless operation, CI/CD integration, or terminal-only workflows:
+
+```bash
+cd apps/backend
+
+# Create a spec interactively
+python spec_runner.py --interactive
+
+# Run autonomous build
+python run.py --spec 001
+
+# Review and merge
+python run.py --spec 001 --review
+python run.py --spec 001 --merge
+```
+
+See [guides/CLI-USAGE.md](guides/CLI-USAGE.md) for complete CLI documentation.
+
+---
+
 ## Development

 Want to build from source or contribute? See [CONTRIBUTING.md](CONTRIBUTING.md) for complete development setup instructions.
@@ -133,7 +157,7 @@ For Linux-specific builds (Flatpak, AppImage), see [guides/linux.md](guides/linu

 ## Security

-Aperant uses a three-layer security model:
+Auto Claude uses a three-layer security model:

 1. **OS Sandbox** - Bash commands run in isolation
 2. **Filesystem Restrictions** - Operations limited to project directory
@@ -150,7 +174,7 @@ All releases are:

 | Command | Description |
 |---------|-------------|
-| `npm run install:all` | Install all dependencies |
+| `npm run install:all` | Install backend and frontend dependencies |
 | `npm start` | Build and run the desktop app |
 | `npm run dev` | Run in development mode with hot reload |
 | `npm run package` | Package for current platform |
@@ -160,6 +184,7 @@ All releases are:
 | `npm run package:flatpak` | Package as Flatpak (see [guides/linux.md](guides/linux.md)) |
 | `npm run lint` | Run linter |
 | `npm test` | Run frontend tests |
+| `npm run test:backend` | Run backend tests |

 ---

@@ -185,7 +210,7 @@ We welcome contributions! Please read [CONTRIBUTING.md](CONTRIBUTING.md) for:

 **AGPL-3.0** - GNU Affero General Public License v3.0

-Aperant is free to use. If you modify and distribute it, or run it as a service, your code must also be open source under AGPL-3.0.
+Auto Claude is free to use. If you modify and distribute it, or run it as a service, your code must also be open source under AGPL-3.0.

 Commercial licensing available for closed-source use cases.

@@ -66,8 +66,9 @@ node scripts/bump-version.js 2.8.0   # Set specific version
 ```

 This will:
- Update `apps/desktop/package.json`
+- Update `apps/frontend/package.json`
 - Update `package.json` (root)
+- Update `apps/backend/__init__.py`
 - Check if `CHANGELOG.md` has an entry for the new version (warns if missing)
 - Create a commit with message `chore: bump version to X.Y.Z`

@@ -194,7 +195,7 @@ The release workflow **validates** that `CHANGELOG.md` has an entry for the vers
 1. Check if version in `package.json` is greater than latest tag:
   ```bash
   git tag -l 'v*' --sort=-version:refname | head -1
-   cat apps/desktop/package.json | grep version
+   cat apps/frontend/package.json | grep version
   ```

 2. Ensure the merge commit touched `package.json`:
@@ -0,0 +1,372 @@
+# Auto Claude Environment Variables
+# Copy this file to .env and fill in your values
+
+# =============================================================================
+# AUTHENTICATION (REQUIRED)
+# =============================================================================
+# Auto Claude uses Claude Code OAuth authentication.
+# Direct API keys (ANTHROPIC_API_KEY) are NOT supported to prevent silent billing.
+#
+# Option 1: Run `claude setup-token` to save token to system keychain (recommended)
+#           (macOS: Keychain, Windows: Credential Manager, Linux: secret-service)
+# Option 2: Set the token explicitly:
+# CLAUDE_CODE_OAUTH_TOKEN=your-oauth-token-here
+#
+# For enterprise/proxy setups (CCR):
+# ANTHROPIC_AUTH_TOKEN=sk-zcf-x-ccr
+
+# =============================================================================
+# CUSTOM API ENDPOINT (OPTIONAL)
+# =============================================================================
+# Override the default Anthropic API endpoint. Useful for:
+#   - Local proxies (ccr, litellm)
+#   - API gateways
+#   - Self-hosted Claude instances
+#
+# ANTHROPIC_BASE_URL=http://127.0.0.1:3456
+#
+# Related settings (usually set together with ANTHROPIC_BASE_URL):
+# NO_PROXY=127.0.0.1
+# DISABLE_TELEMETRY=true
+# DISABLE_COST_WARNINGS=true
+# API_TIMEOUT_MS=600000
+
+# Model override (OPTIONAL)
+# Default: claude-opus-4-6
+# AUTO_BUILD_MODEL=claude-opus-4-6
+
+
+# =============================================================================
+# GIT/WORKTREE SETTINGS (OPTIONAL)
+# =============================================================================
+# Configure how Auto Claude handles git worktrees for isolated builds.
+
+# Default base branch for worktree creation (OPTIONAL)
+# If not set, Auto Claude will auto-detect main/master, or fall back to current branch.
+# Common values: main, master, develop
+# DEFAULT_BRANCH=main
+
+# =============================================================================
+# DEBUG MODE (OPTIONAL)
+# =============================================================================
+# Enable debug logging for development and troubleshooting.
+# Shows detailed information about runner execution, agent calls, file operations.
+
+# Enable debug mode (default: false)
+# DEBUG=true
+
+# Debug log level: 1=basic, 2=detailed, 3=verbose (default: 1)
+# DEBUG_LEVEL=1
+
+# Log to file instead of stdout (OPTIONAL)
+# DEBUG_LOG_FILE=auto-claude/debug.log
+
+# =============================================================================
+# LINEAR INTEGRATION (OPTIONAL)
+# =============================================================================
+# Enable Linear integration for real-time progress tracking in Linear.
+# Get your API key from: https://linear.app/YOUR-TEAM/settings/api
+
+# Linear API Key (OPTIONAL - enables Linear integration)
+# LINEAR_API_KEY=lin_api_xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
+
+# Pre-configured Team ID (OPTIONAL - will auto-detect if not set)
+# LINEAR_TEAM_ID=
+
+# Pre-configured Project ID (OPTIONAL - will create project if not set)
+# LINEAR_PROJECT_ID=
+
+# =============================================================================
+# GITLAB INTEGRATION (OPTIONAL)
+# =============================================================================
+# Enable GitLab integration for issue tracking and merge requests.
+# Supports both GitLab.com and self-hosted GitLab instances.
+#
+# Authentication Options (choose one):
+#
+#   Option 1: glab CLI OAuth (Recommended)
+#   Install glab CLI: https://gitlab.com/gitlab-org/cli#installation
+#   Then run: glab auth login
+#   This opens your browser for OAuth authentication. Once complete,
+#   Auto Claude will automatically use your glab credentials (no env vars needed).
+#   For self-hosted: glab auth login --hostname gitlab.example.com
+#
+#   Option 2: Personal Access Token
+#   Set GITLAB_TOKEN below. Token auth is used if set, otherwise falls back to glab CLI.
+
+# GitLab Instance URL (OPTIONAL - defaults to gitlab.com)
+# For self-hosted: GITLAB_INSTANCE_URL=https://gitlab.example.com
+# GITLAB_INSTANCE_URL=https://gitlab.com
+
+# GitLab Personal Access Token (OPTIONAL - only needed if not using glab CLI)
+# Required scope: api (covers issues, merge requests, releases, project info)
+# Optional scope: write_repository (only if creating new GitLab projects from local repos)
+# Get from: https://gitlab.com/-/user_settings/personal_access_tokens
+# GITLAB_TOKEN=glpat-xxxxxxxxxxxxxxxxxxxx
+
+# GitLab Project (OPTIONAL - format: group/project or numeric ID)
+# If not set, will auto-detect from git remote
+# GITLAB_PROJECT=mygroup/myproject
+
+# =============================================================================
+# UI SETTINGS (OPTIONAL)
+# =============================================================================
+# Enable fancy terminal UI with icons, colors, and interactive menus.
+# Set to "false" to use plain text output (useful for CI/CD or log files).
+
+# Enable fancy UI (default: true)
+# ENABLE_FANCY_UI=true
+
+# =============================================================================
+# ELECTRON MCP SERVER (OPTIONAL)
+# =============================================================================
+# Enable Electron MCP server for AI agents to interact with and validate
+# Electron desktop applications. This allows QA agents to capture screenshots,
+# inspect windows, and validate Electron apps during the review process.
+#
+# The electron-mcp-server connects via Chrome DevTools Protocol to an Electron
+# app running with remote debugging enabled.
+#
+# Prerequisites:
+#   1. Start your Electron app with remote debugging:
+#      ./YourElectronApp --remote-debugging-port=9222
+#
+#   2. For auto-claude-ui specifically (use the MCP-enabled scripts):
+#      cd auto-claude-ui
+#      pnpm run dev:mcp     # Development mode with MCP debugging
+#      # OR for production build:
+#      pnpm run start:mcp   # Production mode with MCP debugging
+#
+# Note: Only QA agents (qa_reviewer, qa_fixer) receive Electron MCP tools.
+# Coder and Planner agents do NOT have access to these tools to minimize
+# context token usage and keep agents focused on their roles.
+#
+# See: https://github.com/anthropics/anthropic-quickstarts/tree/main/mcp-electron-demo
+
+# Enable Electron MCP integration (default: false)
+# ELECTRON_MCP_ENABLED=true
+
+# Chrome DevTools debugging port for Electron connection (default: 9222)
+# ELECTRON_DEBUG_PORT=9222
+
+# =============================================================================
+# GRAPHITI MEMORY INTEGRATION (REQUIRED)
+# =============================================================================
+# Graphiti-based persistent memory layer for cross-session context
+# retention. Uses LadybugDB as the embedded graph database.
+#
+# REQUIREMENTS:
+#   - Python 3.12 or higher
+#   - Install: pip install real_ladybug graphiti-core
+#
+# Supports multiple LLM and embedder providers:
+#   - OpenAI (default)
+#   - Anthropic (LLM only, use with Voyage for embeddings)
+#   - Azure OpenAI
+#   - Ollama (local, fully offline)
+#   - Google AI (Gemini)
+
+# Graphiti is enabled by default. Set to false to disable memory features.
+GRAPHITI_ENABLED=true
+
+# =============================================================================
+# GRAPHITI: Database Settings
+# =============================================================================
+# LadybugDB stores data in a local directory (no Docker required).
+
+# Database name (default: auto_claude_memory)
+# GRAPHITI_DATABASE=auto_claude_memory
+
+# Database storage path (default: ~/.auto-claude/memories)
+# GRAPHITI_DB_PATH=~/.auto-claude/memories
+
+# =============================================================================
+# GRAPHITI: Provider Selection
+# =============================================================================
+# Choose which providers to use for LLM and embeddings.
+# Default is "openai" for both.
+
+# LLM provider: openai | anthropic | azure_openai | ollama | google | openrouter
+# GRAPHITI_LLM_PROVIDER=openai
+
+# Embedder provider: openai | voyage | azure_openai | ollama | google | openrouter
+# GRAPHITI_EMBEDDER_PROVIDER=openai
+
+# =============================================================================
+# GRAPHITI: OpenAI Provider (Default)
+# =============================================================================
+# Use OpenAI for both LLM and embeddings. This is the simplest setup.
+# Required: OPENAI_API_KEY
+
+# OpenAI API Key
+# OPENAI_API_KEY=sk-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
+
+# OpenAI Model for LLM (default: gpt-4o-mini)
+# OPENAI_MODEL=gpt-4o-mini
+
+# OpenAI Model for embeddings (default: text-embedding-3-small)
+# Available: text-embedding-3-small (1536 dim), text-embedding-3-large (3072 dim)
+# OPENAI_EMBEDDING_MODEL=text-embedding-3-small
+
+# =============================================================================
+# GRAPHITI: Anthropic Provider (LLM only)
+# =============================================================================
+# Use Anthropic for LLM. Requires separate embedder (use Voyage or OpenAI).
+# Example: GRAPHITI_LLM_PROVIDER=anthropic, GRAPHITI_EMBEDDER_PROVIDER=voyage
+#
+# Required: ANTHROPIC_API_KEY
+
+# Anthropic API Key
+# ANTHROPIC_API_KEY=sk-ant-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
+
+# Anthropic Model (default: claude-sonnet-4-5-latest)
+# GRAPHITI_ANTHROPIC_MODEL=claude-sonnet-4-5-latest
+
+# =============================================================================
+# GRAPHITI: Voyage AI Provider (Embeddings only)
+# =============================================================================
+# Use Voyage AI for embeddings. Commonly paired with Anthropic LLM.
+# Get API key from: https://www.voyageai.com/
+#
+# Required: VOYAGE_API_KEY
+
+# Voyage AI API Key
+# VOYAGE_API_KEY=pa-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
+
+# Voyage Embedding Model (default: voyage-3)
+# Available: voyage-3 (1024 dim), voyage-3-lite (512 dim)
+# VOYAGE_EMBEDDING_MODEL=voyage-3
+
+# =============================================================================
+# GRAPHITI: Google AI Provider
+# =============================================================================
+# Use Google AI (Gemini) for both LLM and embeddings.
+# Get API key from: https://aistudio.google.com/apikey
+#
+# Required: GOOGLE_API_KEY
+
+# Google AI API Key
+# GOOGLE_API_KEY=AIzaSyxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
+
+# Google LLM Model (default: gemini-2.0-flash)
+# GOOGLE_LLM_MODEL=gemini-2.0-flash
+
+# Google Embedding Model (default: text-embedding-004)
+# GOOGLE_EMBEDDING_MODEL=text-embedding-004
+
+# =============================================================================
+# GRAPHITI: OpenRouter Provider (Multi-provider aggregator)
+# =============================================================================
+# Use OpenRouter to access multiple LLM providers through a single API.
+# OpenRouter provides access to Anthropic, OpenAI, Google, and many other models.
+# Get API key from: https://openrouter.ai/keys
+#
+# Required: OPENROUTER_API_KEY
+
+# OpenRouter API Key
+# OPENROUTER_API_KEY=sk-or-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
+
+# OpenRouter Base URL (default: https://openrouter.ai/api/v1)
+# OPENROUTER_BASE_URL=https://openrouter.ai/api/v1
+
+# OpenRouter LLM Model (default: anthropic/claude-sonnet-4)
+# Popular choices: anthropic/claude-sonnet-4, openai/gpt-4o, google/gemini-2.0-flash
+# OPENROUTER_LLM_MODEL=anthropic/claude-sonnet-4
+
+# OpenRouter Embedding Model (default: openai/text-embedding-3-small)
+# OPENROUTER_EMBEDDING_MODEL=openai/text-embedding-3-small
+
+# =============================================================================
+# GRAPHITI: Azure OpenAI Provider
+# =============================================================================
+# Use Azure OpenAI for both LLM and embeddings.
+# Requires Azure OpenAI deployment with appropriate models.
+#
+# Required: AZURE_OPENAI_API_KEY, AZURE_OPENAI_BASE_URL
+
+# Azure OpenAI API Key
+# AZURE_OPENAI_API_KEY=xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
+
+# Azure OpenAI Base URL (your Azure endpoint)
+# AZURE_OPENAI_BASE_URL=https://your-resource.openai.azure.com/openai/deployments/your-deployment
+
+# Azure OpenAI Deployment Names
+# AZURE_OPENAI_LLM_DEPLOYMENT=gpt-4
+# AZURE_OPENAI_EMBEDDING_DEPLOYMENT=text-embedding-3-small
+
+# =============================================================================
+# GRAPHITI: Ollama Provider (Local/Offline)
+# =============================================================================
+# Use Ollama for fully offline operation. No API keys required.
+# Requires Ollama running locally with appropriate models pulled.
+#
+# Prerequisites:
+#   1. Install Ollama: https://ollama.ai/
+#   2. Pull models: ollama pull deepseek-r1:7b && ollama pull nomic-embed-text
+#   3. Start Ollama server (usually auto-starts)
+#
+# Required: OLLAMA_LLM_MODEL, OLLAMA_EMBEDDING_MODEL, OLLAMA_EMBEDDING_DIM
+
+# Ollama Server URL (default: http://localhost:11434)
+# OLLAMA_BASE_URL=http://localhost:11434
+
+# Ollama LLM Model
+# Popular choices: deepseek-r1:7b, llama3.2:3b, mistral:7b, phi3:medium
+# OLLAMA_LLM_MODEL=deepseek-r1:7b
+
+# Ollama Embedding Model
+# Popular choices: nomic-embed-text (768 dim), mxbai-embed-large (1024 dim)
+# OLLAMA_EMBEDDING_MODEL=nomic-embed-text
+
+# Ollama Embedding Dimension (REQUIRED for Ollama embeddings)
+# Must match your embedding model's output dimension
+# Common values: nomic-embed-text=768, mxbai-embed-large=1024, all-minilm=384
+# OLLAMA_EMBEDDING_DIM=768
+
+# =============================================================================
+# GRAPHITI: Example Configurations
+# =============================================================================
+#
+# --- Example 1: OpenAI (simplest) ---
+# GRAPHITI_ENABLED=true
+# GRAPHITI_LLM_PROVIDER=openai
+# GRAPHITI_EMBEDDER_PROVIDER=openai
+# OPENAI_API_KEY=sk-xxxxxxxx
+#
+# --- Example 2: Anthropic + Voyage (high quality) ---
+# GRAPHITI_ENABLED=true
+# GRAPHITI_LLM_PROVIDER=anthropic
+# GRAPHITI_EMBEDDER_PROVIDER=voyage
+# ANTHROPIC_API_KEY=sk-ant-xxxxxxxx
+# VOYAGE_API_KEY=pa-xxxxxxxx
+#
+# --- Example 3: Ollama (fully offline) ---
+# GRAPHITI_ENABLED=true
+# GRAPHITI_LLM_PROVIDER=ollama
+# GRAPHITI_EMBEDDER_PROVIDER=ollama
+# OLLAMA_LLM_MODEL=deepseek-r1:7b
+# OLLAMA_EMBEDDING_MODEL=nomic-embed-text
+# OLLAMA_EMBEDDING_DIM=768
+#
+# --- Example 4: Azure OpenAI (enterprise) ---
+# GRAPHITI_ENABLED=true
+# GRAPHITI_LLM_PROVIDER=azure_openai
+# GRAPHITI_EMBEDDER_PROVIDER=azure_openai
+# AZURE_OPENAI_API_KEY=xxxxxxxx
+# AZURE_OPENAI_BASE_URL=https://your-resource.openai.azure.com/...
+# AZURE_OPENAI_LLM_DEPLOYMENT=gpt-4
+# AZURE_OPENAI_EMBEDDING_DEPLOYMENT=text-embedding-3-small
+#
+# --- Example 5: Google AI (Gemini) ---
+# GRAPHITI_ENABLED=true
+# GRAPHITI_LLM_PROVIDER=google
+# GRAPHITI_EMBEDDER_PROVIDER=google
+# GOOGLE_API_KEY=AIzaSyxxxxxxxx
+#
+# --- Example 6: OpenRouter (multi-provider aggregator) ---
+# GRAPHITI_ENABLED=true
+# GRAPHITI_LLM_PROVIDER=openrouter
+# GRAPHITI_EMBEDDER_PROVIDER=openrouter
+# OPENROUTER_API_KEY=sk-or-xxxxxxxx
+# OPENROUTER_LLM_MODEL=anthropic/claude-sonnet-4
+# OPENROUTER_EMBEDDING_MODEL=openai/text-embedding-3-small
@@ -0,0 +1,77 @@
+# Environment files
+.env
+.env.local
+.env.*.local
+
+# Virtual environment
+.venv/
+.venv*/
+venv/
+env/
+
+# Python cache
+__pycache__/
+*.py[cod]
+*$py.class
+*.so
+
+# Distribution / packaging
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+
+# Puppeteer / Browser automation
+puppeteer_logs/
+puppeteer-*.log
+*.screenshot.png
+screenshots/
+.puppeteerrc.*
+chrome-profile/
+chromium-profile/
+
+# IDE
+.idea/
+.vscode/
+*.swp
+*.swo
+
+# OS
+.DS_Store
+Thumbs.db
+
+# Git worktrees (used by parallel mode)
+.worktrees/
+
+# Claude Code settings (project-specific)
+.claude_settings.json
+.auto-build-security.json
+
+# Tests (development only)
+tests/
+
+# Exceptions: Allow specific test directories
+!integrations/graphiti/tests/
+!tests/integration/
+
+# Auto Claude data directory
+.auto-claude/
+coverage.json
+
+# Auto Claude generated files
+.auto-claude-security.json
+.auto-claude-status
+.security-key
+logs/security/
@@ -0,0 +1,122 @@
+# Auto Claude Backend
+
+Autonomous coding framework powered by Claude AI. Builds software features through coordinated multi-agent sessions.
+
+## Getting Started
+
+### 1. Install
+
+```bash
+cd apps/backend
+python -m pip install -r requirements.txt
+```
+
+### 2. Configure
+
+```bash
+cp .env.example .env
+```
+
+Authenticate with Claude Code (token auto-saved to Keychain):
+```bash
+claude
+# Type: /login
+# Press Enter to open browser
+```
+
+Token is auto-detected from macOS Keychain / Windows Credential Manager.
+
+### 3. Run
+
+```bash
+# List available specs
+python run.py --list
+
+# Run a spec
+python run.py --spec 001
+```
+
+## Requirements
+
+- Python 3.10+
+- Claude API token
+
+## Commands
+
+| Command | Description |
+|---------|-------------|
+| `--list` | List all specs |
+| `--spec 001` | Run spec 001 |
+| `--spec 001 --isolated` | Run in isolated workspace |
+| `--spec 001 --direct` | Run directly in repo |
+| `--spec 001 --merge` | Merge completed build |
+| `--spec 001 --review` | Review build changes |
+| `--spec 001 --discard` | Discard build |
+| `--spec 001 --qa` | Run QA validation |
+| `--list-worktrees` | List all worktrees |
+| `--help` | Show all options |
+
+## Configuration
+
+Optional `.env` settings:
+
+| Variable | Description |
+|----------|-------------|
+| `AUTO_BUILD_MODEL` | Override Claude model |
+| `DEBUG=true` | Enable debug logging |
+| `LINEAR_API_KEY` | Enable Linear integration |
+| `GRAPHITI_ENABLED=true` | Enable memory system |
+
+## Troubleshooting
+
+**"tree-sitter not available"** - Safe to ignore, uses regex fallback.
+
+**Missing module errors** - Run `python -m pip install -r requirements.txt`
+
+**Debug mode** - Set `DEBUG=true DEBUG_LEVEL=2` before running.
+
+---
+
+## For Developers
+
+### Project Structure
+
+```
+backend/
+├── agents/          # AI agent execution
+├── analysis/        # Code analysis
+├── cli/             # Command-line interface
+├── core/            # Core utilities
+├── integrations/    # External services (Linear, Graphiti)
+├── merge/           # Git merge handling
+├── project/         # Project detection
+├── prompts/         # Prompt templates
+├── qa/              # QA validation
+├── spec/            # Spec management
+└── ui/              # Terminal UI
+```
+
+### Design Principles
+
+- **SOLID** - Single responsibility, clean interfaces
+- **DRY** - Shared utilities in `core/`
+- **KISS** - Simple flat imports via facade modules
+
+### Import Convention
+
+```python
+# Use facade modules for clean imports
+from debug import debug, debug_error
+from progress import count_subtasks
+from workspace import setup_workspace
+```
+
+### Adding Features
+
+1. Create module in appropriate folder
+2. Export API in `__init__.py`
+3. Add facade module at root if commonly imported
+
+## License
+
+AGPL-3.0
@@ -0,0 +1,23 @@
+"""
+Auto Claude Backend - Autonomous Coding Framework
+==================================================
+
+Multi-agent autonomous coding framework that builds software through
+coordinated AI agent sessions.
+
+This package provides:
+- Autonomous agent execution for building features from specs
+- Workspace isolation via git worktrees
+- QA validation loops
+- Memory management (Graphiti + file-based)
+- Linear integration for project management
+
+Quick Start:
+    python run.py --spec 001    # Run a spec
+    python run.py --list        # List all specs
+
+See README.md for full documentation.
+"""
+
+__version__ = "2.7.6-beta.5"
+__author__ = "Auto Claude Team"
@@ -0,0 +1,3 @@
+"""Backward compatibility shim - import from core.agent instead."""
+
+from core.agent import *  # noqa: F403
@@ -0,0 +1,152 @@
+# Agents Module
+
+Modular agent system for autonomous coding. This module refactors the original monolithic `agent.py` (1,446 lines) into focused, maintainable modules.
+
+## Architecture
+
+The agent system is now organized by concern:
+
+```
+auto-claude/agents/
+├── __init__.py          # Public API exports
+├── base.py              # Shared constants and imports
+├── utils.py             # Git operations and plan management
+├── memory.py            # Memory management (Graphiti + file-based)
+├── session.py           # Agent session execution
+├── planner.py           # Follow-up planner logic
+└── coder.py             # Main autonomous agent loop
+```
+
+## Modules
+
+### `base.py` (352 bytes)
+- Shared constants (`AUTO_CONTINUE_DELAY_SECONDS`, `HUMAN_INTERVENTION_FILE`)
+- Common imports and logging setup
+
+### `utils.py` (3.6 KB)
+- Git operations: `get_latest_commit()`, `get_commit_count()`
+- Plan management: `load_implementation_plan()`, `find_subtask_in_plan()`, `find_phase_for_subtask()`
+- Workspace sync: `sync_spec_to_source()`
+
+### `memory.py` (13 KB)
+- Dual-layer memory system (Graphiti primary, file-based fallback)
+- `debug_memory_system_status()` - Memory system diagnostics
+- `get_graphiti_context()` - Retrieve relevant context for subtasks
+- `save_session_memory()` - Save session insights to memory
+- `save_session_to_graphiti()` - Backwards compatibility wrapper
+
+### `session.py` (17 KB)
+- `run_agent_session()` - Execute a single agent session
+- `post_session_processing()` - Process results and update memory
+- Session logging and tool tracking
+- Recovery manager integration
+
+### `planner.py` (5.4 KB)
+- `run_followup_planner()` - Add new subtasks to completed specs
+- Follow-up planning workflow
+- Plan validation and status updates
+
+### `coder.py` (16 KB)
+- `run_autonomous_agent()` - Main autonomous agent loop
+- Planning and coding phase management
+- Linear integration
+- Recovery and stuck subtask handling
+
+## Public API
+
+The `agents` module exports a clean public API:
+
+```python
+from agents import (
+    # Main functions
+    run_autonomous_agent,
+    run_followup_planner,
+
+    # Memory functions
+    save_session_memory,
+    get_graphiti_context,
+
+    # Session management
+    run_agent_session,
+    post_session_processing,
+
+    # Utilities
+    get_latest_commit,
+    load_implementation_plan,
+    sync_spec_to_source,
+)
+```
+
+## Backwards Compatibility
+
+The original `agent.py` is now a facade that re-exports everything from the `agents` module:
+
+```python
+# Old code still works
+from agent import run_autonomous_agent, save_session_memory
+
+# New code can use modular imports
+from agents.coder import run_autonomous_agent
+from agents.memory import save_session_memory
+```
+
+All existing imports continue to work without changes.
+
+## Benefits
+
+1. **Separation of Concerns**: Each module has a clear, focused responsibility
+2. **Maintainability**: Easier to understand and modify individual components
+3. **Testability**: Modules can be tested in isolation
+4. **Backwards Compatible**: No breaking changes to existing code
+5. **Scalability**: Easy to add new agent types or features
+
+## Module Dependencies
+
+```
+coder.py
+  ├── session.py (run_agent_session, post_session_processing)
+  ├── memory.py (get_graphiti_context, debug_memory_system_status)
+  └── utils.py (git operations, plan management)
+
+session.py
+  ├── memory.py (save_session_memory)
+  └── utils.py (git operations, plan management)
+
+planner.py
+  └── session.py (run_agent_session)
+
+memory.py
+  └── base.py (constants, logging)
+```
+
+## Testing
+
+Run the verification script to test the refactoring:
+
+```bash
+python3 auto-claude/agents/test_refactoring.py
+```
+
+This verifies:
+- Module structure is correct
+- All imports work
+- Public API is accessible
+- Backwards compatibility is maintained
+
+## Migration Guide
+
+No migration needed! The refactoring maintains 100% backwards compatibility.
+
+### For new code:
+```python
+# Use focused imports for clarity
+from agents.coder import run_autonomous_agent
+from agents.memory import save_session_memory, get_graphiti_context
+from agents.session import run_agent_session
+```
+
+### For existing code:
+```python
+# Old imports continue to work
+from agent import run_autonomous_agent, save_session_memory
+```
@@ -0,0 +1,96 @@
+"""
+Agents Module
+=============
+
+Modular agent system for autonomous coding.
+
+This module provides:
+- run_autonomous_agent: Main coder agent loop
+- run_followup_planner: Follow-up planner for completed specs
+- Memory management (Graphiti + file-based fallback)
+- Session management and post-processing
+- Utility functions for git and plan management
+
+Uses lazy imports to avoid circular dependencies.
+"""
+
+# Explicit import required by CodeQL static analysis
+# (CodeQL doesn't recognize __getattr__ dynamic exports)
+from .utils import sync_spec_to_source
+
+__all__ = [
+    # Main API
+    "run_autonomous_agent",
+    "run_followup_planner",
+    # Memory
+    "debug_memory_system_status",
+    "get_graphiti_context",
+    "save_session_memory",
+    "save_session_to_graphiti",
+    # Session
+    "run_agent_session",
+    "post_session_processing",
+    # Utils
+    "get_latest_commit",
+    "get_commit_count",
+    "load_implementation_plan",
+    "find_subtask_in_plan",
+    "find_phase_for_subtask",
+    "sync_spec_to_source",
+    # Constants
+    "AUTO_CONTINUE_DELAY_SECONDS",
+    "HUMAN_INTERVENTION_FILE",
+]
+
+
+def __getattr__(name):
+    """Lazy imports to avoid circular dependencies."""
+    if name in ("AUTO_CONTINUE_DELAY_SECONDS", "HUMAN_INTERVENTION_FILE"):
+        from .base import AUTO_CONTINUE_DELAY_SECONDS, HUMAN_INTERVENTION_FILE
+
+        return locals()[name]
+    elif name == "run_autonomous_agent":
+        from .coder import run_autonomous_agent
+
+        return run_autonomous_agent
+    elif name in (
+        "debug_memory_system_status",
+        "get_graphiti_context",
+        "save_session_memory",
+        "save_session_to_graphiti",
+    ):
+        from .memory_manager import (
+            debug_memory_system_status,
+            get_graphiti_context,
+            save_session_memory,
+            save_session_to_graphiti,
+        )
+
+        return locals()[name]
+    elif name == "run_followup_planner":
+        from .planner import run_followup_planner
+
+        return run_followup_planner
+    elif name in ("post_session_processing", "run_agent_session"):
+        from .session import post_session_processing, run_agent_session
+
+        return locals()[name]
+    elif name in (
+        "find_phase_for_subtask",
+        "find_subtask_in_plan",
+        "get_commit_count",
+        "get_latest_commit",
+        "load_implementation_plan",
+        "sync_spec_to_source",
+    ):
+        from .utils import (
+            find_phase_for_subtask,
+            find_subtask_in_plan,
+            get_commit_count,
+            get_latest_commit,
+            load_implementation_plan,
+            sync_spec_to_source,
+        )
+
+        return locals()[name]
+    raise AttributeError(f"module 'agents' has no attribute '{name}'")
@@ -0,0 +1,99 @@
+"""
+Base Module for Agent System
+=============================
+
+Shared imports, types, and constants used across agent modules.
+"""
+
+import logging
+import re
+
+# Configure logging
+logger = logging.getLogger(__name__)
+
+# Configuration constants
+AUTO_CONTINUE_DELAY_SECONDS = 3
+HUMAN_INTERVENTION_FILE = "PAUSE"
+
+# Retry configuration for subtask execution
+MAX_SUBTASK_RETRIES = 5  # Maximum attempts before marking subtask as stuck
+
+# Retry configuration for 400 tool concurrency errors
+MAX_CONCURRENCY_RETRIES = 5  # Maximum number of retries for tool concurrency errors
+INITIAL_RETRY_DELAY_SECONDS = (
+    2  # Initial retry delay (doubles each retry: 2s, 4s, 8s, 16s, 32s)
+)
+MAX_RETRY_DELAY_SECONDS = 32  # Cap retry delay at 32 seconds
+
+# Pause file constants for intelligent error recovery
+# These files signal pause/resume between frontend and backend
+RATE_LIMIT_PAUSE_FILE = "RATE_LIMIT_PAUSE"  # Created when rate limited
+AUTH_FAILURE_PAUSE_FILE = "AUTH_PAUSE"  # Created when auth fails
+RESUME_FILE = "RESUME"  # Created by frontend to signal resume
+
+# Maximum time to wait for rate limit reset (2 hours)
+# If reset time is beyond this, task should fail rather than wait indefinitely
+MAX_RATE_LIMIT_WAIT_SECONDS = 7200
+
+# Wait intervals for pause/resume checking
+RATE_LIMIT_CHECK_INTERVAL_SECONDS = (
+    30  # Check for RESUME file every 30 seconds during rate limit wait
+)
+AUTH_RESUME_CHECK_INTERVAL_SECONDS = 10  # Check for re-authentication every 10 seconds
+AUTH_RESUME_MAX_WAIT_SECONDS = 86400  # Maximum wait for re-authentication (24 hours)
+
+
+def sanitize_error_message(error_message: str, max_length: int = 500) -> str:
+    """
+    Sanitize error messages to remove potentially sensitive information.
+
+    Redacts:
+    - API keys (sk-..., key-...)
+    - Bearer tokens
+    - Token/secret values
+
+    Args:
+        error_message: The raw error message to sanitize
+        max_length: Maximum length to truncate to (default 500)
+
+    Returns:
+        Sanitized and truncated error message
+    """
+    if not error_message:
+        return ""
+
+    # Redact patterns that look like API keys or tokens
+    # Pattern: sk-... (OpenAI/Anthropic keys like sk-ant-api03-...)
+    sanitized = re.sub(
+        r"\bsk-[a-zA-Z0-9._\-]{20,}\b", "[REDACTED_API_KEY]", error_message
+    )
+
+    # Pattern: key-... (generic API keys)
+    sanitized = re.sub(r"\bkey-[a-zA-Z0-9._\-]{20,}\b", "[REDACTED_API_KEY]", sanitized)
+
+    # Pattern: Bearer ... (bearer tokens)
+    sanitized = re.sub(
+        r"\bBearer\s+[a-zA-Z0-9._\-]{20,}\b", "Bearer [REDACTED_TOKEN]", sanitized
+    )
+
+    # Pattern: token= or token: followed by long strings
+    sanitized = re.sub(
+        r"(token[=:]\s*)[a-zA-Z0-9._\-]{20,}\b",
+        r"\1[REDACTED_TOKEN]",
+        sanitized,
+        flags=re.IGNORECASE,
+    )
+
+    # Pattern: secret= or secret: followed by strings
+    sanitized = re.sub(
+        r"(secret[=:]\s*)[a-zA-Z0-9._\-]{20,}\b",
+        r"\1[REDACTED_SECRET]",
+        sanitized,
+        flags=re.IGNORECASE,
+    )
+
+    # Truncate to max length
+    if len(sanitized) > max_length:
+        sanitized = sanitized[:max_length] + "..."
+
+    return sanitized
@@ -0,0 +1,97 @@
+"""
+Investigation context loading for agents.
+
+Provides utilities to load investigation data from spec directories
+for GitHub-sourced tasks.
+"""
+
+import json
+from pathlib import Path
+from typing import Any
+
+
+def load_investigation_context(spec_dir: Path) -> dict[str, Any] | None:
+    """
+    Load investigation context if this spec was created from a GitHub issue.
+
+    Args:
+        spec_dir: Path to the spec directory
+
+    Returns:
+        Structured investigation context with root_cause, fix_approaches,
+        reproducer, gotchas, and patterns_to_follow, or None if no
+        investigation data exists.
+    """
+    investigation_report_path = spec_dir / "investigation_report.json"
+
+    if not investigation_report_path.exists():
+        return None
+
+    try:
+        with open(investigation_report_path) as f:
+            report = json.load(f)
+
+        root_cause = report.get("root_cause", {})
+        fix_advice = report.get("fix_advice", {})
+        reproduction = report.get("reproduction", {})
+
+        # Structure the context for agents
+        return {
+            "root_cause": {
+                "summary": root_cause.get("identified_root_cause"),
+                "evidence": root_cause.get("evidence", ""),
+                "code_paths": root_cause.get("code_paths", []),
+            },
+            "fix_approaches": fix_advice.get("approaches", []),
+            "reproducer": reproduction if reproduction else None,
+            "gotchas": fix_advice.get("gotchas", []),
+            "patterns_to_follow": fix_advice.get("patterns_to_follow", []),
+            "impact": report.get("impact", {}),
+        }
+    except (json.JSONDecodeError, OSError):
+        return None
+
+
+def load_investigation_for_qa(
+    spec_dir: Path, base_branch: str
+) -> dict[str, Any] | None:
+    """
+    Load investigation context for QA validation.
+
+    Similar to load_investigation_context but includes base_branch
+    for QA comparison.
+
+    Args:
+        spec_dir: Path to the spec directory
+        base_branch: Base branch to compare against (e.g., 'main', 'develop')
+
+    Returns:
+        Structured investigation context with root_cause, reproducer,
+        impact, expected_outcome, and base_branch, or None if no
+        investigation data exists.
+    """
+    investigation_report_path = spec_dir / "investigation_report.json"
+
+    if not investigation_report_path.exists():
+        return None
+
+    try:
+        with open(investigation_report_path) as f:
+            report = json.load(f)
+
+        root_cause = report.get("root_cause", {})
+        reproduction = report.get("reproduction", {})
+
+        return {
+            "root_cause": {
+                "summary": root_cause.get("identified_root_cause"),
+                "evidence": root_cause.get("evidence", ""),
+                "code_paths": root_cause.get("code_paths", []),
+            },
+            "reproducer": reproduction if reproduction else None,
+            "impact": report.get("impact", {}),
+            "expected_outcome": report.get("ai_summary"),
+            "base_branch": base_branch,
+        }
+    except (json.JSONDecodeError, OSError):
+        return None
@@ -0,0 +1,494 @@
+"""
+Memory Management for Agent System
+===================================
+
+Handles session memory storage using dual-layer approach:
+- PRIMARY: Graphiti (when enabled) - semantic search, cross-session context
+- FALLBACK: File-based memory - zero dependencies, always available
+"""
+
+import logging
+from pathlib import Path
+
+from core.sentry import capture_exception
+from debug import (
+    debug,
+    debug_detailed,
+    debug_error,
+    debug_section,
+    debug_success,
+    debug_warning,
+    is_debug_enabled,
+)
+from graphiti_config import get_graphiti_status, is_graphiti_enabled
+
+# Import from parent memory package
+# Now safe since this module is named memory_manager (not memory)
+from memory import save_session_insights as save_file_based_memory
+from memory.graphiti_helpers import get_graphiti_memory
+
+logger = logging.getLogger(__name__)
+
+
+def debug_memory_system_status() -> None:
+    """
+    Print memory system status for debugging.
+
+    Called at startup when DEBUG=true to show memory configuration.
+    """
+    if not is_debug_enabled():
+        return
+
+    debug_section("memory", "Memory System Status")
+
+    # Get Graphiti status
+    graphiti_status = get_graphiti_status()
+
+    debug(
+        "memory",
+        "Memory system configuration",
+        primary_system="Graphiti"
+        if graphiti_status.get("available")
+        else "File-based (fallback)",
+        graphiti_enabled=graphiti_status.get("enabled"),
+        graphiti_available=graphiti_status.get("available"),
+    )
+
+    if graphiti_status.get("enabled"):
+        debug_detailed(
+            "memory",
+            "Graphiti configuration",
+            host=graphiti_status.get("host"),
+            port=graphiti_status.get("port"),
+            database=graphiti_status.get("database"),
+            llm_provider=graphiti_status.get("llm_provider"),
+            embedder_provider=graphiti_status.get("embedder_provider"),
+        )
+
+        if not graphiti_status.get("available"):
+            debug_warning(
+                "memory",
+                "Graphiti not available",
+                reason=graphiti_status.get("reason"),
+                errors=graphiti_status.get("errors"),
+            )
+            debug("memory", "Will use file-based memory as fallback")
+        else:
+            debug_success("memory", "Graphiti ready as PRIMARY memory system")
+    else:
+        debug(
+            "memory",
+            "Graphiti disabled, using file-based memory only",
+            note="Set GRAPHITI_ENABLED=true to enable Graphiti",
+        )
+
+
+async def get_graphiti_context(
+    spec_dir: Path,
+    project_dir: Path,
+    subtask: dict,
+) -> str | None:
+    """
+    Retrieve relevant context from Graphiti for the current subtask.
+
+    This searches the knowledge graph for context relevant to the subtask's
+    task description, returning past insights, patterns, and gotchas.
+
+    Args:
+        spec_dir: Spec directory
+        project_dir: Project root directory
+        subtask: The current subtask being worked on
+
+    Returns:
+        Formatted context string or None if unavailable
+    """
+    if is_debug_enabled():
+        debug(
+            "memory",
+            "Retrieving Graphiti context for subtask",
+            subtask_id=subtask.get("id", "unknown"),
+            subtask_desc=subtask.get("description", "")[:100],
+        )
+
+    if not is_graphiti_enabled():
+        if is_debug_enabled():
+            debug("memory", "Graphiti not enabled, skipping context retrieval")
+        return None
+
+    memory = None
+    try:
+        # Use centralized helper for GraphitiMemory instantiation (async)
+        memory = await get_graphiti_memory(spec_dir, project_dir)
+        if memory is None:
+            if is_debug_enabled():
+                debug_warning(
+                    "memory", "GraphitiMemory not available for context retrieval"
+                )
+            return None
+
+        # Build search query from subtask description
+        subtask_desc = subtask.get("description", "")
+        subtask_id = subtask.get("id", "")
+        query = f"{subtask_desc} {subtask_id}".strip()
+
+        if not query:
+            if is_debug_enabled():
+                debug_warning("memory", "Empty query, skipping context retrieval")
+            return None
+
+        if is_debug_enabled():
+            debug_detailed(
+                "memory",
+                "Searching Graphiti knowledge graph",
+                query=query[:200],
+                num_results=5,
+            )
+
+        # Get relevant context
+        context_items = await memory.get_relevant_context(query, num_results=5)
+
+        # Get patterns and gotchas specifically (THE FIX for learning loop!)
+        # This retrieves PATTERN and GOTCHA episode types for cross-session learning
+        patterns, gotchas = await memory.get_patterns_and_gotchas(
+            query, num_results=3, min_score=0.5
+        )
+
+        # Also get recent session history
+        session_history = await memory.get_session_history(limit=3)
+
+        if is_debug_enabled():
+            debug(
+                "memory",
+                "Graphiti context retrieval complete",
+                context_items_found=len(context_items) if context_items else 0,
+                patterns_found=len(patterns) if patterns else 0,
+                gotchas_found=len(gotchas) if gotchas else 0,
+                session_history_found=len(session_history) if session_history else 0,
+            )
+
+        if not context_items and not session_history and not patterns and not gotchas:
+            if is_debug_enabled():
+                debug("memory", "No relevant context found in Graphiti")
+            return None
+
+        # Format the context
+        sections = ["## Graphiti Memory Context\n"]
+        sections.append("_Retrieved from knowledge graph for this subtask:_\n")
+
+        if context_items:
+            sections.append("### Relevant Knowledge\n")
+            for item in context_items:
+                content = item.get("content", "")[:500]  # Truncate
+                item_type = item.get("type", "unknown")
+                sections.append(f"- **[{item_type}]** {content}\n")
+
+        # Add patterns section (cross-session learning)
+        if patterns:
+            sections.append("### Learned Patterns\n")
+            sections.append("_Patterns discovered in previous sessions:_\n")
+            for p in patterns:
+                pattern_text = p.get("pattern", "")
+                applies_to = p.get("applies_to", "")
+                if applies_to:
+                    sections.append(
+                        f"- **Pattern**: {pattern_text}\n  _Applies to:_ {applies_to}\n"
+                    )
+                else:
+                    sections.append(f"- **Pattern**: {pattern_text}\n")
+
+        # Add gotchas section (cross-session learning)
+        if gotchas:
+            sections.append("### Known Gotchas\n")
+            sections.append("_Pitfalls to avoid:_\n")
+            for g in gotchas:
+                gotcha_text = g.get("gotcha", "")
+                solution = g.get("solution", "")
+                if solution:
+                    sections.append(
+                        f"- **Gotcha**: {gotcha_text}\n  _Solution:_ {solution}\n"
+                    )
+                else:
+                    sections.append(f"- **Gotcha**: {gotcha_text}\n")
+
+        if session_history:
+            sections.append("### Recent Session Insights\n")
+            for session in session_history[:2]:  # Only show last 2
+                session_num = session.get("session_number", "?")
+                recommendations = session.get("recommendations_for_next_session", [])
+                if recommendations:
+                    sections.append(f"**Session {session_num} recommendations:**")
+                    for rec in recommendations[:3]:  # Limit to 3
+                        sections.append(f"- {rec}")
+                    sections.append("")
+
+        if is_debug_enabled():
+            debug_success(
+                "memory", "Graphiti context formatted", total_sections=len(sections)
+            )
+
+        return "\n".join(sections)
+
+    except Exception as e:
+        logger.warning(f"Failed to get Graphiti context: {e}")
+        if is_debug_enabled():
+            debug_error("memory", "Graphiti context retrieval failed", error=str(e))
+        # Capture exception to Sentry with full context
+        capture_exception(
+            e,
+            operation="get_graphiti_context",
+            subtask_id=subtask.get("id", "unknown"),
+            subtask_desc=subtask.get("description", "")[:200],
+            spec_dir=str(spec_dir),
+            project_dir=str(project_dir),
+        )
+        return None
+    finally:
+        # Always close the memory connection (swallow exceptions to avoid overriding)
+        if memory is not None:
+            try:
+                await memory.close()
+            except Exception as e:
+                logger.debug(
+                    "Failed to close Graphiti memory connection", exc_info=True
+                )
+
+
+async def save_session_memory(
+    spec_dir: Path,
+    project_dir: Path,
+    subtask_id: str,
+    session_num: int,
+    success: bool,
+    subtasks_completed: list[str],
+    discoveries: dict | None = None,
+) -> tuple[bool, str]:
+    """
+    Save session insights to memory.
+
+    Memory Strategy:
+    - PRIMARY: Graphiti (when enabled) - provides semantic search, cross-session context
+    - FALLBACK: File-based (when Graphiti is disabled) - zero dependencies, always works
+
+    This is called after each session to persist learnings.
+
+    Args:
+        spec_dir: Spec directory
+        project_dir: Project root directory
+        subtask_id: The subtask that was worked on
+        session_num: Current session number
+        success: Whether the subtask was completed successfully
+        subtasks_completed: List of subtask IDs completed this session
+        discoveries: Optional dict with file discoveries, patterns, gotchas
+
+    Returns:
+        Tuple of (success, storage_type) where storage_type is "graphiti" or "file"
+    """
+    # Debug: Log memory save start
+    if is_debug_enabled():
+        debug_section("memory", f"Saving Session {session_num} Memory")
+        debug(
+            "memory",
+            "Memory save initiated",
+            subtask_id=subtask_id,
+            session_num=session_num,
+            success=success,
+            subtasks_completed=subtasks_completed,
+            spec_dir=str(spec_dir),
+        )
+
+    # Build insights structure (same format for both storage systems)
+    insights = {
+        "subtasks_completed": subtasks_completed,
+        "discoveries": discoveries
+        or {
+            "files_understood": {},
+            "patterns_found": [],
+            "gotchas_encountered": [],
+        },
+        "what_worked": [f"Implemented subtask: {subtask_id}"] if success else [],
+        "what_failed": [] if success else [f"Failed to complete subtask: {subtask_id}"],
+        "recommendations_for_next_session": [],
+    }
+
+    if is_debug_enabled():
+        debug_detailed("memory", "Insights structure built", insights=insights)
+
+    # Check Graphiti status for debugging
+    graphiti_enabled = is_graphiti_enabled()
+    if is_debug_enabled():
+        graphiti_status = get_graphiti_status()
+        debug(
+            "memory",
+            "Graphiti status check",
+            enabled=graphiti_status.get("enabled"),
+            available=graphiti_status.get("available"),
+            host=graphiti_status.get("host"),
+            port=graphiti_status.get("port"),
+            database=graphiti_status.get("database"),
+            llm_provider=graphiti_status.get("llm_provider"),
+            embedder_provider=graphiti_status.get("embedder_provider"),
+            reason=graphiti_status.get("reason") or "OK",
+        )
+
+    # PRIMARY: Try Graphiti if enabled
+    if graphiti_enabled:
+        if is_debug_enabled():
+            debug("memory", "Attempting PRIMARY storage: Graphiti")
+
+        memory = None
+        try:
+            # Use centralized helper for GraphitiMemory instantiation (async)
+            memory = await get_graphiti_memory(spec_dir, project_dir)
+            if memory is None:
+                if is_debug_enabled():
+                    debug_warning("memory", "GraphitiMemory not available")
+                    debug(
+                        "memory",
+                        "get_graphiti_memory() returned None - this usually means Graphiti is disabled or provider config is invalid",
+                    )
+                # Continue to file-based fallback
+            if memory is not None and memory.is_enabled:
+                if is_debug_enabled():
+                    debug("memory", "Saving to Graphiti...")
+
+                # Use structured insights if we have rich extracted data
+                if discoveries and discoveries.get("file_insights"):
+                    # Rich insights from insight_extractor
+                    if is_debug_enabled():
+                        debug(
+                            "memory",
+                            "Using save_structured_insights (rich data available)",
+                        )
+                    result = await memory.save_structured_insights(discoveries)
+                else:
+                    # Fallback to basic session insights
+                    result = await memory.save_session_insights(session_num, insights)
+
+                if result:
+                    logger.info(
+                        f"Session {session_num} insights saved to Graphiti (primary)"
+                    )
+                    if is_debug_enabled():
+                        debug_success(
+                            "memory",
+                            f"Session {session_num} saved to Graphiti (PRIMARY)",
+                            storage_type="graphiti",
+                            subtasks_saved=len(subtasks_completed),
+                        )
+                    return True, "graphiti"
+                else:
+                    logger.warning(
+                        "Graphiti save returned False, falling back to file-based"
+                    )
+                    if is_debug_enabled():
+                        debug_warning(
+                            "memory", "Graphiti save returned False, using FALLBACK"
+                        )
+            elif memory is None:
+                if is_debug_enabled():
+                    debug_warning(
+                        "memory", "GraphitiMemory not available, using FALLBACK"
+                    )
+            else:
+                # memory is not None but memory.is_enabled is False
+                logger.warning(
+                    "GraphitiMemory.is_enabled=False, falling back to file-based"
+                )
+                if is_debug_enabled():
+                    debug_warning("memory", "GraphitiMemory disabled, using FALLBACK")
+
+        except Exception as e:
+            logger.warning(f"Graphiti save failed: {e}, falling back to file-based")
+            if is_debug_enabled():
+                debug_error("memory", "Graphiti save failed", error=str(e))
+            # Capture exception to Sentry with full context
+            capture_exception(
+                e,
+                operation="save_session_memory_graphiti",
+                subtask_id=subtask_id,
+                session_num=session_num,
+                success=success,
+                subtasks_completed=subtasks_completed,
+                spec_dir=str(spec_dir),
+                project_dir=str(project_dir),
+            )
+        finally:
+            # Always close the memory connection (swallow exceptions to avoid overriding)
+            if memory is not None:
+                try:
+                    await memory.close()
+                except Exception as e:
+                    logger.debug(
+                        "Failed to close Graphiti memory connection", exc_info=e
+                    )
+    else:
+        if is_debug_enabled():
+            debug("memory", "Graphiti not enabled, skipping to FALLBACK")
+
+    # FALLBACK: File-based memory (when Graphiti is disabled or fails)
+    if is_debug_enabled():
+        debug("memory", "Attempting FALLBACK storage: File-based")
+
+    try:
+        memory_dir = spec_dir / "memory" / "session_insights"
+        if is_debug_enabled():
+            debug_detailed(
+                "memory",
+                "File-based memory path",
+                memory_dir=str(memory_dir),
+                session_file=f"session_{session_num:03d}.json",
+            )
+
+        save_file_based_memory(spec_dir, session_num, insights)
+        logger.info(
+            f"Session {session_num} insights saved to file-based memory (fallback)"
+        )
+
+        if is_debug_enabled():
+            debug_success(
+                "memory",
+                f"Session {session_num} saved to file-based (FALLBACK)",
+                storage_type="file",
+                file_path=str(memory_dir / f"session_{session_num:03d}.json"),
+                subtasks_saved=len(subtasks_completed),
+            )
+        return True, "file"
+    except Exception as e:
+        logger.error(f"File-based memory save also failed: {e}")
+        if is_debug_enabled():
+            debug_error("memory", "File-based memory save FAILED", error=str(e))
+        # Capture exception to Sentry with full context
+        capture_exception(
+            e,
+            operation="save_session_memory_file",
+            subtask_id=subtask_id,
+            session_num=session_num,
+            success=success,
+            subtasks_completed=subtasks_completed,
+            spec_dir=str(spec_dir),
+            project_dir=str(project_dir),
+        )
+        return False, "none"
+
+
+# Keep the old function name as an alias for backwards compatibility
+async def save_session_to_graphiti(
+    spec_dir: Path,
+    project_dir: Path,
+    subtask_id: str,
+    session_num: int,
+    success: bool,
+    subtasks_completed: list[str],
+    discoveries: dict | None = None,
+) -> bool:
+    """Backwards compatibility wrapper for save_session_memory."""
+    result, _ = await save_session_memory(
+        spec_dir,
+        project_dir,
+        subtask_id,
+        session_num,
+        success,
+        subtasks_completed,
+        discoveries,
+    )
+    return result
@@ -0,0 +1,198 @@
+"""
+Planner Agent Module
+====================
+
+Handles follow-up planner sessions for adding new subtasks to completed specs.
+"""
+
+import logging
+from pathlib import Path
+
+from core.client import create_client
+from phase_config import (
+    get_fast_mode,
+    get_phase_client_thinking_kwargs,
+    get_phase_model,
+    get_phase_model_betas,
+)
+from phase_event import ExecutionPhase, emit_phase
+from task_logger import (
+    LogPhase,
+    get_task_logger,
+)
+from ui import (
+    BuildState,
+    Icons,
+    StatusManager,
+    bold,
+    box,
+    highlight,
+    icon,
+    muted,
+    print_status,
+)
+
+from .session import run_agent_session
+
+logger = logging.getLogger(__name__)
+
+
+async def run_followup_planner(
+    project_dir: Path,
+    spec_dir: Path,
+    model: str,
+    verbose: bool = False,
+) -> bool:
+    """
+    Run the follow-up planner to add new subtasks to a completed spec.
+
+    This is a simplified version of run_autonomous_agent that:
+    1. Creates a client
+    2. Loads the followup planner prompt
+    3. Runs a single planning session
+    4. Returns after the plan is updated (doesn't enter coding loop)
+
+    The planner agent will:
+    - Read FOLLOWUP_REQUEST.md for the new task
+    - Read the existing implementation_plan.json
+    - Add new phase(s) with pending subtasks
+    - Update the plan status back to in_progress
+
+    Args:
+        project_dir: Root directory for the project
+        spec_dir: Directory containing the completed spec
+        model: Claude model to use
+        verbose: Whether to show detailed output
+
+    Returns:
+        bool: True if planning completed successfully
+    """
+    from implementation_plan import ImplementationPlan
+    from prompts import get_followup_planner_prompt
+
+    # Initialize status manager for ccstatusline
+    status_manager = StatusManager(project_dir)
+    status_manager.set_active(spec_dir.name, BuildState.PLANNING)
+    emit_phase(ExecutionPhase.PLANNING, "Follow-up planning")
+
+    # Initialize task logger for persistent logging
+    task_logger = get_task_logger(spec_dir)
+
+    # Show header
+    content = [
+        bold(f"{icon(Icons.GEAR)} FOLLOW-UP PLANNER SESSION"),
+        "",
+        f"Spec: {highlight(spec_dir.name)}",
+        muted("Adding follow-up work to completed spec."),
+        "",
+        muted("The agent will read your FOLLOWUP_REQUEST.md and add new subtasks."),
+    ]
+    print()
+    print(box(content, width=70, style="heavy"))
+    print()
+
+    # Start planning phase in task logger
+    if task_logger:
+        task_logger.start_phase(LogPhase.PLANNING, "Starting follow-up planning...")
+        task_logger.set_session(1)
+
+    # Create client with phase-specific model and thinking budget
+    # Respects task_metadata.json configuration when no CLI override
+    planning_model = get_phase_model(spec_dir, "planning", model)
+    planning_betas = get_phase_model_betas(spec_dir, "planning", model)
+    thinking_kwargs = get_phase_client_thinking_kwargs(
+        spec_dir, "planning", planning_model
+    )
+    fast_mode = get_fast_mode(spec_dir)
+    logger.info(
+        f"[Planner] [Fast Mode] {'ENABLED' if fast_mode else 'disabled'} for follow-up planning"
+    )
+    client = create_client(
+        project_dir,
+        spec_dir,
+        planning_model,
+        agent_type="planner",
+        betas=planning_betas,
+        fast_mode=fast_mode,
+        **thinking_kwargs,
+    )
+
+    # Generate follow-up planner prompt
+    prompt = get_followup_planner_prompt(spec_dir)
+
+    print_status("Running follow-up planner...", "progress")
+    print()
+
+    try:
+        # Run single planning session
+        async with client:
+            status, response, error_info = await run_agent_session(
+                client, prompt, spec_dir, verbose, phase=LogPhase.PLANNING
+            )
+
+        # End planning phase in task logger
+        if task_logger:
+            task_logger.end_phase(
+                LogPhase.PLANNING,
+                success=(status != "error"),
+                message="Follow-up planning session completed",
+            )
+
+        if status == "error":
+            print()
+            print_status("Follow-up planning failed", "error")
+            status_manager.update(state=BuildState.ERROR)
+            return False
+
+        # Verify the plan was updated (should have pending subtasks now)
+        plan_file = spec_dir / "implementation_plan.json"
+        if plan_file.exists():
+            plan = ImplementationPlan.load(plan_file)
+
+            # Check if there are any pending subtasks
+            all_subtasks = [c for p in plan.phases for c in p.subtasks]
+            pending_subtasks = [c for c in all_subtasks if c.status.value == "pending"]
+
+            if pending_subtasks:
+                # Reset the plan status to in_progress (in case planner didn't)
+                plan.reset_for_followup()
+                await plan.async_save(plan_file)
+
+                print()
+                content = [
+                    bold(f"{icon(Icons.SUCCESS)} FOLLOW-UP PLANNING COMPLETE"),
+                    "",
+                    f"New pending subtasks: {highlight(str(len(pending_subtasks)))}",
+                    f"Total subtasks: {len(all_subtasks)}",
+                    "",
+                    muted("Next steps:"),
+                    f"  Run: {highlight(f'python auto-claude/run.py --spec {spec_dir.name}')}",
+                ]
+                print(box(content, width=70, style="heavy"))
+                print()
+                status_manager.update(state=BuildState.PAUSED)
+                return True
+            else:
+                print()
+                print_status(
+                    "Warning: No pending subtasks found after planning", "warning"
+                )
+                print(muted("The planner may not have added new subtasks."))
+                print(muted("Check implementation_plan.json manually."))
+                status_manager.update(state=BuildState.PAUSED)
+                return False
+        else:
+            print()
+            print_status(
+                "Error: implementation_plan.json not found after planning", "error"
+            )
+            status_manager.update(state=BuildState.ERROR)
+            return False
+
+    except Exception as e:
+        print()
+        print_status(f"Follow-up planning error: {e}", "error")
+        if task_logger:
+            task_logger.log_error(f"Follow-up planning error: {e}", LogPhase.PLANNING)
+        status_manager.update(state=BuildState.ERROR)
+        return False
@@ -0,0 +1,347 @@
+"""
+PR Template Filler Agent Module
+================================
+
+Detects GitHub PR templates in a project and uses Claude to intelligently
+fill them based on code changes, spec context, commit history, and branch info.
+"""
+
+import logging
+from pathlib import Path
+
+from core.client import create_client
+from task_logger import LogPhase, get_task_logger
+
+from .session import run_agent_session
+
+logger = logging.getLogger(__name__)
+
+# Maximum diff size (in characters) before truncating to file-level summaries
+MAX_DIFF_CHARS = 30_000
+
+
+def detect_pr_template(project_dir: Path | str) -> str | None:
+    """
+    Detect a GitHub PR template in the project.
+
+    Searches for:
+    1. .github/PULL_REQUEST_TEMPLATE.md (single template)
+    2. .github/PULL_REQUEST_TEMPLATE/ directory (picks the first .md file)
+
+    Args:
+        project_dir: Root directory of the project
+
+    Returns:
+        The template content as a string, or None if no template is found.
+    """
+    project_dir = Path(project_dir)
+    # Check for single template file
+    single_template = project_dir / ".github" / "PULL_REQUEST_TEMPLATE.md"
+    if single_template.is_file():
+        try:
+            content = single_template.read_text(encoding="utf-8")
+            if content.strip():
+                logger.info(f"Found PR template: {single_template}")
+                return content
+        except Exception as e:
+            logger.warning(f"Failed to read PR template {single_template}: {e}")
+
+    # Check for template directory (pick first .md file alphabetically)
+    template_dir = project_dir / ".github" / "PULL_REQUEST_TEMPLATE"
+    if template_dir.is_dir():
+        try:
+            md_files = sorted(template_dir.glob("*.md"))
+            if md_files:
+                content = md_files[0].read_text(encoding="utf-8")
+                if content.strip():
+                    logger.info(f"Found PR template: {md_files[0]}")
+                    return content
+        except Exception as e:
+            logger.warning(f"Failed to read PR template from {template_dir}: {e}")
+
+    logger.info("No GitHub PR template found in project")
+    return None
+
+
+def _truncate_diff(diff_summary: str) -> str:
+    """
+    Truncate a large diff to file-level summaries to stay within token limits.
+
+    If the diff is within MAX_DIFF_CHARS, return it unchanged.
+    Otherwise, extract only file-level change summaries (e.g. file names
+    with insertions/deletions counts) and discard line-level detail.
+
+    Args:
+        diff_summary: The full diff summary text
+
+    Returns:
+        The original or truncated diff summary.
+    """
+    if len(diff_summary) <= MAX_DIFF_CHARS:
+        return diff_summary
+
+    lines = diff_summary.splitlines()
+    summary_lines: list[str] = []
+    summary_lines.append("(Diff truncated to file-level summaries due to size)")
+    summary_lines.append("")
+
+    for line in lines:
+        # Keep file-level summary lines (stat lines, file headers, etc.)
+        stripped = line.strip()
+        if (
+            stripped.startswith("diff --git")
+            or stripped.startswith("---")
+            or stripped.startswith("+++")
+            or "file changed" in stripped.lower()
+            or "files changed" in stripped.lower()
+            or "insertion" in stripped.lower()
+            or "deletion" in stripped.lower()
+            or stripped.startswith("rename")
+            or stripped.startswith("new file")
+            or stripped.startswith("deleted file")
+            or stripped.startswith("Binary files")
+        ):
+            summary_lines.append(line)
+
+    # If we couldn't extract meaningful summaries, take the first chunk
+    if len(summary_lines) <= 2:
+        truncated = diff_summary[:MAX_DIFF_CHARS]
+        return truncated + "\n\n(... diff truncated due to size)"
+
+    return "\n".join(summary_lines)
+
+
+def _strip_markdown_fences(content: str) -> str:
+    """
+    Strip markdown code fences from the response if present.
+
+    The AI sometimes wraps the output in ```markdown ... ``` even when instructed
+    not to. This ensures the PR body renders correctly on GitHub.
+
+    Args:
+        content: The response content to clean
+
+    Returns:
+        The content with markdown fences stripped.
+    """
+    result = content
+
+    # Strip opening fence (```markdown or just ```)
+    if result.startswith("```markdown"):
+        result = result[len("```markdown") :].lstrip("\n")
+    elif result.startswith("```md"):
+        result = result[len("```md") :].lstrip("\n")
+    elif result.startswith("```"):
+        result = result[3:].lstrip("\n")
+
+    # Strip closing fence
+    if result.endswith("```"):
+        result = result[:-3].rstrip("\n")
+
+    return result.strip()
+
+
+def _build_prompt(
+    template_content: str,
+    diff_summary: str,
+    spec_overview: str,
+    commit_log: str,
+    branch_name: str,
+    target_branch: str,
+) -> str:
+    """
+    Build the prompt for the PR template filler agent.
+
+    Combines the system prompt context variables into a single message
+    that includes the template and all change context.
+
+    Args:
+        template_content: The PR template markdown
+        diff_summary: Git diff summary (possibly truncated)
+        spec_overview: Spec.md content or summary
+        commit_log: Git log of commits in the PR
+        branch_name: Source branch name
+        target_branch: Target branch name
+
+    Returns:
+        The assembled prompt string.
+    """
+    return f"""Fill out the following GitHub PR template using the provided context.
+Return ONLY the filled template markdown — no preamble, no explanation, no code fences.
+
+## Checkbox Guidelines
+
+IMPORTANT: Be accurate and honest about what has and hasn't been verified.
+
+**Check these based on context (you can infer from the diff/spec):**
+- Base Branch targeting — check based on target_branch value
+- Type of Change (bug fix, feature, docs, refactor, test) — infer from diff and spec
+- Area (Frontend, Backend, Fullstack) — infer from changed file paths
+- Feature Toggle "N/A" — if the feature appears complete and not behind a flag
+- Breaking Changes "No" — if changes appear backward compatible
+
+**Leave UNCHECKED (these require human verification you cannot perform):**
+- "I've tested my changes locally" — you have not tested anything
+- "All CI checks pass" — CI has not run yet
+- "Windows/macOS/Linux tested" — requires manual testing on each platform
+- "All existing tests pass" — CI has not run yet
+- "New features include test coverage" — unless test files are clearly visible in the diff
+- "Bug fixes include regression tests" — unless test files are clearly visible in the diff
+
+**For platform/code quality checkboxes:**
+- "Used centralized platform/ module" — leave unchecked unless you can verify from the diff
+- "No hardcoded paths" — leave unchecked unless you can verify from the diff
+- "PR is small and focused (< 400 lines)" — check only if diff stats show < 400 lines changed
+
+**For the "I've synced with develop branch" checkbox:**
+- Leave unchecked — you cannot verify the sync status
+
+## PR Template
+
+{template_content}
+
+## Change Context
+
+### Branch Information
+- **Source branch:** {branch_name}
+- **Target branch:** {target_branch}
+
+### Git Diff Summary
+```
+{diff_summary}
+```
+
+### Spec Overview
+{spec_overview}
+
+### Commit History
+```
+{commit_log}
+```
+
+Fill every section of the PR template. Follow the checkbox guidelines above carefully.
+Output ONLY the completed template — no code fences, no preamble."""
+
+
+def _load_spec_overview(spec_dir: Path) -> str:
+    """
+    Load the spec.md content for context. Falls back to a brief note if unavailable.
+
+    Args:
+        spec_dir: Directory containing the spec files
+
+    Returns:
+        The spec content or a fallback message.
+    """
+    spec_file = spec_dir / "spec.md"
+    if spec_file.is_file():
+        try:
+            content = spec_file.read_text(encoding="utf-8")
+            # Truncate very long specs to keep prompt manageable
+            if len(content) > 8000:
+                return content[:8000] + "\n\n(... spec truncated for brevity)"
+            return content
+        except Exception as e:
+            logger.warning(f"Failed to read spec.md: {e}")
+    return "(No spec overview available)"
+
+
+async def run_pr_template_filler(
+    project_dir: Path,
+    spec_dir: Path,
+    model: str,
+    thinking_budget: int | None = None,
+    branch_name: str = "",
+    target_branch: str = "develop",
+    diff_summary: str = "",
+    commit_log: str = "",
+    verbose: bool = False,
+) -> str | None:
+    """
+    Run the PR template filler agent to generate a filled PR body.
+
+    Detects the project's PR template, gathers change context, and invokes
+    Claude to intelligently fill out the template sections.
+
+    Args:
+        project_dir: Root directory of the project
+        spec_dir: Directory containing the spec files
+        model: Claude model to use
+        thinking_budget: Max thinking tokens (None to disable extended thinking)
+        branch_name: Source branch name for the PR
+        target_branch: Target branch name for the PR
+        diff_summary: Git diff summary of changes
+        commit_log: Git log of commits included in the PR
+        verbose: Whether to show detailed output
+
+    Returns:
+        The filled template markdown string, or None if template detection fails
+        or the agent encounters an error.
+    """
+    # Detect PR template
+    template_content = detect_pr_template(project_dir)
+    if template_content is None:
+        logger.info("No PR template detected — skipping template filler")
+        return None
+
+    # Load spec overview
+    spec_overview = _load_spec_overview(spec_dir)
+
+    # Truncate diff if too large
+    truncated_diff = _truncate_diff(diff_summary)
+
+    # Build the prompt
+    prompt = _build_prompt(
+        template_content=template_content,
+        diff_summary=truncated_diff,
+        spec_overview=spec_overview,
+        commit_log=commit_log,
+        branch_name=branch_name,
+        target_branch=target_branch,
+    )
+
+    # Initialize task logger
+    task_logger = get_task_logger(spec_dir)
+    if task_logger:
+        task_logger.start_phase(LogPhase.CODING, "PR template filling")
+
+    # Create client following the pattern from planner.py
+    client = create_client(
+        project_dir,
+        spec_dir,
+        model,
+        agent_type="pr_template_filler",
+        max_thinking_tokens=thinking_budget,
+    )
+
+    try:
+        async with client:
+            status, response, _ = await run_agent_session(
+                client, prompt, spec_dir, verbose, phase=LogPhase.CODING
+            )
+
+        if task_logger:
+            task_logger.end_phase(
+                LogPhase.CODING,
+                success=(status != "error"),
+                message="PR template filling completed",
+            )
+
+        if status == "error":
+            logger.error("PR template filler agent returned an error")
+            return None
+
+        # The agent should return only the filled template markdown
+        if response and response.strip():
+            result = _strip_markdown_fences(response.strip())
+            logger.info("PR template filled successfully")
+            return result
+
+        logger.warning("PR template filler returned empty response")
+        return None
+
+    except Exception as e:
+        logger.error(f"PR template filler error: {e}")
+        if task_logger:
+            task_logger.log_error(f"PR template filler error: {e}", LogPhase.CODING)
+        return None
@@ -0,0 +1,726 @@
+"""
+Agent Session Management
+========================
+
+Handles running agent sessions and post-session processing including
+memory updates, recovery tracking, and Linear integration.
+"""
+
+import logging
+from pathlib import Path
+
+from claude_agent_sdk import ClaudeSDKClient
+from core.error_utils import (
+    is_authentication_error,
+    is_rate_limit_error,
+    is_tool_concurrency_error,
+)
+from core.file_utils import write_json_atomic
+from debug import debug, debug_detailed, debug_error, debug_section, debug_success
+from insight_extractor import extract_session_insights
+from linear_updater import (
+    linear_subtask_completed,
+    linear_subtask_failed,
+)
+from progress import (
+    count_subtasks_detailed,
+    is_build_complete,
+)
+from recovery import RecoveryManager, check_and_recover, reset_subtask
+from security.tool_input_validator import get_safe_tool_input
+from task_logger import (
+    LogEntryType,
+    LogPhase,
+    get_task_logger,
+)
+from ui import (
+    StatusManager,
+    muted,
+    print_key_value,
+    print_status,
+)
+
+from .base import sanitize_error_message
+from .memory_manager import save_session_memory
+from .utils import (
+    find_subtask_in_plan,
+    get_commit_count,
+    get_latest_commit,
+    load_implementation_plan,
+    sync_spec_to_source,
+)
+
+logger = logging.getLogger(__name__)
+
+
+def _execute_recovery_action(
+    recovery_action,
+    recovery_manager: RecoveryManager,
+    spec_dir: Path,
+    project_dir: Path,
+    subtask_id: str,
+) -> None:
+    """Execute a recovery action (rollback/retry/skip/escalate)."""
+    if not recovery_action:
+        return
+
+    print_status(f"Recovery action: {recovery_action.action}", "info")
+    print_status(f"Reason: {recovery_action.reason}", "info")
+
+    if recovery_action.action == "rollback":
+        print_status(f"Rolling back to {recovery_action.target[:8]}", "warning")
+        if recovery_manager.rollback_to_commit(recovery_action.target):
+            print_status("Rollback successful", "success")
+        else:
+            print_status("Rollback failed", "error")
+
+    elif recovery_action.action == "retry":
+        print_status(f"Resetting subtask {subtask_id} for retry", "info")
+        reset_subtask(spec_dir, project_dir, subtask_id)
+        print_status("Subtask reset - will retry with different approach", "success")
+
+    elif recovery_action.action in ("skip", "escalate"):
+        print_status(f"Marking subtask {subtask_id} as stuck", "warning")
+        recovery_manager.mark_subtask_stuck(subtask_id, recovery_action.reason)
+        print_status("Subtask marked for human intervention", "warning")
+
+
+async def post_session_processing(
+    spec_dir: Path,
+    project_dir: Path,
+    subtask_id: str,
+    session_num: int,
+    commit_before: str | None,
+    commit_count_before: int,
+    recovery_manager: RecoveryManager,
+    linear_enabled: bool = False,
+    status_manager: StatusManager | None = None,
+    source_spec_dir: Path | None = None,
+    error_info: dict | None = None,
+) -> bool:
+    """
+    Process session results and update memory automatically.
+
+    This runs in Python (100% reliable) instead of relying on agent compliance.
+
+    Args:
+        spec_dir: Spec directory containing memory/
+        project_dir: Project root for git operations
+        subtask_id: The subtask that was being worked on
+        session_num: Current session number
+        commit_before: Git commit hash before session
+        commit_count_before: Number of commits before session
+        recovery_manager: Recovery manager instance
+        linear_enabled: Whether Linear integration is enabled
+        status_manager: Optional status manager for ccstatusline
+        source_spec_dir: Original spec directory (for syncing back from worktree)
+        error_info: Error information from run_agent_session (for rate limit detection)
+
+    Returns:
+        True if subtask was completed successfully
+    """
+    print()
+    print(muted("--- Post-Session Processing ---"))
+
+    # Sync implementation plan back to source (for worktree mode)
+    if sync_spec_to_source(spec_dir, source_spec_dir):
+        print_status("Implementation plan synced to main project", "success")
+
+    # Check if implementation plan was updated
+    plan = load_implementation_plan(spec_dir)
+    if not plan:
+        print("  Warning: Could not load implementation plan")
+        return False
+
+    subtask = find_subtask_in_plan(plan, subtask_id)
+    if not subtask:
+        print(f"  Warning: Subtask {subtask_id} not found in plan")
+        return False
+
+    subtask_status = subtask.get("status", "pending")
+
+    # Check for new commits
+    commit_after = get_latest_commit(project_dir)
+    commit_count_after = get_commit_count(project_dir)
+    new_commits = commit_count_after - commit_count_before
+
+    print_key_value("Subtask status", subtask_status)
+    print_key_value("New commits", str(new_commits))
+
+    if subtask_status == "completed":
+        # Success! Record the attempt and good commit
+        print_status(f"Subtask {subtask_id} completed successfully", "success")
+
+        # Update status file
+        if status_manager:
+            subtasks = count_subtasks_detailed(spec_dir)
+            status_manager.update_subtasks(
+                completed=subtasks["completed"],
+                total=subtasks["total"],
+                in_progress=0,
+            )
+
+        # Record successful attempt
+        recovery_manager.record_attempt(
+            subtask_id=subtask_id,
+            session=session_num,
+            success=True,
+            approach=f"Implemented: {subtask.get('description', 'subtask')[:100]}",
+        )
+
+        # Record good commit for rollback safety
+        if commit_after and commit_after != commit_before:
+            recovery_manager.record_good_commit(commit_after, subtask_id)
+            print_status(f"Recorded good commit: {commit_after[:8]}", "success")
+
+        # Record Linear session result (if enabled)
+        if linear_enabled:
+            # Get progress counts for the comment
+            subtasks_detail = count_subtasks_detailed(spec_dir)
+            await linear_subtask_completed(
+                spec_dir=spec_dir,
+                subtask_id=subtask_id,
+                completed_count=subtasks_detail["completed"],
+                total_count=subtasks_detail["total"],
+            )
+            print_status("Linear progress recorded", "success")
+
+        # Extract rich insights from session (LLM-powered analysis)
+        try:
+            extracted_insights = await extract_session_insights(
+                spec_dir=spec_dir,
+                project_dir=project_dir,
+                subtask_id=subtask_id,
+                session_num=session_num,
+                commit_before=commit_before,
+                commit_after=commit_after,
+                success=True,
+                recovery_manager=recovery_manager,
+            )
+            insight_count = len(extracted_insights.get("file_insights", []))
+            pattern_count = len(extracted_insights.get("patterns_discovered", []))
+            if insight_count > 0 or pattern_count > 0:
+                print_status(
+                    f"Extracted {insight_count} file insights, {pattern_count} patterns",
+                    "success",
+                )
+        except Exception as e:
+            logger.warning(f"Insight extraction failed: {e}")
+            extracted_insights = None
+
+        # Save session memory (Graphiti=primary, file-based=fallback)
+        try:
+            save_success, storage_type = await save_session_memory(
+                spec_dir=spec_dir,
+                project_dir=project_dir,
+                subtask_id=subtask_id,
+                session_num=session_num,
+                success=True,
+                subtasks_completed=[subtask_id],
+                discoveries=extracted_insights,
+            )
+            if save_success:
+                if storage_type == "graphiti":
+                    print_status("Session saved to Graphiti memory", "success")
+                else:
+                    print_status(
+                        "Session saved to file-based memory (fallback)", "info"
+                    )
+            else:
+                print_status("Failed to save session memory", "warning")
+        except Exception as e:
+            logger.warning(f"Error saving session memory: {e}")
+            print_status("Memory save failed", "warning")
+
+        return True
+
+    elif subtask_status == "in_progress":
+        # Session ended without completion
+        print_status(f"Subtask {subtask_id} still in progress", "warning")
+
+        recovery_manager.record_attempt(
+            subtask_id=subtask_id,
+            session=session_num,
+            success=False,
+            approach="Session ended with subtask in_progress",
+            error="Subtask not marked as completed",
+        )
+
+        # Check if this was a concurrency error - if so, reset subtask to pending for retry
+        is_concurrency_error = (
+            error_info and error_info.get("type") == "tool_concurrency"
+        )
+
+        if is_concurrency_error:
+            print_status(
+                f"Rate limit detected - resetting subtask {subtask_id} to pending for retry",
+                "info",
+            )
+
+            # Use recovery system's reset_subtask for consistency
+            reset_subtask(spec_dir, project_dir, subtask_id)
+
+            # Also reset in implementation plan
+            plan = load_implementation_plan(spec_dir)
+            if plan:
+                # Find and reset the subtask
+                subtask_found = False
+                for phase in plan.get("phases", []):
+                    for subtask in phase.get("subtasks", []):
+                        if subtask.get("id") == subtask_id:
+                            # Reset subtask to pending state
+                            subtask["status"] = "pending"
+                            subtask["started_at"] = None
+                            subtask["completed_at"] = None
+                            subtask_found = True
+                            break
+                    if subtask_found:
+                        break
+
+                if subtask_found:
+                    # Save plan atomically to prevent corruption
+                    try:
+                        plan_path = spec_dir / "implementation_plan.json"
+                        write_json_atomic(plan_path, plan, indent=2)
+                        print_status(
+                            f"Subtask {subtask_id} reset to pending status", "success"
+                        )
+                    except Exception as e:
+                        logger.error(
+                            f"Failed to save implementation plan after reset: {e}"
+                        )
+                        print_status("Failed to save plan after reset", "error")
+                else:
+                    print_status(
+                        f"Warning: Could not find subtask {subtask_id} in plan",
+                        "warning",
+                    )
+            else:
+                print_status(
+                    "Warning: Could not load implementation plan for reset", "warning"
+                )
+        else:
+            # Non-rate-limit error - use automatic recovery flow
+            error_message = (
+                error_info.get("message", "Subtask not marked as completed")
+                if error_info
+                else "Subtask not marked as completed"
+            )
+
+            recovery_action = check_and_recover(
+                spec_dir=spec_dir,
+                project_dir=project_dir,
+                subtask_id=subtask_id,
+                error=error_message,
+            )
+            _execute_recovery_action(
+                recovery_action, recovery_manager, spec_dir, project_dir, subtask_id
+            )
+
+        # Still record commit if one was made (partial progress)
+        if commit_after and commit_after != commit_before:
+            recovery_manager.record_good_commit(commit_after, subtask_id)
+            print_status(
+                f"Recorded partial progress commit: {commit_after[:8]}", "info"
+            )
+
+        # Record Linear session result (if enabled)
+        if linear_enabled:
+            attempt_count = recovery_manager.get_attempt_count(subtask_id)
+            await linear_subtask_failed(
+                spec_dir=spec_dir,
+                subtask_id=subtask_id,
+                attempt=attempt_count,
+                error_summary="Session ended without completion",
+            )
+
+        # Extract insights even from failed sessions (valuable for future attempts)
+        try:
+            extracted_insights = await extract_session_insights(
+                spec_dir=spec_dir,
+                project_dir=project_dir,
+                subtask_id=subtask_id,
+                session_num=session_num,
+                commit_before=commit_before,
+                commit_after=commit_after,
+                success=False,
+                recovery_manager=recovery_manager,
+            )
+        except Exception as e:
+            logger.debug(f"Insight extraction failed for incomplete session: {e}")
+            extracted_insights = None
+
+        # Save failed session memory (to track what didn't work)
+        try:
+            await save_session_memory(
+                spec_dir=spec_dir,
+                project_dir=project_dir,
+                subtask_id=subtask_id,
+                session_num=session_num,
+                success=False,
+                subtasks_completed=[],
+                discoveries=extracted_insights,
+            )
+        except Exception as e:
+            logger.debug(f"Failed to save incomplete session memory: {e}")
+
+        return False
+
+    else:
+        # Subtask still pending or failed
+        print_status(
+            f"Subtask {subtask_id} not completed (status: {subtask_status})", "error"
+        )
+
+        recovery_manager.record_attempt(
+            subtask_id=subtask_id,
+            session=session_num,
+            success=False,
+            approach="Session ended without progress",
+            error=f"Subtask status is {subtask_status}",
+        )
+
+        # Automatic recovery flow - determine and execute recovery action
+        error_message = f"Subtask status is {subtask_status}"
+        if error_info:
+            error_message = error_info.get("message", error_message)
+
+        recovery_action = check_and_recover(
+            spec_dir=spec_dir,
+            project_dir=project_dir,
+            subtask_id=subtask_id,
+            error=error_message,
+        )
+        _execute_recovery_action(
+            recovery_action, recovery_manager, spec_dir, project_dir, subtask_id
+        )
+
+        # Record Linear session result (if enabled)
+        if linear_enabled:
+            attempt_count = recovery_manager.get_attempt_count(subtask_id)
+            await linear_subtask_failed(
+                spec_dir=spec_dir,
+                subtask_id=subtask_id,
+                attempt=attempt_count,
+                error_summary=f"Subtask status: {subtask_status}",
+            )
+
+        # Extract insights even from completely failed sessions
+        try:
+            extracted_insights = await extract_session_insights(
+                spec_dir=spec_dir,
+                project_dir=project_dir,
+                subtask_id=subtask_id,
+                session_num=session_num,
+                commit_before=commit_before,
+                commit_after=commit_after,
+                success=False,
+                recovery_manager=recovery_manager,
+            )
+        except Exception as e:
+            logger.debug(f"Insight extraction failed for failed session: {e}")
+            extracted_insights = None
+
+        # Save failed session memory (to track what didn't work)
+        try:
+            await save_session_memory(
+                spec_dir=spec_dir,
+                project_dir=project_dir,
+                subtask_id=subtask_id,
+                session_num=session_num,
+                success=False,
+                subtasks_completed=[],
+                discoveries=extracted_insights,
+            )
+        except Exception as e:
+            logger.debug(f"Failed to save failed session memory: {e}")
+
+        return False
+
+
+async def run_agent_session(
+    client: ClaudeSDKClient,
+    message: str,
+    spec_dir: Path,
+    verbose: bool = False,
+    phase: LogPhase = LogPhase.CODING,
+) -> tuple[str, str, dict]:
+    """
+    Run a single agent session using Claude Agent SDK.
+
+    Args:
+        client: Claude SDK client
+        message: The prompt to send
+        spec_dir: Spec directory path
+        verbose: Whether to show detailed output
+        phase: Current execution phase for logging
+
+    Returns:
+        (status, response_text, error_info) where:
+        - status: "continue", "complete", or "error"
+        - response_text: Agent's response text
+        - error_info: Dict with error details (empty if no error):
+            - "type": "tool_concurrency" or "other"
+            - "message": Error message string
+            - "exception_type": Exception class name string
+    """
+    debug_section("session", f"Agent Session - {phase.value}")
+    debug(
+        "session",
+        "Starting agent session",
+        spec_dir=str(spec_dir),
+        phase=phase.value,
+        prompt_length=len(message),
+        prompt_preview=message[:200] + "..." if len(message) > 200 else message,
+    )
+    print("Sending prompt to Claude Agent SDK...\n")
+
+    # Get task logger for this spec
+    task_logger = get_task_logger(spec_dir)
+    current_tool = None
+    message_count = 0
+    tool_count = 0
+
+    try:
+        # Send the query
+        debug("session", "Sending query to Claude SDK...")
+        await client.query(message)
+        debug_success("session", "Query sent successfully")
+
+        # Collect response text and show tool use
+        response_text = ""
+        debug("session", "Starting to receive response stream...")
+        async for msg in client.receive_response():
+            msg_type = type(msg).__name__
+            message_count += 1
+            debug_detailed(
+                "session",
+                f"Received message #{message_count}",
+                msg_type=msg_type,
+            )
+
+            # Handle AssistantMessage (text and tool use)
+            if msg_type == "AssistantMessage" and hasattr(msg, "content"):
+                for block in msg.content:
+                    block_type = type(block).__name__
+
+                    if block_type == "TextBlock" and hasattr(block, "text"):
+                        response_text += block.text
+                        print(block.text, end="", flush=True)
+                        # Log text to task logger (persist without double-printing)
+                        if task_logger and block.text.strip():
+                            task_logger.log(
+                                block.text,
+                                LogEntryType.TEXT,
+                                phase,
+                                print_to_console=False,
+                            )
+                    elif block_type == "ToolUseBlock" and hasattr(block, "name"):
+                        tool_name = block.name
+                        tool_input_display = None
+                        tool_count += 1
+
+                        # Safely extract tool input (handles None, non-dict, etc.)
+                        inp = get_safe_tool_input(block)
+
+                        # Extract meaningful tool input for display
+                        if inp:
+                            if "pattern" in inp:
+                                tool_input_display = f"pattern: {inp['pattern']}"
+                            elif "file_path" in inp:
+                                fp = inp["file_path"]
+                                if len(fp) > 50:
+                                    fp = "..." + fp[-47:]
+                                tool_input_display = fp
+                            elif "command" in inp:
+                                cmd = inp["command"]
+                                if len(cmd) > 50:
+                                    cmd = cmd[:47] + "..."
+                                tool_input_display = cmd
+                            elif "path" in inp:
+                                tool_input_display = inp["path"]
+
+                        debug(
+                            "session",
+                            f"Tool call #{tool_count}: {tool_name}",
+                            tool_input=tool_input_display,
+                            full_input=str(inp)[:500] if inp else None,
+                        )
+
+                        # Log tool start (handles printing too)
+                        if task_logger:
+                            task_logger.tool_start(
+                                tool_name,
+                                tool_input_display,
+                                phase,
+                                print_to_console=True,
+                            )
+                        else:
+                            print(f"\n[Tool: {tool_name}]", flush=True)
+
+                        if verbose and hasattr(block, "input"):
+                            input_str = str(block.input)
+                            if len(input_str) > 300:
+                                print(f"   Input: {input_str[:300]}...", flush=True)
+                            else:
+                                print(f"   Input: {input_str}", flush=True)
+                        current_tool = tool_name
+
+            # Handle UserMessage (tool results)
+            elif msg_type == "UserMessage" and hasattr(msg, "content"):
+                for block in msg.content:
+                    block_type = type(block).__name__
+
+                    if block_type == "ToolResultBlock":
+                        result_content = getattr(block, "content", "")
+                        is_error = getattr(block, "is_error", False)
+
+                        # Check if this is an error (not just content containing "blocked")
+                        if is_error and "blocked" in str(result_content).lower():
+                            # Actual blocked command by security hook
+                            debug_error(
+                                "session",
+                                f"Tool BLOCKED: {current_tool}",
+                                result=str(result_content)[:300],
+                            )
+                            print(f"   [BLOCKED] {result_content}", flush=True)
+                            if task_logger and current_tool:
+                                task_logger.tool_end(
+                                    current_tool,
+                                    success=False,
+                                    result="BLOCKED",
+                                    detail=str(result_content),
+                                    phase=phase,
+                                )
+                        elif is_error:
+                            # Show errors (truncated)
+                            error_str = str(result_content)[:500]
+                            debug_error(
+                                "session",
+                                f"Tool error: {current_tool}",
+                                error=error_str[:200],
+                            )
+                            print(f"   [Error] {error_str}", flush=True)
+                            if task_logger and current_tool:
+                                # Store full error in detail for expandable view
+                                task_logger.tool_end(
+                                    current_tool,
+                                    success=False,
+                                    result=error_str[:100],
+                                    detail=str(result_content),
+                                    phase=phase,
+                                )
+                        else:
+                            # Tool succeeded
+                            debug_detailed(
+                                "session",
+                                f"Tool success: {current_tool}",
+                                result_length=len(str(result_content)),
+                            )
+                            if verbose:
+                                result_str = str(result_content)[:200]
+                                print(f"   [Done] {result_str}", flush=True)
+                            else:
+                                print("   [Done]", flush=True)
+                            if task_logger and current_tool:
+                                # Store full result in detail for expandable view (only for certain tools)
+                                # Skip storing for very large outputs like Glob results
+                                detail_content = None
+                                if current_tool in (
+                                    "Read",
+                                    "Grep",
+                                    "Bash",
+                                    "Edit",
+                                    "Write",
+                                ):
+                                    result_str = str(result_content)
+                                    # Only store if not too large (detail truncation happens in logger)
+                                    if (
+                                        len(result_str) < 50000
+                                    ):  # 50KB max before truncation
+                                        detail_content = result_str
+                                task_logger.tool_end(
+                                    current_tool,
+                                    success=True,
+                                    detail=detail_content,
+                                    phase=phase,
+                                )
+
+                        current_tool = None
+
+        print("\n" + "-" * 70 + "\n")
+
+        # Check if build is complete
+        if is_build_complete(spec_dir):
+            debug_success(
+                "session",
+                "Session completed - build is complete",
+                message_count=message_count,
+                tool_count=tool_count,
+                response_length=len(response_text),
+            )
+            return "complete", response_text, {}
+
+        debug_success(
+            "session",
+            "Session completed - continuing",
+            message_count=message_count,
+            tool_count=tool_count,
+            response_length=len(response_text),
+        )
+        return "continue", response_text, {}
+
+    except Exception as e:
+        # Detect specific error types for better retry handling
+        is_concurrency = is_tool_concurrency_error(e)
+        is_rate_limit = is_rate_limit_error(e)
+        is_auth = is_authentication_error(e)
+
+        # Classify error type for appropriate handling
+        if is_concurrency:
+            error_type = "tool_concurrency"
+        elif is_rate_limit:
+            error_type = "rate_limit"
+        elif is_auth:
+            error_type = "authentication"
+        else:
+            error_type = "other"
+
+        debug_error(
+            "session",
+            f"Session error: {e}",
+            exception_type=type(e).__name__,
+            error_category=error_type,
+            message_count=message_count,
+            tool_count=tool_count,
+        )
+
+        # Sanitize error message to remove potentially sensitive data
+        # Must happen BEFORE printing to stdout, since stdout is captured by the frontend
+        sanitized_error = sanitize_error_message(str(e))
+
+        # Log errors prominently based on type
+        if is_concurrency:
+            print("\n⚠️  Tool concurrency limit reached (400 error)")
+            print("   Claude API limits concurrent tool use in a single request")
+            print(f"   Error: {sanitized_error[:200]}\n")
+        elif is_rate_limit:
+            print("\n⚠️  Rate limit reached")
+            print("   API usage quota exceeded - waiting for reset")
+            print(f"   Error: {sanitized_error[:200]}\n")
+        elif is_auth:
+            print("\n⚠️  Authentication error")
+            print("   OAuth token may be invalid or expired")
+            print(f"   Error: {sanitized_error[:200]}\n")
+        else:
+            print(f"Error during agent session: {sanitized_error}")
+
+        if task_logger:
+            task_logger.log_error(f"Session error: {sanitized_error}", phase)
+
+        error_info = {
+            "type": error_type,
+            "message": sanitized_error,
+            "exception_type": type(e).__name__,
+        }
+        return "error", sanitized_error, error_info
@@ -0,0 +1,91 @@
+"""
+Custom MCP Tools for Auto-Claude Agents
+========================================
+
+This module provides custom MCP tools that agents can use for reliable
+operations on auto-claude data structures. These tools replace prompt-based
+JSON manipulation with guaranteed-correct operations.
+
+Benefits:
+- 100% reliable JSON operations (no malformed output)
+- Reduced context usage (tool definitions << prompt instructions)
+- Type-safe with proper error handling
+- Each agent only sees tools relevant to their role via allowed_tools
+
+Usage:
+    from auto_claude_tools import create_auto_claude_mcp_server, get_allowed_tools
+
+    # Create the MCP server
+    mcp_server = create_auto_claude_mcp_server(spec_dir, project_dir)
+
+    # Get allowed tools for a specific agent type
+    allowed_tools = get_allowed_tools("coder")
+
+    # Use in ClaudeAgentOptions
+    options = ClaudeAgentOptions(
+        mcp_servers={"auto-claude": mcp_server},
+        allowed_tools=allowed_tools,
+        ...
+    )
+"""
+
+from .models import (
+    # Agent configuration registry
+    AGENT_CONFIGS,
+    # Base tools
+    BASE_READ_TOOLS,
+    BASE_WRITE_TOOLS,
+    # MCP tool lists
+    CONTEXT7_TOOLS,
+    ELECTRON_TOOLS,
+    GRAPHITI_MCP_TOOLS,
+    LINEAR_TOOLS,
+    PUPPETEER_TOOLS,
+    # Auto-Claude tool names
+    TOOL_GET_BUILD_PROGRESS,
+    TOOL_GET_SESSION_CONTEXT,
+    TOOL_RECORD_DISCOVERY,
+    TOOL_RECORD_GOTCHA,
+    TOOL_UPDATE_QA_STATUS,
+    TOOL_UPDATE_SUBTASK_STATUS,
+    WEB_TOOLS,
+    # Config functions
+    get_agent_config,
+    get_default_thinking_level,
+    get_required_mcp_servers,
+    is_electron_mcp_enabled,
+)
+from .permissions import get_all_agent_types, get_allowed_tools
+from .registry import create_auto_claude_mcp_server, is_tools_available
+
+__all__ = [
+    # Main API
+    "create_auto_claude_mcp_server",
+    "get_allowed_tools",
+    "is_tools_available",
+    # Agent configuration registry
+    "AGENT_CONFIGS",
+    "get_agent_config",
+    "get_required_mcp_servers",
+    "get_default_thinking_level",
+    "get_all_agent_types",
+    # Base tool lists
+    "BASE_READ_TOOLS",
+    "BASE_WRITE_TOOLS",
+    "WEB_TOOLS",
+    # MCP tool lists
+    "CONTEXT7_TOOLS",
+    "LINEAR_TOOLS",
+    "GRAPHITI_MCP_TOOLS",
+    "ELECTRON_TOOLS",
+    "PUPPETEER_TOOLS",
+    # Auto-Claude tool name constants
+    "TOOL_UPDATE_SUBTASK_STATUS",
+    "TOOL_GET_BUILD_PROGRESS",
+    "TOOL_RECORD_DISCOVERY",
+    "TOOL_RECORD_GOTCHA",
+    "TOOL_GET_SESSION_CONTEXT",
+    "TOOL_UPDATE_QA_STATUS",
+    # Config
+    "is_electron_mcp_enabled",
+]
@@ -0,0 +1,545 @@
+"""
+Tool Models and Constants
+==========================
+
+Defines tool name constants and configuration for auto-claude MCP tools.
+
+This module is the single source of truth for all tool definitions used by
+the Claude Agent SDK client. Tool lists are organized by category:
+
+- Base tools: Core file operations (Read, Write, Edit, etc.)
+- Web tools: Documentation and research (WebFetch, WebSearch)
+- MCP tools: External integrations (Context7, Linear, Graphiti, etc.)
+- Auto-Claude tools: Custom build management tools
+"""
+
+import os
+
+# =============================================================================
+# Base Tools (Built-in Claude Code tools)
+# =============================================================================
+
+# Core file operation tools
+BASE_READ_TOOLS = ["Read", "Glob", "Grep"]
+BASE_WRITE_TOOLS = ["Write", "Edit", "Bash"]
+
+# Web tools for documentation lookup and research
+# Always available to all agents for accessing external information
+WEB_TOOLS = ["WebFetch", "WebSearch"]
+
+# =============================================================================
+# Auto-Claude MCP Tools (Custom build management)
+# =============================================================================
+
+# Auto-Claude MCP tool names (prefixed with mcp__auto-claude__)
+TOOL_UPDATE_SUBTASK_STATUS = "mcp__auto-claude__update_subtask_status"
+TOOL_GET_BUILD_PROGRESS = "mcp__auto-claude__get_build_progress"
+TOOL_RECORD_DISCOVERY = "mcp__auto-claude__record_discovery"
+TOOL_RECORD_GOTCHA = "mcp__auto-claude__record_gotcha"
+TOOL_GET_SESSION_CONTEXT = "mcp__auto-claude__get_session_context"
+TOOL_UPDATE_QA_STATUS = "mcp__auto-claude__update_qa_status"
+
+# =============================================================================
+# External MCP Tools
+# =============================================================================
+
+# Context7 MCP tools for documentation lookup (always enabled)
+CONTEXT7_TOOLS = [
+    "mcp__context7__resolve-library-id",
+    "mcp__context7__query-docs",
+]
+
+# Linear MCP tools for project management (when LINEAR_API_KEY is set)
+LINEAR_TOOLS = [
+    "mcp__linear-server__list_teams",
+    "mcp__linear-server__get_team",
+    "mcp__linear-server__list_projects",
+    "mcp__linear-server__get_project",
+    "mcp__linear-server__create_project",
+    "mcp__linear-server__update_project",
+    "mcp__linear-server__list_issues",
+    "mcp__linear-server__get_issue",
+    "mcp__linear-server__create_issue",
+    "mcp__linear-server__update_issue",
+    "mcp__linear-server__list_comments",
+    "mcp__linear-server__create_comment",
+    "mcp__linear-server__list_issue_statuses",
+    "mcp__linear-server__list_issue_labels",
+    "mcp__linear-server__list_users",
+    "mcp__linear-server__get_user",
+]
+
+# Graphiti MCP tools for knowledge graph memory (when GRAPHITI_MCP_URL is set)
+# See: https://github.com/getzep/graphiti
+GRAPHITI_MCP_TOOLS = [
+    "mcp__graphiti-memory__search_nodes",  # Search entity summaries
+    "mcp__graphiti-memory__search_facts",  # Search relationships between entities
+    "mcp__graphiti-memory__add_episode",  # Add data to knowledge graph
+    "mcp__graphiti-memory__get_episodes",  # Retrieve recent episodes
+    "mcp__graphiti-memory__get_entity_edge",  # Get specific entity/relationship
+]
+
+# =============================================================================
+# Browser Automation MCP Tools (QA agents only)
+# =============================================================================
+
+# Puppeteer MCP tools for web browser automation
+# Used for web frontend validation (non-Electron web apps)
+# NOTE: Screenshots must be compressed (1280x720, quality 60, JPEG) to stay under
+# Claude SDK's 1MB JSON message buffer limit. See GitHub issue #74.
+PUPPETEER_TOOLS = [
+    "mcp__puppeteer__puppeteer_connect_active_tab",
+    "mcp__puppeteer__puppeteer_navigate",
+    "mcp__puppeteer__puppeteer_screenshot",
+    "mcp__puppeteer__puppeteer_click",
+    "mcp__puppeteer__puppeteer_fill",
+    "mcp__puppeteer__puppeteer_select",
+    "mcp__puppeteer__puppeteer_hover",
+    "mcp__puppeteer__puppeteer_evaluate",
+]
+
+# Electron MCP tools for desktop app automation (when ELECTRON_MCP_ENABLED is set)
+# Uses electron-mcp-server to connect to Electron apps via Chrome DevTools Protocol.
+# Electron app must be started with --remote-debugging-port=9222 (or ELECTRON_DEBUG_PORT).
+# These tools are only available to QA agents (qa_reviewer, qa_fixer), not Coder/Planner.
+# NOTE: Screenshots must be compressed to stay under Claude SDK's 1MB JSON message buffer limit.
+ELECTRON_TOOLS = [
+    "mcp__electron__get_electron_window_info",  # Get info about running Electron windows
+    "mcp__electron__take_screenshot",  # Capture screenshot of Electron window
+    "mcp__electron__send_command_to_electron",  # Send commands (click, fill, evaluate JS)
+    "mcp__electron__read_electron_logs",  # Read console logs from Electron app
+]
+
+# =============================================================================
+# Configuration
+# =============================================================================
+
+
+def is_electron_mcp_enabled() -> bool:
+    """
+    Check if Electron MCP server integration is enabled.
+
+    Requires ELECTRON_MCP_ENABLED to be set to 'true'.
+    When enabled, QA agents can use Electron MCP tools to connect to Electron apps
+    via Chrome DevTools Protocol on the configured debug port.
+    """
+    return os.environ.get("ELECTRON_MCP_ENABLED", "").lower() == "true"
+
+
+# =============================================================================
+# Agent Configuration Registry
+# =============================================================================
+# Single source of truth for phase → tools → MCP servers mapping.
+# This enables phase-aware tool control and context window optimization.
+
+AGENT_CONFIGS = {
+    # ═══════════════════════════════════════════════════════════════════════
+    # SPEC CREATION PHASES (Minimal tools, fast startup)
+    # ═══════════════════════════════════════════════════════════════════════
+    "spec_gatherer": {
+        "tools": BASE_READ_TOOLS + WEB_TOOLS,
+        "mcp_servers": [],  # No MCP needed - just reads project
+        "auto_claude_tools": [],
+        "thinking_default": "medium",
+    },
+    "spec_researcher": {
+        "tools": BASE_READ_TOOLS + WEB_TOOLS,
+        "mcp_servers": ["context7"],  # Needs docs lookup
+        "auto_claude_tools": [],
+        "thinking_default": "medium",
+    },
+    "spec_writer": {
+        "tools": BASE_READ_TOOLS + BASE_WRITE_TOOLS,
+        "mcp_servers": [],  # Just writes spec.md
+        "auto_claude_tools": [],
+        "thinking_default": "high",
+    },
+    "spec_critic": {
+        "tools": BASE_READ_TOOLS,
+        "mcp_servers": [],  # Self-critique, no external tools
+        "auto_claude_tools": [],
+        "thinking_default": "high",
+    },
+    "spec_discovery": {
+        "tools": BASE_READ_TOOLS + WEB_TOOLS,
+        "mcp_servers": [],
+        "auto_claude_tools": [],
+        "thinking_default": "medium",
+    },
+    "spec_context": {
+        "tools": BASE_READ_TOOLS,
+        "mcp_servers": [],
+        "auto_claude_tools": [],
+        "thinking_default": "medium",
+    },
+    "spec_validation": {
+        "tools": BASE_READ_TOOLS,
+        "mcp_servers": [],
+        "auto_claude_tools": [],
+        "thinking_default": "high",
+    },
+    "spec_compaction": {
+        "tools": BASE_READ_TOOLS + BASE_WRITE_TOOLS,
+        "mcp_servers": [],
+        "auto_claude_tools": [],
+        "thinking_default": "medium",
+    },
+    # ═══════════════════════════════════════════════════════════════════════
+    # BUILD PHASES (Full tools + Graphiti memory)
+    # Note: "linear" is conditional on project setting "update_linear_with_tasks"
+    # ═══════════════════════════════════════════════════════════════════════
+    "planner": {
+        "tools": BASE_READ_TOOLS + BASE_WRITE_TOOLS + WEB_TOOLS,
+        "mcp_servers": ["context7", "graphiti", "auto-claude"],
+        "mcp_servers_optional": ["linear"],  # Only if project setting enabled
+        "auto_claude_tools": [
+            TOOL_GET_BUILD_PROGRESS,
+            TOOL_GET_SESSION_CONTEXT,
+            TOOL_RECORD_DISCOVERY,
+        ],
+        "thinking_default": "high",
+    },
+    "coder": {
+        "tools": BASE_READ_TOOLS + BASE_WRITE_TOOLS + WEB_TOOLS,
+        "mcp_servers": ["context7", "graphiti", "auto-claude"],
+        "mcp_servers_optional": ["linear"],
+        "auto_claude_tools": [
+            TOOL_UPDATE_SUBTASK_STATUS,
+            TOOL_GET_BUILD_PROGRESS,
+            TOOL_RECORD_DISCOVERY,
+            TOOL_RECORD_GOTCHA,
+            TOOL_GET_SESSION_CONTEXT,
+        ],
+        "thinking_default": "low",  # Coding uses minimal thinking (effort: low for Opus, 1024 tokens for Sonnet/Haiku)
+    },
+    # ═══════════════════════════════════════════════════════════════════════
+    # QA PHASES (Read + test + browser + Graphiti memory)
+    # ═══════════════════════════════════════════════════════════════════════
+    "qa_reviewer": {
+        # Read + Write/Edit (for QA reports and plan updates) + Bash (for tests)
+        # Note: Reviewer writes to spec directory only (qa_report.md, implementation_plan.json)
+        "tools": BASE_READ_TOOLS + BASE_WRITE_TOOLS + WEB_TOOLS,
+        "mcp_servers": ["context7", "graphiti", "auto-claude", "browser"],
+        "mcp_servers_optional": ["linear"],  # For updating issue status
+        "auto_claude_tools": [
+            TOOL_GET_BUILD_PROGRESS,
+            TOOL_UPDATE_QA_STATUS,
+            TOOL_GET_SESSION_CONTEXT,
+        ],
+        "thinking_default": "high",
+    },
+    "qa_fixer": {
+        "tools": BASE_READ_TOOLS + BASE_WRITE_TOOLS + WEB_TOOLS,
+        "mcp_servers": ["context7", "graphiti", "auto-claude", "browser"],
+        "mcp_servers_optional": ["linear"],
+        "auto_claude_tools": [
+            TOOL_UPDATE_SUBTASK_STATUS,
+            TOOL_GET_BUILD_PROGRESS,
+            TOOL_UPDATE_QA_STATUS,
+            TOOL_RECORD_GOTCHA,
+        ],
+        "thinking_default": "medium",
+    },
+    # ═══════════════════════════════════════════════════════════════════════
+    # UTILITY PHASES (Minimal, no MCP)
+    # ═══════════════════════════════════════════════════════════════════════
+    "insights": {
+        "tools": BASE_READ_TOOLS + WEB_TOOLS,
+        "mcp_servers": [],
+        "auto_claude_tools": [],
+        # Note: Default to "low" for minimal thinking overhead
+        # Haiku doesn't support thinking; create_simple_client() handles this
+        "thinking_default": "low",
+    },
+    "merge_resolver": {
+        "tools": [],  # Text-only analysis
+        "mcp_servers": [],
+        "auto_claude_tools": [],
+        "thinking_default": "low",
+    },
+    "commit_message": {
+        "tools": [],
+        "mcp_servers": [],
+        "auto_claude_tools": [],
+        "thinking_default": "low",
+    },
+    "pr_template_filler": {
+        "tools": BASE_READ_TOOLS,  # Read-only — reads diff, template, spec
+        "mcp_servers": [],  # No MCP needed, context passed via prompt
+        "auto_claude_tools": [],
+        "thinking_default": "low",  # Fast utility task for structured fill-in
+    },
+    "pr_reviewer": {
+        "tools": BASE_READ_TOOLS + WEB_TOOLS,  # Read-only
+        "mcp_servers": ["context7"],
+        "auto_claude_tools": [],
+        "thinking_default": "high",
+    },
+    "pr_orchestrator_parallel": {
+        # Read-only for parallel PR orchestrator
+        # NOTE: Do NOT add "Task" here - the SDK auto-allows Task when agents are defined
+        # via the --agents flag. Explicitly adding it interferes with agent registration.
+        "tools": BASE_READ_TOOLS + WEB_TOOLS,
+        "mcp_servers": ["context7"],
+        "auto_claude_tools": [],
+        "thinking_default": "high",
+    },
+    "pr_followup_parallel": {
+        # Read-only for parallel followup reviewer
+        # NOTE: Do NOT add "Task" here - same reason as pr_orchestrator_parallel
+        "tools": BASE_READ_TOOLS + WEB_TOOLS,
+        "mcp_servers": ["context7"],
+        "auto_claude_tools": [],
+        "thinking_default": "high",
+    },
+    "pr_followup_extraction": {
+        # Lightweight extraction call for recovering data when structured output fails
+        # Pure structured output extraction, no tools needed
+        "tools": [],
+        "mcp_servers": [],
+        "auto_claude_tools": [],
+        "thinking_default": "low",
+    },
+    "pr_finding_validator": {
+        # Standalone validator for re-checking findings against actual code
+        # Called separately from orchestrator to validate findings with fresh context
+        "tools": BASE_READ_TOOLS,
+        "mcp_servers": [],
+        "auto_claude_tools": [],
+        "thinking_default": "medium",
+    },
+    "investigation_specialist": {
+        # Read-only specialist for issue investigation (root cause, impact, fix, reproduction)
+        "tools": BASE_READ_TOOLS,
+        "mcp_servers": [],
+        "auto_claude_tools": [],
+        "thinking_default": "medium",
+    },
+    # ═══════════════════════════════════════════════════════════════════════
+    # ANALYSIS PHASES
+    # ═══════════════════════════════════════════════════════════════════════
+    "analysis": {
+        "tools": BASE_READ_TOOLS + WEB_TOOLS,
+        "mcp_servers": ["context7"],
+        "auto_claude_tools": [],
+        "thinking_default": "medium",
+    },
+    "batch_analysis": {
+        "tools": BASE_READ_TOOLS + WEB_TOOLS,
+        "mcp_servers": [],
+        "auto_claude_tools": [],
+        "thinking_default": "low",
+    },
+    "batch_validation": {
+        "tools": BASE_READ_TOOLS,
+        "mcp_servers": [],
+        "auto_claude_tools": [],
+        "thinking_default": "low",
+    },
+    # ═══════════════════════════════════════════════════════════════════════
+    # ROADMAP & IDEATION
+    # ═══════════════════════════════════════════════════════════════════════
+    "roadmap_discovery": {
+        "tools": BASE_READ_TOOLS + WEB_TOOLS,
+        "mcp_servers": ["context7"],
+        "auto_claude_tools": [],
+        "thinking_default": "high",
+    },
+    "competitor_analysis": {
+        "tools": BASE_READ_TOOLS + WEB_TOOLS,
+        "mcp_servers": ["context7"],  # WebSearch for competitor research
+        "auto_claude_tools": [],
+        "thinking_default": "high",
+    },
+    "ideation": {
+        "tools": BASE_READ_TOOLS + WEB_TOOLS,
+        "mcp_servers": [],
+        "auto_claude_tools": [],
+        "thinking_default": "high",
+    },
+}
+
+
+# =============================================================================
+# Agent Config Helper Functions
+# =============================================================================
+
+
+def get_agent_config(agent_type: str) -> dict:
+    """
+    Get full configuration for an agent type.
+
+    Args:
+        agent_type: The agent type identifier (e.g., 'coder', 'planner', 'qa_reviewer')
+
+    Returns:
+        Configuration dict containing tools, mcp_servers, auto_claude_tools, thinking_default
+
+    Raises:
+        ValueError: If agent_type is not found in AGENT_CONFIGS (strict mode)
+    """
+    if agent_type not in AGENT_CONFIGS:
+        raise ValueError(
+            f"Unknown agent type: '{agent_type}'. "
+            f"Valid types: {sorted(AGENT_CONFIGS.keys())}"
+        )
+    return AGENT_CONFIGS[agent_type]
+
+
+def _map_mcp_server_name(
+    name: str, custom_server_ids: list[str] | None = None
+) -> str | None:
+    """
+    Map user-friendly MCP server names to internal identifiers.
+    Also accepts custom server IDs directly.
+
+    Args:
+        name: User-provided MCP server name
+        custom_server_ids: List of custom server IDs to accept as-is
+
+    Returns:
+        Internal server identifier or None if not recognized
+    """
+    if not name:
+        return None
+    mappings = {
+        "context7": "context7",
+        "graphiti-memory": "graphiti",
+        "graphiti": "graphiti",
+        "linear": "linear",
+        "electron": "electron",
+        "puppeteer": "puppeteer",
+        "auto-claude": "auto-claude",
+    }
+    # Check if it's a known mapping
+    mapped = mappings.get(name.lower().strip())
+    if mapped:
+        return mapped
+    # Check if it's a custom server ID (accept as-is)
+    if custom_server_ids and name in custom_server_ids:
+        return name
+    return None
+
+
+def get_required_mcp_servers(
+    agent_type: str,
+    project_capabilities: dict | None = None,
+    linear_enabled: bool = False,
+    mcp_config: dict | None = None,
+) -> list[str]:
+    """
+    Get MCP servers required for this agent type.
+
+    Handles dynamic server selection:
+    - "browser" → electron (if is_electron) or puppeteer (if is_web_frontend)
+    - "linear" → only if in mcp_servers_optional AND linear_enabled is True
+    - "graphiti" → only if GRAPHITI_MCP_URL is set
+    - Respects per-project MCP config overrides from .auto-claude/.env
+    - Applies per-agent ADD/REMOVE overrides from AGENT_MCP_<agent>_ADD/REMOVE
+
+    Args:
+        agent_type: The agent type identifier
+        project_capabilities: Dict from detect_project_capabilities() or None
+        linear_enabled: Whether Linear integration is enabled for this project
+        mcp_config: Per-project MCP server toggles from .auto-claude/.env
+                   Keys: CONTEXT7_ENABLED, LINEAR_MCP_ENABLED, ELECTRON_MCP_ENABLED,
+                         PUPPETEER_MCP_ENABLED, AGENT_MCP_<agent>_ADD/REMOVE
+
+    Returns:
+        List of MCP server names to start
+    """
+    config = get_agent_config(agent_type)
+    servers = list(config.get("mcp_servers", []))
+
+    # Load per-project config (or use defaults)
+    if mcp_config is None:
+        mcp_config = {}
+
+    # Filter context7 if explicitly disabled by project config
+    if "context7" in servers:
+        context7_enabled = mcp_config.get("CONTEXT7_ENABLED", "true")
+        if str(context7_enabled).lower() == "false":
+            servers = [s for s in servers if s != "context7"]
+
+    # Handle optional servers (e.g., Linear if project setting enabled)
+    optional = config.get("mcp_servers_optional", [])
+    if "linear" in optional and linear_enabled:
+        # Also check per-project LINEAR_MCP_ENABLED override
+        linear_mcp_enabled = mcp_config.get("LINEAR_MCP_ENABLED", "true")
+        if str(linear_mcp_enabled).lower() != "false":
+            servers.append("linear")
+
+    # Handle dynamic "browser" → electron/puppeteer based on project type and config
+    if "browser" in servers:
+        servers = [s for s in servers if s != "browser"]
+        if project_capabilities:
+            is_electron = project_capabilities.get("is_electron", False)
+            is_web_frontend = project_capabilities.get("is_web_frontend", False)
+
+            # Check per-project overrides (default false for both)
+            electron_enabled = mcp_config.get("ELECTRON_MCP_ENABLED", "false")
+            puppeteer_enabled = mcp_config.get("PUPPETEER_MCP_ENABLED", "false")
+
+            # Electron: enabled by project config OR global env var
+            if is_electron and (
+                str(electron_enabled).lower() == "true" or is_electron_mcp_enabled()
+            ):
+                servers.append("electron")
+            # Puppeteer: enabled by project config (no global env var)
+            elif is_web_frontend and not is_electron:
+                if str(puppeteer_enabled).lower() == "true":
+                    servers.append("puppeteer")
+
+    # Filter graphiti if not enabled
+    if "graphiti" in servers:
+        if not os.environ.get("GRAPHITI_MCP_URL"):
+            servers = [s for s in servers if s != "graphiti"]
+
+    # ========== Apply per-agent MCP overrides ==========
+    # Format: AGENT_MCP_<agent_type>_ADD=server1,server2
+    #         AGENT_MCP_<agent_type>_REMOVE=server1,server2
+    add_key = f"AGENT_MCP_{agent_type}_ADD"
+    remove_key = f"AGENT_MCP_{agent_type}_REMOVE"
+
+    # Extract custom server IDs for mapping (allows custom servers to be recognized)
+    custom_servers = mcp_config.get("CUSTOM_MCP_SERVERS", [])
+    custom_server_ids = [s.get("id") for s in custom_servers if s.get("id")]
+
+    # Process additions
+    if add_key in mcp_config:
+        additions = [
+            s.strip() for s in str(mcp_config[add_key]).split(",") if s.strip()
+        ]
+        for server in additions:
+            mapped = _map_mcp_server_name(server, custom_server_ids)
+            if mapped and mapped not in servers:
+                servers.append(mapped)
+
+    # Process removals (but never remove auto-claude)
+    if remove_key in mcp_config:
+        removals = [
+            s.strip() for s in str(mcp_config[remove_key]).split(",") if s.strip()
+        ]
+        for server in removals:
+            mapped = _map_mcp_server_name(server, custom_server_ids)
+            if mapped and mapped != "auto-claude":  # auto-claude cannot be removed
+                servers = [s for s in servers if s != mapped]
+
+    return servers
+
+
+def get_default_thinking_level(agent_type: str) -> str:
+    """
+    Get default thinking level string for agent type.
+
+    This returns the thinking level name (e.g., 'medium', 'high'), not the token budget.
+    To convert to tokens, use phase_config.get_thinking_budget(level).
+
+    Args:
+        agent_type: The agent type identifier
+
+    Returns:
+        Thinking level string (low, medium, high)
+    """
+    config = get_agent_config(agent_type)
+    return config.get("thinking_default", "medium")
@@ -0,0 +1,120 @@
+"""
+Agent Tool Permissions
+======================
+
+Manages which tools are allowed for each agent type to prevent context
+pollution and accidental misuse.
+
+Supports dynamic tool filtering based on project capabilities to optimize
+context window usage. For example, Electron tools are only included for
+Electron projects, not for Next.js or CLI projects.
+
+This module now uses AGENT_CONFIGS from models.py as the single source of truth
+for tool permissions. The get_allowed_tools() function remains the primary API
+for backwards compatibility.
+"""
+
+from .models import (
+    AGENT_CONFIGS,
+    CONTEXT7_TOOLS,
+    ELECTRON_TOOLS,
+    GRAPHITI_MCP_TOOLS,
+    LINEAR_TOOLS,
+    PUPPETEER_TOOLS,
+    get_agent_config,
+    get_required_mcp_servers,
+)
+from .registry import is_tools_available
+
+
+def get_allowed_tools(
+    agent_type: str,
+    project_capabilities: dict | None = None,
+    linear_enabled: bool = False,
+    mcp_config: dict | None = None,
+) -> list[str]:
+    """
+    Get the list of allowed tools for a specific agent type.
+
+    This ensures each agent only sees tools relevant to their role,
+    preventing context pollution and accidental misuse.
+
+    Uses AGENT_CONFIGS as the single source of truth for tool permissions.
+    Dynamic MCP tools are added based on project capabilities and required servers.
+
+    Args:
+        agent_type: Agent type identifier (e.g., 'coder', 'planner', 'qa_reviewer')
+        project_capabilities: Optional dict from detect_project_capabilities()
+                            containing flags like is_electron, is_web_frontend, etc.
+        linear_enabled: Whether Linear integration is enabled for this project
+        mcp_config: Per-project MCP server toggles from .auto-claude/.env
+
+    Returns:
+        List of allowed tool names
+
+    Raises:
+        ValueError: If agent_type is not found in AGENT_CONFIGS
+    """
+    # Get agent configuration (raises ValueError if unknown type)
+    config = get_agent_config(agent_type)
+
+    # Start with base tools from config
+    tools = list(config.get("tools", []))
+
+    # Get required MCP servers for this agent
+    required_servers = get_required_mcp_servers(
+        agent_type,
+        project_capabilities,
+        linear_enabled,
+        mcp_config,
+    )
+
+    # Add auto-claude tools ONLY if the MCP server is available
+    # This prevents allowing tools that won't work because the server isn't running
+    if "auto-claude" in required_servers and is_tools_available():
+        tools.extend(config.get("auto_claude_tools", []))
+
+    # Add MCP tool names based on required servers
+    tools.extend(_get_mcp_tools_for_servers(required_servers))
+
+    return tools
+
+
+def _get_mcp_tools_for_servers(servers: list[str]) -> list[str]:
+    """
+    Get the list of MCP tools for a list of required servers.
+
+    Maps server names to their corresponding tool lists.
+
+    Args:
+        servers: List of MCP server names (e.g., ['context7', 'linear', 'electron'])
+
+    Returns:
+        List of MCP tool names for all specified servers
+    """
+    tools = []
+
+    for server in servers:
+        if server == "context7":
+            tools.extend(CONTEXT7_TOOLS)
+        elif server == "linear":
+            tools.extend(LINEAR_TOOLS)
+        elif server == "graphiti":
+            tools.extend(GRAPHITI_MCP_TOOLS)
+        elif server == "electron":
+            tools.extend(ELECTRON_TOOLS)
+        elif server == "puppeteer":
+            tools.extend(PUPPETEER_TOOLS)
+        # auto-claude tools are already added via config["auto_claude_tools"]
+
+    return tools
+
+
+def get_all_agent_types() -> list[str]:
+    """
+    Get all registered agent types.
+
+    Returns:
+        Sorted list of all agent type identifiers
+    """
+    return sorted(AGENT_CONFIGS.keys())
@@ -0,0 +1,72 @@
+"""
+Tool Registry
+=============
+
+Central registry for creating and managing auto-claude MCP tools.
+"""
+
+from pathlib import Path
+
+try:
+    from claude_agent_sdk import create_sdk_mcp_server
+
+    SDK_TOOLS_AVAILABLE = True
+except ImportError:
+    SDK_TOOLS_AVAILABLE = False
+    create_sdk_mcp_server = None
+
+from .tools import (
+    create_memory_tools,
+    create_progress_tools,
+    create_qa_tools,
+    create_subtask_tools,
+)
+
+
+def create_all_tools(spec_dir: Path, project_dir: Path) -> list:
+    """
+    Create all custom tools with the given spec and project directories.
+
+    Args:
+        spec_dir: Path to the spec directory
+        project_dir: Path to the project root
+
+    Returns:
+        List of all tool functions
+    """
+    if not SDK_TOOLS_AVAILABLE:
+        return []
+
+    all_tools = []
+
+    # Create tools by category
+    all_tools.extend(create_subtask_tools(spec_dir, project_dir))
+    all_tools.extend(create_progress_tools(spec_dir, project_dir))
+    all_tools.extend(create_memory_tools(spec_dir, project_dir))
+    all_tools.extend(create_qa_tools(spec_dir, project_dir))
+
+    return all_tools
+
+
+def create_auto_claude_mcp_server(spec_dir: Path, project_dir: Path):
+    """
+    Create an MCP server with auto-claude custom tools.
+
+    Args:
+        spec_dir: Path to the spec directory
+        project_dir: Path to the project root
+
+    Returns:
+        MCP server instance, or None if SDK tools not available
+    """
+    if not SDK_TOOLS_AVAILABLE:
+        return None
+
+    tools = create_all_tools(spec_dir, project_dir)
+
+    return create_sdk_mcp_server(name="auto-claude", version="1.0.0", tools=tools)
+
+
+def is_tools_available() -> bool:
+    """Check if SDK tools functionality is available."""
+    return SDK_TOOLS_AVAILABLE
@@ -0,0 +1,18 @@
+"""
+Auto-Claude MCP Tools
+=====================
+
+Individual tool implementations organized by functionality.
+"""
+
+from .memory import create_memory_tools
+from .progress import create_progress_tools
+from .qa import create_qa_tools
+from .subtask import create_subtask_tools
+
+__all__ = [
+    "create_subtask_tools",
+    "create_progress_tools",
+    "create_memory_tools",
+    "create_qa_tools",
+]
@@ -0,0 +1,356 @@
+"""
+Session Memory Tools
+====================
+
+Tools for recording and retrieving session memory, including discoveries,
+gotchas, and patterns.
+
+Dual-storage approach:
+- File-based: Always available, works offline, spec-specific
+- LadybugDB: When Graphiti is enabled, also saves to graph database for
+  cross-session retrieval and Memory UI display
+"""
+
+import asyncio
+import json
+import logging
+from datetime import datetime, timezone
+from pathlib import Path
+from typing import Any
+
+try:
+    from claude_agent_sdk import tool
+
+    SDK_TOOLS_AVAILABLE = True
+except ImportError:
+    SDK_TOOLS_AVAILABLE = False
+    tool = None
+
+logger = logging.getLogger(__name__)
+
+
+async def _save_to_graphiti_async(
+    spec_dir: Path,
+    project_dir: Path,
+    save_type: str,
+    data: dict,
+) -> bool:
+    """
+    Save data to Graphiti/LadybugDB (async implementation).
+
+    Args:
+        spec_dir: Spec directory for GraphitiMemory initialization
+        project_dir: Project root directory
+        save_type: Type of save - 'discovery', 'gotcha', or 'pattern'
+        data: Data to save
+
+    Returns:
+        True if save succeeded, False otherwise
+    """
+    try:
+        # Use centralized helper for GraphitiMemory instantiation
+        # The helper handles enablement checks internally
+        from memory.graphiti_helpers import get_graphiti_memory
+
+        memory = await get_graphiti_memory(spec_dir, project_dir)
+        if memory is None:
+            return False
+
+        try:
+            if save_type == "discovery":
+                # Save as codebase discovery
+                # Format: {file_path: description}
+                result = await memory.save_codebase_discoveries(
+                    {data["file_path"]: data["description"]}
+                )
+            elif save_type == "gotcha":
+                # Save as gotcha
+                gotcha_text = data["gotcha"]
+                if data.get("context"):
+                    gotcha_text += f" (Context: {data['context']})"
+                result = await memory.save_gotcha(gotcha_text)
+            elif save_type == "pattern":
+                # Save as pattern
+                result = await memory.save_pattern(data["pattern"])
+            else:
+                result = False
+            return result
+        finally:
+            # Always close the memory connection (swallow exceptions to avoid overriding)
+            try:
+                await memory.close()
+            except Exception as e:
+                logger.debug(
+                    "Failed to close Graphiti memory connection", exc_info=True
+                )
+
+    except Exception as e:
+        logger.warning(f"Failed to save to Graphiti: {e}")
+        return False
+
+
+def _save_to_graphiti_sync(
+    spec_dir: Path,
+    project_dir: Path,
+    save_type: str,
+    data: dict,
+) -> bool:
+    """
+    Save data to Graphiti/LadybugDB (synchronous wrapper for sync contexts only).
+
+    NOTE: This should only be called from synchronous code. For async callers,
+    use _save_to_graphiti_async() directly to ensure proper resource cleanup.
+
+    Args:
+        spec_dir: Spec directory for GraphitiMemory initialization
+        project_dir: Project root directory
+        save_type: Type of save - 'discovery', 'gotcha', or 'pattern'
+        data: Data to save
+
+    Returns:
+        True if save succeeded, False otherwise
+    """
+    try:
+        # Check if we're already in an async context
+        try:
+            asyncio.get_running_loop()
+            # We're in an async context - caller should use _save_to_graphiti_async
+            # Log a warning and return False to avoid the resource leak bug
+            logger.warning(
+                "_save_to_graphiti_sync called from async context. "
+                "Use _save_to_graphiti_async instead for proper cleanup."
+            )
+            return False
+        except RuntimeError:
+            # No running loop - safe to create one
+            return asyncio.run(
+                _save_to_graphiti_async(spec_dir, project_dir, save_type, data)
+            )
+    except Exception as e:
+        logger.warning(f"Failed to save to Graphiti: {e}")
+        return False
+
+
+def create_memory_tools(spec_dir: Path, project_dir: Path) -> list:
+    """
+    Create session memory tools.
+
+    Args:
+        spec_dir: Path to the spec directory
+        project_dir: Path to the project root
+
+    Returns:
+        List of memory tool functions
+    """
+    if not SDK_TOOLS_AVAILABLE:
+        return []
+
+    tools = []
+
+    # -------------------------------------------------------------------------
+    # Tool: record_discovery
+    # -------------------------------------------------------------------------
+    @tool(
+        "record_discovery",
+        "Record a codebase discovery to session memory. Use this when you learn something important about the codebase.",
+        {"file_path": str, "description": str, "category": str},
+    )
+    async def record_discovery(args: dict[str, Any]) -> dict[str, Any]:
+        """Record a discovery to the codebase map (file + Graphiti)."""
+        file_path = args["file_path"]
+        description = args["description"]
+        category = args.get("category", "general")
+
+        memory_dir = spec_dir / "memory"
+        memory_dir.mkdir(exist_ok=True)
+
+        codebase_map_file = memory_dir / "codebase_map.json"
+        saved_to_graphiti = False
+
+        try:
+            # PRIMARY: Save to file-based storage (always works)
+            # Load existing map or create new
+            if codebase_map_file.exists():
+                with open(codebase_map_file, encoding="utf-8") as f:
+                    codebase_map = json.load(f)
+            else:
+                codebase_map = {
+                    "discovered_files": {},
+                    "last_updated": None,
+                }
+
+            # Add or update the discovery
+            codebase_map["discovered_files"][file_path] = {
+                "description": description,
+                "category": category,
+                "discovered_at": datetime.now(timezone.utc).isoformat(),
+            }
+            codebase_map["last_updated"] = datetime.now(timezone.utc).isoformat()
+
+            with open(codebase_map_file, "w", encoding="utf-8") as f:
+                json.dump(codebase_map, f, indent=2)
+
+            # SECONDARY: Also save to Graphiti/LadybugDB (for Memory UI)
+            saved_to_graphiti = await _save_to_graphiti_async(
+                spec_dir,
+                project_dir,
+                "discovery",
+                {
+                    "file_path": file_path,
+                    "description": f"[{category}] {description}",
+                },
+            )
+
+            storage_note = " (also saved to memory graph)" if saved_to_graphiti else ""
+            return {
+                "content": [
+                    {
+                        "type": "text",
+                        "text": f"Recorded discovery for '{file_path}': {description}{storage_note}",
+                    }
+                ]
+            }
+
+        except Exception as e:
+            return {
+                "content": [{"type": "text", "text": f"Error recording discovery: {e}"}]
+            }
+
+    tools.append(record_discovery)
+
+    # -------------------------------------------------------------------------
+    # Tool: record_gotcha
+    # -------------------------------------------------------------------------
+    @tool(
+        "record_gotcha",
+        "Record a gotcha or pitfall to avoid. Use this when you encounter something that future sessions should know.",
+        {"gotcha": str, "context": str},
+    )
+    async def record_gotcha(args: dict[str, Any]) -> dict[str, Any]:
+        """Record a gotcha to session memory (file + Graphiti)."""
+        gotcha = args["gotcha"]
+        context = args.get("context", "")
+
+        memory_dir = spec_dir / "memory"
+        memory_dir.mkdir(exist_ok=True)
+
+        gotchas_file = memory_dir / "gotchas.md"
+        saved_to_graphiti = False
+
+        try:
+            # PRIMARY: Save to file-based storage (always works)
+            timestamp = datetime.now(timezone.utc).strftime("%Y-%m-%d %H:%M")
+
+            entry = f"\n## [{timestamp}]\n{gotcha}"
+            if context:
+                entry += f"\n\n_Context: {context}_"
+            entry += "\n"
+
+            with open(gotchas_file, "a", encoding="utf-8") as f:
+                if not gotchas_file.exists() or gotchas_file.stat().st_size == 0:
+                    f.write(
+                        "# Gotchas & Pitfalls\n\nThings to watch out for in this codebase.\n"
+                    )
+                f.write(entry)
+
+            # SECONDARY: Also save to Graphiti/LadybugDB (for Memory UI)
+            saved_to_graphiti = await _save_to_graphiti_async(
+                spec_dir,
+                project_dir,
+                "gotcha",
+                {"gotcha": gotcha, "context": context},
+            )
+
+            storage_note = " (also saved to memory graph)" if saved_to_graphiti else ""
+            return {
+                "content": [
+                    {"type": "text", "text": f"Recorded gotcha: {gotcha}{storage_note}"}
+                ]
+            }
+
+        except Exception as e:
+            return {
+                "content": [{"type": "text", "text": f"Error recording gotcha: {e}"}]
+            }
+
+    tools.append(record_gotcha)
+
+    # -------------------------------------------------------------------------
+    # Tool: get_session_context
+    # -------------------------------------------------------------------------
+    @tool(
+        "get_session_context",
+        "Get context from previous sessions including discoveries, gotchas, and patterns.",
+        {},
+    )
+    async def get_session_context(args: dict[str, Any]) -> dict[str, Any]:
+        """Get accumulated session context."""
+        memory_dir = spec_dir / "memory"
+
+        if not memory_dir.exists():
+            return {
+                "content": [
+                    {
+                        "type": "text",
+                        "text": "No session memory found. This appears to be the first session.",
+                    }
+                ]
+            }
+
+        result_parts = []
+
+        # Load codebase map
+        codebase_map_file = memory_dir / "codebase_map.json"
+        if codebase_map_file.exists():
+            try:
+                with open(codebase_map_file, encoding="utf-8") as f:
+                    codebase_map = json.load(f)
+
+                discoveries = codebase_map.get("discovered_files", {})
+                if discoveries:
+                    result_parts.append("## Codebase Discoveries")
+                    for path, info in list(discoveries.items())[:20]:  # Limit to 20
+                        desc = info.get("description", "No description")
+                        result_parts.append(f"- `{path}`: {desc}")
+            except Exception:
+                pass
+
+        # Load gotchas
+        gotchas_file = memory_dir / "gotchas.md"
+        if gotchas_file.exists():
+            try:
+                content = gotchas_file.read_text(encoding="utf-8")
+                if content.strip():
+                    result_parts.append("\n## Gotchas")
+                    # Take last 1000 chars to avoid too much context
+                    result_parts.append(
+                        content[-1000:] if len(content) > 1000 else content
+                    )
+            except Exception:
+                pass
+
+        # Load patterns
+        patterns_file = memory_dir / "patterns.md"
+        if patterns_file.exists():
+            try:
+                content = patterns_file.read_text(encoding="utf-8")
+                if content.strip():
+                    result_parts.append("\n## Patterns")
+                    result_parts.append(
+                        content[-1000:] if len(content) > 1000 else content
+                    )
+            except Exception:
+                pass
+
+        if not result_parts:
+            return {
+                "content": [
+                    {"type": "text", "text": "No session context available yet."}
+                ]
+            }
+
+        return {"content": [{"type": "text", "text": "\n".join(result_parts)}]}
+
+    tools.append(get_session_context)
+
+    return tools
@@ -0,0 +1,142 @@
+"""
+Build Progress Tools
+====================
+
+Tools for tracking and reporting build progress.
+"""
+
+import json
+from pathlib import Path
+from typing import Any
+
+try:
+    from claude_agent_sdk import tool
+
+    SDK_TOOLS_AVAILABLE = True
+except ImportError:
+    SDK_TOOLS_AVAILABLE = False
+    tool = None
+
+
+def create_progress_tools(spec_dir: Path, project_dir: Path) -> list:
+    """
+    Create build progress tracking tools.
+
+    Args:
+        spec_dir: Path to the spec directory
+        project_dir: Path to the project root
+
+    Returns:
+        List of progress tool functions
+    """
+    if not SDK_TOOLS_AVAILABLE:
+        return []
+
+    tools = []
+
+    # -------------------------------------------------------------------------
+    # Tool: get_build_progress
+    # -------------------------------------------------------------------------
+    @tool(
+        "get_build_progress",
+        "Get the current build progress including completed subtasks, pending subtasks, and next subtask to work on.",
+        {},
+    )
+    async def get_build_progress(args: dict[str, Any]) -> dict[str, Any]:
+        """Get current build progress."""
+        plan_file = spec_dir / "implementation_plan.json"
+
+        if not plan_file.exists():
+            return {
+                "content": [
+                    {
+                        "type": "text",
+                        "text": "No implementation plan found. Run the planner first.",
+                    }
+                ]
+            }
+
+        try:
+            with open(plan_file, encoding="utf-8") as f:
+                plan = json.load(f)
+
+            stats = {
+                "total": 0,
+                "completed": 0,
+                "in_progress": 0,
+                "pending": 0,
+                "failed": 0,
+            }
+
+            phases_summary = []
+            next_subtask = None
+
+            for phase in plan.get("phases", []):
+                phase_id = phase.get("id") or phase.get("phase")
+                phase_name = phase.get("name", phase_id)
+                phase_subtasks = phase.get("subtasks", [])
+
+                phase_stats = {"completed": 0, "total": len(phase_subtasks)}
+
+                for subtask in phase_subtasks:
+                    stats["total"] += 1
+                    status = subtask.get("status", "pending")
+
+                    if status == "completed":
+                        stats["completed"] += 1
+                        phase_stats["completed"] += 1
+                    elif status == "in_progress":
+                        stats["in_progress"] += 1
+                    elif status == "failed":
+                        stats["failed"] += 1
+                    else:
+                        stats["pending"] += 1
+                        # Track next subtask to work on
+                        if next_subtask is None:
+                            next_subtask = {
+                                "id": subtask.get("id"),
+                                "description": subtask.get("description"),
+                                "phase": phase_name,
+                            }
+
+                phases_summary.append(
+                    f"  {phase_name}: {phase_stats['completed']}/{phase_stats['total']}"
+                )
+
+            progress_pct = (
+                (stats["completed"] / stats["total"] * 100) if stats["total"] > 0 else 0
+            )
+
+            result = f"""Build Progress: {stats["completed"]}/{stats["total"]} subtasks ({progress_pct:.0f}%)
+
+Status breakdown:
+  Completed: {stats["completed"]}
+  In Progress: {stats["in_progress"]}
+  Pending: {stats["pending"]}
+  Failed: {stats["failed"]}
+
+Phases:
+{chr(10).join(phases_summary)}"""
+
+            if next_subtask:
+                result += f"""
+
+Next subtask to work on:
+  ID: {next_subtask["id"]}
+  Phase: {next_subtask["phase"]}
+  Description: {next_subtask["description"]}"""
+            elif stats["completed"] == stats["total"]:
+                result += "\n\nAll subtasks completed! Build is ready for QA."
+
+            return {"content": [{"type": "text", "text": result}]}
+
+        except Exception as e:
+            return {
+                "content": [
+                    {"type": "text", "text": f"Error reading build progress: {e}"}
+                ]
+            }
+
+    tools.append(get_build_progress)
+
+    return tools
@@ -0,0 +1,204 @@
+"""
+QA Management Tools
+===================
+
+Tools for managing QA status and sign-off in implementation_plan.json.
+"""
+
+import json
+import logging
+from datetime import datetime, timezone
+from pathlib import Path
+from typing import Any
+
+from core.file_utils import write_json_atomic
+from spec.validate_pkg.auto_fix import auto_fix_plan
+
+try:
+    from claude_agent_sdk import tool
+
+    SDK_TOOLS_AVAILABLE = True
+except ImportError:
+    SDK_TOOLS_AVAILABLE = False
+    tool = None
+
+
+def _apply_qa_update(
+    plan: dict[str, Any],
+    status: str,
+    issues: list[Any],
+    tests_passed: dict[str, Any],
+) -> int:
+    """
+    Apply QA update to the plan and return the new QA session number.
+
+    Args:
+        plan: The implementation plan dict
+        status: QA status (pending, in_review, approved, rejected, fixes_applied)
+        issues: List of issues found
+        tests_passed: Dict of test results
+
+    Returns:
+        The new QA session number
+    """
+    # Get current QA session number
+    current_qa = plan.get("qa_signoff", {})
+    qa_session = current_qa.get("qa_session", 0)
+    if status in ["in_review", "rejected"]:
+        qa_session += 1
+
+    plan["qa_signoff"] = {
+        "status": status,
+        "qa_session": qa_session,
+        "issues_found": issues,
+        "tests_passed": tests_passed,
+        "timestamp": datetime.now(timezone.utc).isoformat(),
+        "ready_for_qa_revalidation": status == "fixes_applied",
+    }
+
+    # NOTE: Do NOT write plan["status"] or plan["planStatus"] here.
+    # The frontend XState task state machine owns status transitions.
+    # Writing status here races with XState's persistPlanStatusAndReasonSync()
+    # and can clobber the reviewReason field, causing tasks to appear "incomplete".
+
+    plan["last_updated"] = datetime.now(timezone.utc).isoformat()
+
+    return qa_session
+
+
+def create_qa_tools(spec_dir: Path, project_dir: Path) -> list:
+    """
+    Create QA management tools.
+
+    Args:
+        spec_dir: Path to the spec directory
+        project_dir: Path to the project root
+
+    Returns:
+        List of QA tool functions
+    """
+    if not SDK_TOOLS_AVAILABLE:
+        return []
+
+    tools = []
+
+    # -------------------------------------------------------------------------
+    # Tool: update_qa_status
+    # -------------------------------------------------------------------------
+    @tool(
+        "update_qa_status",
+        "Update the QA sign-off status in implementation_plan.json. Use after QA review.",
+        {"status": str, "issues": str, "tests_passed": str},
+    )
+    async def update_qa_status(args: dict[str, Any]) -> dict[str, Any]:
+        """Update QA status in the implementation plan."""
+        status = args["status"]
+        issues_str = args.get("issues", "[]")
+        tests_str = args.get("tests_passed", "{}")
+
+        valid_statuses = [
+            "pending",
+            "in_review",
+            "approved",
+            "rejected",
+            "fixes_applied",
+        ]
+        if status not in valid_statuses:
+            return {
+                "content": [
+                    {
+                        "type": "text",
+                        "text": f"Error: Invalid QA status '{status}'. Must be one of: {valid_statuses}",
+                    }
+                ]
+            }
+
+        plan_file = spec_dir / "implementation_plan.json"
+        if not plan_file.exists():
+            return {
+                "content": [
+                    {
+                        "type": "text",
+                        "text": "Error: implementation_plan.json not found",
+                    }
+                ]
+            }
+
+        try:
+            # Parse issues and tests
+            try:
+                issues = json.loads(issues_str) if issues_str else []
+            except json.JSONDecodeError:
+                issues = [{"description": issues_str}] if issues_str else []
+
+            try:
+                tests_passed = json.loads(tests_str) if tests_str else {}
+            except json.JSONDecodeError:
+                tests_passed = {}
+
+            with open(plan_file, encoding="utf-8") as f:
+                plan = json.load(f)
+
+            qa_session = _apply_qa_update(plan, status, issues, tests_passed)
+
+            # Use atomic write to prevent file corruption
+            write_json_atomic(plan_file, plan, indent=2)
+
+            return {
+                "content": [
+                    {
+                        "type": "text",
+                        "text": f"Updated QA status to '{status}' (session {qa_session})",
+                    }
+                ]
+            }
+
+        except json.JSONDecodeError as e:
+            # Attempt to auto-fix the plan and retry
+            if auto_fix_plan(spec_dir):
+                # Retry after fix
+                try:
+                    with open(plan_file, encoding="utf-8") as f:
+                        plan = json.load(f)
+
+                    qa_session = _apply_qa_update(plan, status, issues, tests_passed)
+                    write_json_atomic(plan_file, plan, indent=2)
+
+                    return {
+                        "content": [
+                            {
+                                "type": "text",
+                                "text": f"Updated QA status to '{status}' (session {qa_session}) (after auto-fix)",
+                            }
+                        ]
+                    }
+                except Exception as retry_err:
+                    logging.warning(
+                        f"QA update retry failed after auto-fix: {retry_err} (original error: {e})"
+                    )
+                    return {
+                        "content": [
+                            {
+                                "type": "text",
+                                "text": f"Error: QA update failed after auto-fix: {retry_err} (original JSON error: {e})",
+                            }
+                        ]
+                    }
+
+            return {
+                "content": [
+                    {
+                        "type": "text",
+                        "text": f"Error: Invalid JSON in implementation_plan.json: {e}",
+                    }
+                ]
+            }
+
+        except Exception as e:
+            return {
+                "content": [{"type": "text", "text": f"Error updating QA status: {e}"}]
+            }
+
+    tools.append(update_qa_status)
+
+    return tools
@@ -0,0 +1,204 @@
+"""
+Subtask Management Tools
+========================
+
+Tools for managing subtask status in implementation_plan.json.
+"""
+
+import json
+import logging
+from datetime import datetime, timezone
+from pathlib import Path
+from typing import Any
+
+from core.file_utils import write_json_atomic
+from spec.validate_pkg.auto_fix import auto_fix_plan
+
+try:
+    from claude_agent_sdk import tool
+
+    SDK_TOOLS_AVAILABLE = True
+except ImportError:
+    SDK_TOOLS_AVAILABLE = False
+    tool = None
+
+
+def _update_subtask_in_plan(
+    plan: dict[str, Any],
+    subtask_id: str,
+    status: str,
+    notes: str,
+) -> bool:
+    """
+    Update a subtask in the plan.
+
+    Args:
+        plan: The implementation plan dict
+        subtask_id: ID of the subtask to update
+        status: New status (pending, in_progress, completed, failed)
+        notes: Optional notes to add
+
+    Returns:
+        True if subtask was found and updated, False otherwise
+    """
+    subtask_found = False
+    for phase in plan.get("phases", []):
+        for subtask in phase.get("subtasks", []):
+            if subtask.get("id") == subtask_id:
+                subtask["status"] = status
+                if notes:
+                    subtask["notes"] = notes
+                subtask["updated_at"] = datetime.now(timezone.utc).isoformat()
+                subtask_found = True
+                break
+        if subtask_found:
+            break
+
+    if subtask_found:
+        plan["last_updated"] = datetime.now(timezone.utc).isoformat()
+
+    return subtask_found
+
+
+def create_subtask_tools(spec_dir: Path, project_dir: Path) -> list:
+    """
+    Create subtask management tools.
+
+    Args:
+        spec_dir: Path to the spec directory
+        project_dir: Path to the project root
+
+    Returns:
+        List of subtask tool functions
+    """
+    if not SDK_TOOLS_AVAILABLE:
+        return []
+
+    tools = []
+
+    # -------------------------------------------------------------------------
+    # Tool: update_subtask_status
+    # -------------------------------------------------------------------------
+    @tool(
+        "update_subtask_status",
+        "Update the status of a subtask in implementation_plan.json. Use this when completing or starting a subtask.",
+        {"subtask_id": str, "status": str, "notes": str},
+    )
+    async def update_subtask_status(args: dict[str, Any]) -> dict[str, Any]:
+        """Update subtask status in the implementation plan."""
+        subtask_id = args["subtask_id"]
+        status = args["status"]
+        notes = args.get("notes", "")
+
+        valid_statuses = ["pending", "in_progress", "completed", "failed"]
+        if status not in valid_statuses:
+            return {
+                "content": [
+                    {
+                        "type": "text",
+                        "text": f"Error: Invalid status '{status}'. Must be one of: {valid_statuses}",
+                    }
+                ]
+            }
+
+        plan_file = spec_dir / "implementation_plan.json"
+        if not plan_file.exists():
+            return {
+                "content": [
+                    {
+                        "type": "text",
+                        "text": "Error: implementation_plan.json not found",
+                    }
+                ]
+            }
+
+        try:
+            with open(plan_file, encoding="utf-8") as f:
+                plan = json.load(f)
+
+            subtask_found = _update_subtask_in_plan(plan, subtask_id, status, notes)
+
+            if not subtask_found:
+                return {
+                    "content": [
+                        {
+                            "type": "text",
+                            "text": f"Error: Subtask '{subtask_id}' not found in implementation plan",
+                        }
+                    ]
+                }
+
+            # Use atomic write to prevent file corruption
+            write_json_atomic(plan_file, plan, indent=2)
+
+            return {
+                "content": [
+                    {
+                        "type": "text",
+                        "text": f"Successfully updated subtask '{subtask_id}' to status '{status}'",
+                    }
+                ]
+            }
+
+        except json.JSONDecodeError as e:
+            # Attempt to auto-fix the plan and retry
+            if auto_fix_plan(spec_dir):
+                # Retry after fix
+                try:
+                    with open(plan_file, encoding="utf-8") as f:
+                        plan = json.load(f)
+
+                    subtask_found = _update_subtask_in_plan(
+                        plan, subtask_id, status, notes
+                    )
+
+                    if subtask_found:
+                        write_json_atomic(plan_file, plan, indent=2)
+                        return {
+                            "content": [
+                                {
+                                    "type": "text",
+                                    "text": f"Successfully updated subtask '{subtask_id}' to status '{status}' (after auto-fix)",
+                                }
+                            ]
+                        }
+                    else:
+                        return {
+                            "content": [
+                                {
+                                    "type": "text",
+                                    "text": f"Error: Subtask '{subtask_id}' not found in implementation plan (after auto-fix)",
+                                }
+                            ]
+                        }
+                except Exception as retry_err:
+                    logging.warning(
+                        f"Subtask update retry failed after auto-fix: {retry_err}"
+                    )
+                    return {
+                        "content": [
+                            {
+                                "type": "text",
+                                "text": f"Error: Subtask update failed after auto-fix: {retry_err}",
+                            }
+                        ]
+                    }
+
+            return {
+                "content": [
+                    {
+                        "type": "text",
+                        "text": f"Error: Invalid JSON in implementation_plan.json: {e}",
+                    }
+                ]
+            }
+        except Exception as e:
+            return {
+                "content": [
+                    {"type": "text", "text": f"Error updating subtask status: {e}"}
+                ]
+            }
+
+    tools.append(update_subtask_status)
+
+    return tools
@@ -0,0 +1,181 @@
+"""
+Utility Functions for Agent System
+===================================
+
+Helper functions for git operations, plan management, and file syncing.
+"""
+
+import json
+import logging
+import shutil
+from pathlib import Path
+
+from core.git_executable import run_git
+
+logger = logging.getLogger(__name__)
+
+
+def get_latest_commit(project_dir: Path) -> str | None:
+    """Get the hash of the latest git commit."""
+    result = run_git(
+        ["rev-parse", "HEAD"],
+        cwd=project_dir,
+        timeout=10,
+    )
+    if result.returncode == 0:
+        return result.stdout.strip()
+    return None
+
+
+def get_commit_count(project_dir: Path) -> int:
+    """Get the total number of commits."""
+    result = run_git(
+        ["rev-list", "--count", "HEAD"],
+        cwd=project_dir,
+        timeout=10,
+    )
+    if result.returncode == 0:
+        try:
+            return int(result.stdout.strip())
+        except ValueError:
+            return 0
+    return 0
+
+
+def load_implementation_plan(spec_dir: Path) -> dict | None:
+    """Load the implementation plan JSON."""
+    plan_file = spec_dir / "implementation_plan.json"
+    if not plan_file.exists():
+        return None
+    try:
+        with open(plan_file, encoding="utf-8") as f:
+            return json.load(f)
+    except (OSError, json.JSONDecodeError, UnicodeDecodeError):
+        return None
+
+
+def find_subtask_in_plan(plan: dict, subtask_id: str) -> dict | None:
+    """Find a subtask by ID in the plan."""
+    for phase in plan.get("phases", []):
+        for subtask in phase.get("subtasks", []):
+            if subtask.get("id") == subtask_id:
+                return subtask
+    return None
+
+
+def find_phase_for_subtask(plan: dict, subtask_id: str) -> dict | None:
+    """Find the phase containing a subtask."""
+    for phase in plan.get("phases", []):
+        for subtask in phase.get("subtasks", []):
+            if subtask.get("id") == subtask_id:
+                return phase
+    return None
+
+
+def sync_spec_to_source(spec_dir: Path, source_spec_dir: Path | None) -> bool:
+    """
+    Sync ALL spec files from worktree back to source spec directory.
+
+    When running in isolated mode (worktrees), the agent creates and updates
+    many files inside the worktree's spec directory. This function syncs ALL
+    of them back to the main project's spec directory.
+
+    IMPORTANT: Since .auto-claude/ is gitignored, this sync happens to the
+    local filesystem regardless of what branch the user is on. The worktree
+    may be on a different branch (e.g., auto-claude/093-task), but the sync
+    target is always the main project's .auto-claude/specs/ directory.
+
+    Files synced (all files in spec directory):
+    - implementation_plan.json - Task status and subtask completion
+    - build-progress.txt - Session-by-session progress notes
+    - task_logs.json - Execution logs
+    - review_state.json - QA review state
+    - critique_report.json - Spec critique findings
+    - suggested_commit_message.txt - Commit suggestions
+    - REGRESSION_TEST_REPORT.md - Test regression report
+    - spec.md, context.json, etc. - Original spec files (for completeness)
+    - memory/ directory - Codebase map, patterns, gotchas, session insights
+
+    Args:
+        spec_dir: Current spec directory (inside worktree)
+        source_spec_dir: Original spec directory in main project (outside worktree)
+
+    Returns:
+        True if sync was performed, False if not needed or failed
+    """
+    # Skip if no source specified or same path (not in worktree mode)
+    if not source_spec_dir:
+        return False
+
+    # Resolve paths and check if they're different
+    spec_dir_resolved = spec_dir.resolve()
+    source_spec_dir_resolved = source_spec_dir.resolve()
+
+    if spec_dir_resolved == source_spec_dir_resolved:
+        return False  # Same directory, no sync needed
+
+    synced_any = False
+
+    # Ensure source directory exists
+    source_spec_dir.mkdir(parents=True, exist_ok=True)
+
+    try:
+        # Sync all files and directories from worktree spec to source spec
+        for item in spec_dir.iterdir():
+            # Skip symlinks to prevent path traversal attacks
+            if item.is_symlink():
+                logger.warning(f"Skipping symlink during sync: {item.name}")
+                continue
+
+            source_item = source_spec_dir / item.name
+
+            if item.is_file():
+                # Copy file (preserves timestamps)
+                shutil.copy2(item, source_item)
+                logger.debug(f"Synced {item.name} to source")
+                synced_any = True
+
+            elif item.is_dir():
+                # Recursively sync directory
+                _sync_directory(item, source_item)
+                synced_any = True
+
+    except Exception as e:
+        logger.warning(f"Failed to sync spec directory to source: {e}")
+
+    return synced_any
+
+
+def _sync_directory(source_dir: Path, target_dir: Path) -> None:
+    """
+    Recursively sync a directory from source to target.
+
+    Args:
+        source_dir: Source directory (in worktree)
+        target_dir: Target directory (in main project)
+    """
+    # Create target directory if needed
+    target_dir.mkdir(parents=True, exist_ok=True)
+
+    for item in source_dir.iterdir():
+        # Skip symlinks to prevent path traversal attacks
+        if item.is_symlink():
+            logger.warning(
+                f"Skipping symlink during sync: {source_dir.name}/{item.name}"
+            )
+            continue
+
+        target_item = target_dir / item.name
+
+        if item.is_file():
+            shutil.copy2(item, target_item)
+            logger.debug(f"Synced {source_dir.name}/{item.name} to source")
+        elif item.is_dir():
+            # Recurse into subdirectories
+            _sync_directory(item, target_item)
+
+
+# Keep the old name as an alias for backward compatibility
+def sync_plan_to_source(spec_dir: Path, source_spec_dir: Path | None) -> bool:
+    """Alias for sync_spec_to_source for backward compatibility."""
+    return sync_spec_to_source(spec_dir, source_spec_dir)
@@ -0,0 +1,42 @@
+"""
+Analysis Module
+===============
+
+Code analysis and project scanning tools.
+"""
+
+# Import from analyzers subpackage (these are the modular analyzers)
+
+from __future__ import annotations
+
+from .analyzers import (
+    ProjectAnalyzer as ModularProjectAnalyzer,
+)
+from .analyzers import (
+    ServiceAnalyzer,
+    analyze_project,
+    analyze_service,
+)
+from .ci_discovery import CIDiscovery
+
+# Import from analysis module root (these are other analysis tools)
+from .project_analyzer import ProjectAnalyzer
+from .risk_classifier import RiskClassifier
+from .security_scanner import SecurityScanner
+
+# TestDiscovery was removed - tests are now co-located in their respective modules
+
+# insight_extractor is a module with functions, not a class, so don't import it here
+# Import it directly when needed: from analysis import insight_extractor
+
+__all__ = [
+    "ProjectAnalyzer",
+    "ModularProjectAnalyzer",
+    "ServiceAnalyzer",
+    "analyze_project",
+    "analyze_service",
+    "RiskClassifier",
+    "SecurityScanner",
+    "CIDiscovery",
+    # "TestDiscovery",  # Removed - tests now co-located in their modules
+]
@@ -0,0 +1,102 @@
+#!/usr/bin/env python3
+"""
+Codebase Analyzer
+=================
+
+Automatically detects project structure, frameworks, and services.
+Supports monorepos with multiple services.
+
+Usage:
+    # Index entire project (creates project_index.json)
+    python auto-claude/analyzer.py --index
+
+    # Analyze specific service
+    python auto-claude/analyzer.py --service backend
+
+    # Output to specific file
+    python auto-claude/analyzer.py --index --output path/to/output.json
+
+The analyzer will:
+1. Detect if this is a monorepo or single project
+2. Find all services/packages and analyze each separately
+3. Map interdependencies between services
+4. Identify infrastructure (Docker, CI/CD)
+5. Document conventions (linting, testing)
+
+This module now serves as a facade to the modular analyzer system in the analyzers/ package.
+All actual implementation is in focused submodules for better maintainability.
+"""
+
+from __future__ import annotations
+
+import json
+from pathlib import Path
+
+# Import from the new modular structure
+from .analyzers import (
+    ProjectAnalyzer,
+    ServiceAnalyzer,
+    analyze_project,
+    analyze_service,
+)
+
+# Re-export for backward compatibility
+__all__ = [
+    "ServiceAnalyzer",
+    "ProjectAnalyzer",
+    "analyze_project",
+    "analyze_service",
+]
+
+
+def main():
+    """CLI entry point."""
+    import argparse
+
+    parser = argparse.ArgumentParser(
+        description="Analyze project structure, frameworks, and services"
+    )
+    parser.add_argument(
+        "--project-dir",
+        type=Path,
+        default=Path.cwd(),
+        help="Project directory to analyze (default: current directory)",
+    )
+    parser.add_argument(
+        "--index",
+        action="store_true",
+        help="Create full project index (default behavior)",
+    )
+    parser.add_argument(
+        "--service",
+        type=str,
+        default=None,
+        help="Analyze a specific service only",
+    )
+    parser.add_argument(
+        "--output",
+        type=Path,
+        default=None,
+        help="Output file for JSON results",
+    )
+    parser.add_argument(
+        "--quiet",
+        action="store_true",
+        help="Only output JSON, no status messages",
+    )
+
+    args = parser.parse_args()
+
+    # Determine what to analyze
+    if args.service:
+        results = analyze_service(args.project_dir, args.service, args.output)
+    else:
+        results = analyze_project(args.project_dir, args.output)
+
+    # Print results
+    if not args.quiet or not args.output:
+        print(json.dumps(results, indent=2))
+
+
+if __name__ == "__main__":
+    main()
@@ -0,0 +1,94 @@
+"""
+Analyzers Package
+=================
+
+Modular analyzer system for detecting project structure, frameworks, and services.
+
+Main exports:
+- ServiceAnalyzer: Analyzes a single service/package
+- ProjectAnalyzer: Analyzes entire projects (single or monorepo)
+- analyze_project: Convenience function for project analysis
+- analyze_service: Convenience function for service analysis
+"""
+
+from __future__ import annotations
+
+from pathlib import Path
+from typing import Any
+
+from .project_analyzer_module import ProjectAnalyzer
+from .service_analyzer import ServiceAnalyzer
+
+# Re-export main classes
+__all__ = [
+    "ServiceAnalyzer",
+    "ProjectAnalyzer",
+    "analyze_project",
+    "analyze_service",
+]
+
+
+def analyze_project(project_dir: Path, output_file: Path | None = None) -> dict:
+    """
+    Analyze a project and optionally save results.
+
+    Args:
+        project_dir: Path to the project root
+        output_file: Optional path to save JSON output
+
+    Returns:
+        Project index as a dictionary
+    """
+    import json
+
+    analyzer = ProjectAnalyzer(project_dir)
+    results = analyzer.analyze()
+
+    if output_file:
+        output_file.parent.mkdir(parents=True, exist_ok=True)
+        with open(output_file, "w", encoding="utf-8") as f:
+            json.dump(results, f, indent=2)
+        print(f"Project index saved to: {output_file}")
+
+    return results
+
+
+def analyze_service(
+    project_dir: Path, service_name: str, output_file: Path | None = None
+) -> dict:
+    """
+    Analyze a specific service within a project.
+
+    Args:
+        project_dir: Path to the project root
+        service_name: Name of the service to analyze
+        output_file: Optional path to save JSON output
+
+    Returns:
+        Service analysis as a dictionary
+    """
+    import json
+
+    # Find the service
+    service_path = project_dir / service_name
+    if not service_path.exists():
+        # Check common locations
+        for parent in ["packages", "apps", "services"]:
+            candidate = project_dir / parent / service_name
+            if candidate.exists():
+                service_path = candidate
+                break
+
+    if not service_path.exists():
+        raise ValueError(f"Service '{service_name}' not found in {project_dir}")
+
+    analyzer = ServiceAnalyzer(service_path, service_name)
+    results = analyzer.analyze()
+
+    if output_file:
+        output_file.parent.mkdir(parents=True, exist_ok=True)
+        with open(output_file, "w", encoding="utf-8") as f:
+            json.dump(results, f, indent=2)
+        print(f"Service analysis saved to: {output_file}")
+
+    return results
@@ -0,0 +1,151 @@
+"""
+Base Analyzer Module
+====================
+
+Provides common constants, utilities, and base functionality shared across all analyzers.
+"""
+
+from __future__ import annotations
+
+import json
+from pathlib import Path
+
+# Directories to skip during analysis
+SKIP_DIRS = {
+    "node_modules",
+    ".git",
+    "__pycache__",
+    ".venv",
+    "venv",
+    ".env",
+    "env",
+    "dist",
+    "build",
+    ".next",
+    ".nuxt",
+    "target",
+    "vendor",
+    ".idea",
+    ".vscode",
+    ".pytest_cache",
+    ".mypy_cache",
+    "coverage",
+    ".coverage",
+    "htmlcov",
+    "eggs",
+    "*.egg-info",
+    ".turbo",
+    ".cache",
+    ".worktrees",  # Skip git worktrees directory
+    ".auto-claude",  # Skip auto-claude metadata directory
+}
+
+# Common service directory names
+SERVICE_INDICATORS = {
+    "backend",
+    "frontend",
+    "api",
+    "web",
+    "app",
+    "server",
+    "client",
+    "worker",
+    "workers",
+    "services",
+    "packages",
+    "apps",
+    "libs",
+    "scraper",
+    "crawler",
+    "proxy",
+    "gateway",
+    "admin",
+    "dashboard",
+    "mobile",
+    "desktop",
+    "cli",
+    "sdk",
+    "core",
+    "shared",
+    "common",
+}
+
+# Files that indicate a service root
+SERVICE_ROOT_FILES = {
+    "package.json",
+    "requirements.txt",
+    "pyproject.toml",
+    "Cargo.toml",
+    "go.mod",
+    "Gemfile",
+    "composer.json",
+    "pom.xml",
+    "build.gradle",
+    "Makefile",
+    "Dockerfile",
+}
+
+
+class BaseAnalyzer:
+    """Base class with common utilities for all analyzers."""
+
+    def __init__(self, path: Path):
+        self.path = path.resolve()
+
+    def _exists(self, path: str) -> bool:
+        """Check if a file exists relative to the analyzer's path."""
+        return (self.path / path).exists()
+
+    def _read_file(self, path: str) -> str:
+        """Read a file relative to the analyzer's path."""
+        try:
+            return (self.path / path).read_text(encoding="utf-8")
+        except (OSError, UnicodeDecodeError):
+            return ""
+
+    def _read_json(self, path: str) -> dict | None:
+        """Read and parse a JSON file relative to the analyzer's path."""
+        content = self._read_file(path)
+        if content:
+            try:
+                return json.loads(content)
+            except json.JSONDecodeError:
+                return None
+        return None
+
+    def _infer_env_var_type(self, value: str) -> str:
+        """Infer the type of an environment variable from its value."""
+        if not value:
+            return "string"
+
+        # Boolean
+        if value.lower() in ["true", "false", "1", "0", "yes", "no"]:
+            return "boolean"
+
+        # Number
+        if value.isdigit():
+            return "number"
+
+        # URL
+        if value.startswith(
+            (
+                "http://",
+                "https://",
+                "postgres://",
+                "postgresql://",
+                "mysql://",
+                "mongodb://",
+                "redis://",
+            )
+        ):
+            return "url"
+
+        # Email
+        if "@" in value and "." in value:
+            return "email"
+
+        # Path
+        if "/" in value or "\\" in value:
+            return "path"
+
+        return "string"
@@ -0,0 +1,26 @@
+"""
+Context Analyzer Package
+=========================
+
+Contains specialized detectors for comprehensive project context analysis.
+"""
+
+from __future__ import annotations
+
+from .api_docs_detector import ApiDocsDetector
+from .auth_detector import AuthDetector
+from .env_detector import EnvironmentDetector
+from .jobs_detector import JobsDetector
+from .migrations_detector import MigrationsDetector
+from .monitoring_detector import MonitoringDetector
+from .services_detector import ServicesDetector
+
+__all__ = [
+    "ApiDocsDetector",
+    "AuthDetector",
+    "EnvironmentDetector",
+    "JobsDetector",
+    "MigrationsDetector",
+    "MonitoringDetector",
+    "ServicesDetector",
+]
@@ -0,0 +1,95 @@
+"""
+API Documentation Detector Module
+==================================
+
+Detects API documentation tools and configurations:
+- OpenAPI/Swagger (FastAPI auto-generated, swagger-ui-express)
+- GraphQL playground
+- API documentation endpoints
+"""
+
+from __future__ import annotations
+
+from pathlib import Path
+from typing import Any
+
+from ..base import BaseAnalyzer
+
+
+class ApiDocsDetector(BaseAnalyzer):
+    """Detects API documentation setup."""
+
+    def __init__(self, path: Path, analysis: dict[str, Any]):
+        super().__init__(path)
+        self.analysis = analysis
+
+    def detect(self) -> None:
+        """
+        Detect API documentation setup.
+
+        Detects: OpenAPI/Swagger, GraphQL playground, API docs endpoints.
+        """
+        docs_info = {}
+
+        # Detect OpenAPI/Swagger
+        openapi_info = self._detect_fastapi() or self._detect_swagger_nodejs()
+        if openapi_info:
+            docs_info.update(openapi_info)
+
+        # Detect GraphQL
+        graphql_info = self._detect_graphql()
+        if graphql_info:
+            docs_info["graphql"] = graphql_info
+
+        if docs_info:
+            self.analysis["api_documentation"] = docs_info
+
+    def _detect_fastapi(self) -> dict[str, Any] | None:
+        """Detect FastAPI auto-generated OpenAPI docs."""
+        if self.analysis.get("framework") != "FastAPI":
+            return None
+
+        return {
+            "type": "openapi",
+            "auto_generated": True,
+            "docs_url": "/docs",
+            "redoc_url": "/redoc",
+            "openapi_url": "/openapi.json",
+        }
+
+    def _detect_swagger_nodejs(self) -> dict[str, Any] | None:
+        """Detect Swagger for Node.js projects."""
+        if not self._exists("package.json"):
+            return None
+
+        pkg = self._read_json("package.json")
+        if not pkg:
+            return None
+
+        deps = {**pkg.get("dependencies", {}), **pkg.get("devDependencies", {})}
+        if "swagger-ui-express" in deps or "swagger-jsdoc" in deps:
+            return {
+                "type": "openapi",
+                "library": "swagger-ui-express",
+                "docs_url": "/api-docs",
+            }
+
+        return None
+
+    def _detect_graphql(self) -> dict[str, str] | None:
+        """Detect GraphQL API and playground."""
+        if not self._exists("package.json"):
+            return None
+
+        pkg = self._read_json("package.json")
+        if not pkg:
+            return None
+
+        deps = {**pkg.get("dependencies", {}), **pkg.get("devDependencies", {})}
+        if "graphql" in deps or "apollo-server" in deps or "@apollo/server" in deps:
+            return {
+                "playground_url": "/graphql",
+                "library": "apollo-server" if "apollo-server" in deps else "graphql",
+            }
+
+        return None
@@ -0,0 +1,141 @@
+"""
+Authentication Patterns Detector Module
+========================================
+
+Detects authentication and authorization patterns:
+- JWT authentication
+- OAuth providers
+- Session-based authentication
+- API key authentication
+- User models
+- Auth middleware and decorators
+"""
+
+from __future__ import annotations
+
+import re
+from pathlib import Path
+from typing import Any
+
+from ..base import BaseAnalyzer
+
+
+class AuthDetector(BaseAnalyzer):
+    """Detects authentication and authorization patterns."""
+
+    JWT_LIBS = ["python-jose", "pyjwt", "jsonwebtoken", "jose"]
+    OAUTH_LIBS = ["authlib", "passport", "next-auth", "@auth/core", "oauth2"]
+    SESSION_LIBS = ["flask-login", "express-session", "django.contrib.auth"]
+
+    USER_MODEL_FILES = [
+        "models/user.py",
+        "models/User.py",
+        "app/models/user.py",
+        "models/user.ts",
+        "models/User.ts",
+        "src/models/user.ts",
+    ]
+
+    def __init__(self, path: Path, analysis: dict[str, Any]):
+        super().__init__(path)
+        self.analysis = analysis
+
+    def detect(self) -> None:
+        """
+        Detect authentication and authorization patterns.
+
+        Detects: JWT, OAuth, session-based, API keys, user models, protected routes.
+        """
+        auth_info = {
+            "strategies": [],
+            "libraries": [],
+            "user_model": None,
+            "middleware": [],
+        }
+
+        # Get all dependencies
+        all_deps = self._get_all_dependencies()
+
+        # Detect auth strategies and libraries
+        self._detect_jwt(all_deps, auth_info)
+        self._detect_oauth(all_deps, auth_info)
+        self._detect_session(all_deps, auth_info)
+
+        # Find user model
+        auth_info["user_model"] = self._find_user_model()
+
+        # Detect auth middleware/decorators
+        auth_info["middleware"] = self._find_auth_middleware()
+
+        # Remove duplicates from strategies
+        auth_info["strategies"] = list(set(auth_info["strategies"]))
+
+        if auth_info["strategies"] or auth_info["libraries"]:
+            self.analysis["auth"] = auth_info
+
+    def _get_all_dependencies(self) -> set[str]:
+        """Extract all dependencies from Python and Node.js projects."""
+        all_deps = set()
+
+        if self._exists("requirements.txt"):
+            content = self._read_file("requirements.txt")
+            all_deps.update(re.findall(r"^([a-zA-Z0-9_-]+)", content, re.MULTILINE))
+
+        pkg = self._read_json("package.json")
+        if pkg:
+            all_deps.update(pkg.get("dependencies", {}).keys())
+
+        return all_deps
+
+    def _detect_jwt(self, all_deps: set[str], auth_info: dict[str, Any]) -> None:
+        """Detect JWT authentication libraries."""
+        for lib in self.JWT_LIBS:
+            if lib in all_deps:
+                auth_info["strategies"].append("jwt")
+                auth_info["libraries"].append(lib)
+                break
+
+    def _detect_oauth(self, all_deps: set[str], auth_info: dict[str, Any]) -> None:
+        """Detect OAuth authentication libraries."""
+        for lib in self.OAUTH_LIBS:
+            if lib in all_deps:
+                auth_info["strategies"].append("oauth")
+                auth_info["libraries"].append(lib)
+                break
+
+    def _detect_session(self, all_deps: set[str], auth_info: dict[str, Any]) -> None:
+        """Detect session-based authentication libraries."""
+        for lib in self.SESSION_LIBS:
+            if lib in all_deps:
+                auth_info["strategies"].append("session")
+                auth_info["libraries"].append(lib)
+                break
+
+    def _find_user_model(self) -> str | None:
+        """Find the user model file."""
+        for model_file in self.USER_MODEL_FILES:
+            if self._exists(model_file):
+                return model_file
+        return None
+
+    def _find_auth_middleware(self) -> list[str]:
+        """Detect auth middleware and decorators from Python files."""
+        # Limit to first 20 files for performance
+        all_py_files = list(self.path.glob("**/*.py"))[:20]
+        auth_decorators = set()
+
+        for py_file in all_py_files:
+            try:
+                content = py_file.read_text(encoding="utf-8")
+                # Find custom decorators
+                if (
+                    "@require" in content
+                    or "@login_required" in content
+                    or "@authenticate" in content
+                ):
+                    decorators = re.findall(r"@(\w*(?:require|auth|login)\w*)", content)
+                    auth_decorators.update(decorators)
+            except (OSError, UnicodeDecodeError):
+                continue
+
+        return list(auth_decorators) if auth_decorators else []
@@ -0,0 +1,223 @@
+"""
+Environment Variable Detector Module
+=====================================
+
+Detects and analyzes environment variables from multiple sources:
+- .env files and variants
+- .env.example files
+- docker-compose.yml
+- Source code (os.getenv, process.env)
+"""
+
+from __future__ import annotations
+
+import re
+from pathlib import Path
+from typing import Any
+
+from ..base import BaseAnalyzer
+
+
+class EnvironmentDetector(BaseAnalyzer):
+    """Detects environment variables and their configurations."""
+
+    def __init__(self, path: Path, analysis: dict[str, Any]):
+        super().__init__(path)
+        self.analysis = analysis
+
+    def detect(self) -> None:
+        """
+        Discover all environment variables from multiple sources.
+
+        Extracts from: .env files, docker-compose, example files.
+        Categorizes as required/optional and detects sensitive data.
+        """
+        env_vars = {}
+        required_vars = set()
+        optional_vars = set()
+
+        # Parse various sources
+        self._parse_env_files(env_vars)
+        self._parse_env_example(env_vars, required_vars)
+        self._parse_docker_compose(env_vars)
+        self._parse_code_references(env_vars, optional_vars)
+
+        # Mark required vs optional
+        for key in env_vars:
+            if "required" not in env_vars[key]:
+                env_vars[key]["required"] = key in required_vars
+
+        if env_vars:
+            self.analysis["environment"] = {
+                "variables": env_vars,
+                "required_count": len(required_vars),
+                "optional_count": len(optional_vars),
+                "detected_count": len(env_vars),
+            }
+
+    def _parse_env_files(self, env_vars: dict[str, Any]) -> None:
+        """Parse .env files and variants."""
+        env_files = [
+            ".env",
+            ".env.local",
+            ".env.development",
+            ".env.production",
+            ".env.dev",
+            ".env.prod",
+            ".env.test",
+            ".env.staging",
+            "config/.env",
+            "../.env",
+        ]
+
+        for env_file in env_files:
+            content = self._read_file(env_file)
+            if not content:
+                continue
+
+            for line in content.split("\n"):
+                line = line.strip()
+                if not line or line.startswith("#"):
+                    continue
+
+                # Parse KEY=value or KEY="value" or KEY='value'
+                match = re.match(r"^([A-Z_][A-Z0-9_]*)\s*=\s*(.*)$", line)
+                if match:
+                    key = match.group(1)
+                    value = match.group(2).strip().strip('"').strip("'")
+
+                    # Detect if sensitive
+                    is_sensitive = self._is_sensitive_key(key)
+
+                    # Detect type
+                    var_type = self._infer_env_var_type(value)
+
+                    env_vars[key] = {
+                        "value": "<REDACTED>" if is_sensitive else value,
+                        "source": env_file,
+                        "type": var_type,
+                        "sensitive": is_sensitive,
+                    }
+
+    def _parse_env_example(
+        self, env_vars: dict[str, Any], required_vars: set[str]
+    ) -> None:
+        """Parse .env.example to find required variables."""
+        example_content = self._read_file(".env.example") or self._read_file(
+            ".env.sample"
+        )
+        if not example_content:
+            return
+
+        for line in example_content.split("\n"):
+            line = line.strip()
+            if not line or line.startswith("#"):
+                continue
+
+            match = re.match(r"^([A-Z_][A-Z0-9_]*)\s*=", line)
+            if match:
+                key = match.group(1)
+                required_vars.add(key)
+
+                if key not in env_vars:
+                    env_vars[key] = {
+                        "value": None,
+                        "source": ".env.example",
+                        "type": "string",
+                        "sensitive": self._is_sensitive_key(key),
+                        "required": True,
+                    }
+
+    def _parse_docker_compose(self, env_vars: dict[str, Any]) -> None:
+        """Parse docker-compose.yml environment section."""
+        for compose_file in ["docker-compose.yml", "../docker-compose.yml"]:
+            content = self._read_file(compose_file)
+            if not content:
+                continue
+
+            # Look for environment variables in docker-compose
+            in_env_section = False
+            for line in content.split("\n"):
+                if "environment:" in line:
+                    in_env_section = True
+                    continue
+
+                if in_env_section:
+                    # Check if we left the environment section
+                    if line and not line.startswith((" ", "\t", "-")):
+                        in_env_section = False
+                        continue
+
+                    # Parse - KEY=value or - KEY
+                    match = re.match(r"^\s*-\s*([A-Z_][A-Z0-9_]*)", line)
+                    if match:
+                        key = match.group(1)
+                        if key not in env_vars:
+                            env_vars[key] = {
+                                "value": None,
+                                "source": compose_file,
+                                "type": "string",
+                                "sensitive": False,
+                            }
+
+    def _parse_code_references(
+        self, env_vars: dict[str, Any], optional_vars: set[str]
+    ) -> None:
+        """Scan code for os.getenv() / process.env usage to find optional vars."""
+        entry_files = [
+            "app.py",
+            "main.py",
+            "config.py",
+            "settings.py",
+            "src/config.py",
+            "src/settings.py",
+            "index.js",
+            "index.ts",
+            "config.js",
+            "config.ts",
+        ]
+
+        for entry_file in entry_files:
+            content = self._read_file(entry_file)
+            if not content:
+                continue
+
+            # Python: os.getenv("VAR") or os.environ.get("VAR")
+            python_patterns = [
+                r'os\.getenv\(["\']([A-Z_][A-Z0-9_]*)["\']',
+                r'os\.environ\.get\(["\']([A-Z_][A-Z0-9_]*)["\']',
+                r'os\.environ\[["\']([A-Z_][A-Z0-9_]*)["\']',
+            ]
+
+            # JavaScript: process.env.VAR
+            js_patterns = [
+                r"process\.env\.([A-Z_][A-Z0-9_]*)",
+            ]
+
+            for pattern in python_patterns + js_patterns:
+                matches = re.findall(pattern, content)
+                for var_name in matches:
+                    if var_name not in env_vars:
+                        optional_vars.add(var_name)
+                        env_vars[var_name] = {
+                            "value": None,
+                            "source": f"code:{entry_file}",
+                            "type": "string",
+                            "sensitive": self._is_sensitive_key(var_name),
+                            "required": False,
+                        }
+
+    @staticmethod
+    def _is_sensitive_key(key: str) -> bool:
+        """Determine if an environment variable key contains sensitive data."""
+        sensitive_keywords = [
+            "secret",
+            "key",
+            "password",
+            "token",
+            "api_key",
+            "private",
+            "credential",
+            "auth",
+        ]
+        return any(keyword in key.lower() for keyword in sensitive_keywords)
@@ -0,0 +1,118 @@
+"""
+Background Jobs Detector Module
+================================
+
+Detects background job and task queue systems:
+- Celery (Python)
+- BullMQ/Bull (Node.js)
+- Sidekiq (Ruby)
+- Scheduled tasks and cron jobs
+"""
+
+from __future__ import annotations
+
+import re
+from pathlib import Path
+from typing import Any
+
+from ..base import BaseAnalyzer
+
+
+class JobsDetector(BaseAnalyzer):
+    """Detects background job and task queue systems."""
+
+    def __init__(self, path: Path, analysis: dict[str, Any]):
+        super().__init__(path)
+        self.analysis = analysis
+
+    def detect(self) -> None:
+        """
+        Detect background job/task queue systems.
+
+        Detects: Celery, BullMQ, Sidekiq, cron jobs, scheduled tasks.
+        """
+        jobs_info = None
+
+        # Try each job system in order
+        jobs_info = (
+            self._detect_celery() or self._detect_bullmq() or self._detect_sidekiq()
+        )
+
+        if jobs_info:
+            self.analysis["background_jobs"] = jobs_info
+
+    def _detect_celery(self) -> dict[str, Any] | None:
+        """Detect Celery (Python) task queue."""
+        celery_files = list(self.path.glob("**/celery.py")) + list(
+            self.path.glob("**/tasks.py")
+        )
+        if not celery_files:
+            return None
+
+        tasks = []
+        for task_file in celery_files:
+            try:
+                content = task_file.read_text(encoding="utf-8")
+                # Find @celery.task or @shared_task decorators
+                task_pattern = r"@(?:celery\.task|shared_task|app\.task)\s*(?:\([^)]*\))?\s*def\s+(\w+)"
+                task_matches = re.findall(task_pattern, content)
+
+                for task_name in task_matches:
+                    tasks.append(
+                        {
+                            "name": task_name,
+                            "file": str(task_file.relative_to(self.path)),
+                        }
+                    )
+
+            except (OSError, UnicodeDecodeError):
+                continue
+
+        if not tasks:
+            return None
+
+        return {
+            "system": "celery",
+            "tasks": tasks,
+            "total_tasks": len(tasks),
+            "worker_command": "celery -A app worker",
+        }
+
+    def _detect_bullmq(self) -> dict[str, Any] | None:
+        """Detect BullMQ/Bull (Node.js) task queue."""
+        if not self._exists("package.json"):
+            return None
+
+        pkg = self._read_json("package.json")
+        if not pkg:
+            return None
+
+        deps = pkg.get("dependencies", {})
+        if "bullmq" in deps:
+            return {
+                "system": "bullmq",
+                "tasks": [],
+                "worker_command": "node worker.js",
+            }
+        elif "bull" in deps:
+            return {
+                "system": "bull",
+                "tasks": [],
+                "worker_command": "node worker.js",
+            }
+
+        return None
+
+    def _detect_sidekiq(self) -> dict[str, Any] | None:
+        """Detect Sidekiq (Ruby) background jobs."""
+        if not self._exists("Gemfile"):
+            return None
+
+        gemfile = self._read_file("Gemfile")
+        if "sidekiq" not in gemfile.lower():
+            return None
+
+        return {
+            "system": "sidekiq",
+            "worker_command": "bundle exec sidekiq",
+        }
@@ -0,0 +1,129 @@
+"""
+Database Migrations Detector Module
+====================================
+
+Detects database migration tools and configurations:
+- Alembic (Python)
+- Django migrations
+- Knex (Node.js)
+- TypeORM
+- Prisma
+"""
+
+from __future__ import annotations
+
+from pathlib import Path
+from typing import Any
+
+from ..base import BaseAnalyzer
+
+
+class MigrationsDetector(BaseAnalyzer):
+    """Detects database migration setup and tools."""
+
+    def __init__(self, path: Path, analysis: dict[str, Any]):
+        super().__init__(path)
+        self.analysis = analysis
+
+    def detect(self) -> None:
+        """
+        Detect database migration setup.
+
+        Detects: Alembic, Django migrations, Knex, TypeORM, Prisma migrations.
+        """
+        migration_info = None
+
+        # Try each migration tool in order
+        migration_info = (
+            self._detect_alembic()
+            or self._detect_django()
+            or self._detect_knex()
+            or self._detect_typeorm()
+            or self._detect_prisma()
+        )
+
+        if migration_info:
+            self.analysis["migrations"] = migration_info
+
+    def _detect_alembic(self) -> dict[str, Any] | None:
+        """Detect Alembic (Python) migrations."""
+        if not (self._exists("alembic.ini") or self._exists("alembic")):
+            return None
+
+        return {
+            "tool": "alembic",
+            "directory": "alembic/versions"
+            if self._exists("alembic/versions")
+            else "alembic",
+            "config_file": "alembic.ini",
+            "commands": {
+                "upgrade": "alembic upgrade head",
+                "downgrade": "alembic downgrade -1",
+                "create": "alembic revision --autogenerate -m 'message'",
+            },
+        }
+
+    def _detect_django(self) -> dict[str, Any] | None:
+        """Detect Django migrations."""
+        if not self._exists("manage.py"):
+            return None
+
+        migration_dirs = list(self.path.glob("**/migrations"))
+        if not migration_dirs:
+            return None
+
+        return {
+            "tool": "django",
+            "directories": [str(d.relative_to(self.path)) for d in migration_dirs],
+            "commands": {
+                "migrate": "python manage.py migrate",
+                "makemigrations": "python manage.py makemigrations",
+            },
+        }
+
+    def _detect_knex(self) -> dict[str, Any] | None:
+        """Detect Knex (Node.js) migrations."""
+        if not (self._exists("knexfile.js") or self._exists("knexfile.ts")):
+            return None
+
+        return {
+            "tool": "knex",
+            "directory": "migrations",
+            "config_file": "knexfile.js",
+            "commands": {
+                "migrate": "knex migrate:latest",
+                "rollback": "knex migrate:rollback",
+                "create": "knex migrate:make migration_name",
+            },
+        }
+
+    def _detect_typeorm(self) -> dict[str, Any] | None:
+        """Detect TypeORM migrations."""
+        if not (self._exists("ormconfig.json") or self._exists("data-source.ts")):
+            return None
+
+        return {
+            "tool": "typeorm",
+            "directory": "migrations",
+            "commands": {
+                "run": "typeorm migration:run",
+                "revert": "typeorm migration:revert",
+                "create": "typeorm migration:create",
+            },
+        }
+
+    def _detect_prisma(self) -> dict[str, Any] | None:
+        """Detect Prisma migrations."""
+        if not self._exists("prisma/schema.prisma"):
+            return None
+
+        return {
+            "tool": "prisma",
+            "directory": "prisma/migrations",
+            "config_file": "prisma/schema.prisma",
+            "commands": {
+                "migrate": "prisma migrate deploy",
+                "dev": "prisma migrate dev",
+                "create": "prisma migrate dev --name migration_name",
+            },
+        }
@@ -0,0 +1,109 @@
+"""
+Monitoring Detector Module
+===========================
+
+Detects monitoring and observability setup:
+- Health check endpoints
+- Prometheus metrics endpoints
+- APM tools (Sentry, Datadog, New Relic)
+- Logging infrastructure
+"""
+
+from __future__ import annotations
+
+from pathlib import Path
+from typing import Any
+
+from ..base import BaseAnalyzer
+
+
+class MonitoringDetector(BaseAnalyzer):
+    """Detects monitoring and observability setup."""
+
+    def __init__(self, path: Path, analysis: dict[str, Any]):
+        super().__init__(path)
+        self.analysis = analysis
+
+    def detect(self) -> None:
+        """
+        Detect monitoring and observability setup.
+
+        Detects: Health checks, metrics endpoints, APM tools, logging.
+        """
+        monitoring_info = {}
+
+        # Detect health check endpoints from existing API analysis
+        health_checks = self._detect_health_checks()
+        if health_checks:
+            monitoring_info["health_checks"] = health_checks
+
+        # Detect Prometheus metrics
+        metrics_info = self._detect_prometheus()
+        if metrics_info:
+            monitoring_info.update(metrics_info)
+
+        # Reference APM tools from services analysis
+        apm_tools = self._get_apm_tools()
+        if apm_tools:
+            monitoring_info["apm_tools"] = apm_tools
+
+        if monitoring_info:
+            self.analysis["monitoring"] = monitoring_info
+
+    def _detect_health_checks(self) -> list[str] | None:
+        """Detect health check endpoints from API routes."""
+        if "api" not in self.analysis:
+            return None
+
+        routes = self.analysis["api"].get("routes", [])
+        health_routes = [
+            r["path"]
+            for r in routes
+            if "health" in r["path"].lower() or "ping" in r["path"].lower()
+        ]
+
+        return health_routes if health_routes else None
+
+    def _detect_prometheus(self) -> dict[str, str] | None:
+        """Detect Prometheus metrics endpoint."""
+        # Look for actual Prometheus imports/usage, not just keywords
+        all_files = (
+            list(self.path.glob("**/*.py"))[:30] + list(self.path.glob("**/*.js"))[:30]
+        )
+
+        for file_path in all_files:
+            # Skip analyzer files to avoid self-detection
+            if "analyzers" in str(file_path) or "analyzer.py" in str(file_path):
+                continue
+
+            try:
+                content = file_path.read_text(encoding="utf-8")
+                # Look for actual Prometheus imports or usage patterns
+                prometheus_patterns = [
+                    "from prometheus_client import",
+                    "import prometheus_client",
+                    "prometheus_client.",
+                    "@app.route('/metrics')",  # Flask
+                    "app.get('/metrics'",  # Express/Fastify
+                    "router.get('/metrics'",  # Express Router
+                ]
+
+                if any(pattern in content for pattern in prometheus_patterns):
+                    return {
+                        "metrics_endpoint": "/metrics",
+                        "metrics_type": "prometheus",
+                    }
+            except (OSError, UnicodeDecodeError):
+                continue
+
+        return None
+
+    def _get_apm_tools(self) -> list[str] | None:
+        """Get APM tools from existing services analysis."""
+        if (
+            "services" not in self.analysis
+            or "monitoring" not in self.analysis["services"]
+        ):
+            return None
+
+        return [s["type"] for s in self.analysis["services"]["monitoring"]]
@@ -0,0 +1,215 @@
+"""
+External Services Detector Module
+==================================
+
+Detects external service integrations based on dependencies:
+- Databases (PostgreSQL, MySQL, MongoDB, Redis, SQLite)
+- Cache services (Redis, Memcached)
+- Message queues (Celery, BullMQ, Kafka, RabbitMQ)
+- Email services (SendGrid, Mailgun, Postmark)
+- Payment processors (Stripe, PayPal, Square)
+- Storage services (AWS S3, Google Cloud Storage, Azure)
+- Auth providers (OAuth, JWT)
+- Monitoring tools (Sentry, Datadog, New Relic)
+"""
+
+from __future__ import annotations
+
+import re
+from pathlib import Path
+from typing import Any
+
+from ..base import BaseAnalyzer
+
+
+class ServicesDetector(BaseAnalyzer):
+    """Detects external service integrations."""
+
+    # Service indicator mappings
+    DATABASE_INDICATORS = {
+        "psycopg2": "postgresql",
+        "psycopg2-binary": "postgresql",
+        "pg": "postgresql",
+        "mysql": "mysql",
+        "mysql2": "mysql",
+        "pymongo": "mongodb",
+        "mongodb": "mongodb",
+        "mongoose": "mongodb",
+        "redis": "redis",
+        "redis-py": "redis",
+        "ioredis": "redis",
+        "sqlite3": "sqlite",
+        "better-sqlite3": "sqlite",
+    }
+
+    CACHE_INDICATORS = ["redis", "memcached", "node-cache"]
+
+    QUEUE_INDICATORS = {
+        "celery": "celery",
+        "bullmq": "bullmq",
+        "bull": "bull",
+        "kafka-python": "kafka",
+        "kafkajs": "kafka",
+        "amqplib": "rabbitmq",
+        "amqp": "rabbitmq",
+    }
+
+    EMAIL_INDICATORS = {
+        "sendgrid": "sendgrid",
+        "@sendgrid/mail": "sendgrid",
+        "nodemailer": "smtp",
+        "mailgun": "mailgun",
+        "postmark": "postmark",
+    }
+
+    PAYMENT_INDICATORS = {
+        "stripe": "stripe",
+        "paypal": "paypal",
+        "square": "square",
+        "braintree": "braintree",
+    }
+
+    STORAGE_INDICATORS = {
+        "boto3": "aws_s3",
+        "@aws-sdk/client-s3": "aws_s3",
+        "aws-sdk": "aws_s3",
+        "@google-cloud/storage": "google_cloud_storage",
+        "azure-storage-blob": "azure_blob_storage",
+    }
+
+    AUTH_INDICATORS = {
+        "authlib": "oauth",
+        "python-jose": "jwt",
+        "pyjwt": "jwt",
+        "jsonwebtoken": "jwt",
+        "passport": "oauth",
+        "next-auth": "oauth",
+        "@auth/core": "oauth",
+    }
+
+    MONITORING_INDICATORS = {
+        "sentry-sdk": "sentry",
+        "@sentry/node": "sentry",
+        "datadog": "datadog",
+        "newrelic": "new_relic",
+        "loguru": "logging",
+        "winston": "logging",
+        "pino": "logging",
+    }
+
+    def __init__(self, path: Path, analysis: dict[str, Any]):
+        super().__init__(path)
+        self.analysis = analysis
+
+    def detect(self) -> None:
+        """
+        Detect external service integrations.
+
+        Detects: databases, cache, email, payments, storage, monitoring, etc.
+        """
+        services = {
+            "databases": [],
+            "cache": [],
+            "message_queues": [],
+            "email": [],
+            "payments": [],
+            "storage": [],
+            "auth_providers": [],
+            "monitoring": [],
+        }
+
+        # Get all dependencies
+        all_deps = self._get_all_dependencies()
+
+        # Detect each service category
+        self._detect_databases(all_deps, services["databases"])
+        self._detect_cache(all_deps, services["cache"])
+        self._detect_message_queues(all_deps, services["message_queues"])
+        self._detect_email(all_deps, services["email"])
+        self._detect_payments(all_deps, services["payments"])
+        self._detect_storage(all_deps, services["storage"])
+        self._detect_auth_providers(all_deps, services["auth_providers"])
+        self._detect_monitoring(all_deps, services["monitoring"])
+
+        # Remove empty categories
+        services = {k: v for k, v in services.items() if v}
+
+        if services:
+            self.analysis["services"] = services
+
+    def _get_all_dependencies(self) -> set[str]:
+        """Extract all dependencies from Python and Node.js projects."""
+        all_deps = set()
+
+        # Python dependencies
+        if self._exists("requirements.txt"):
+            content = self._read_file("requirements.txt")
+            all_deps.update(re.findall(r"^([a-zA-Z0-9_-]+)", content, re.MULTILINE))
+
+        # Node.js dependencies
+        pkg = self._read_json("package.json")
+        if pkg:
+            all_deps.update(pkg.get("dependencies", {}).keys())
+            all_deps.update(pkg.get("devDependencies", {}).keys())
+
+        return all_deps
+
+    def _detect_databases(
+        self, all_deps: set[str], databases: list[dict[str, str]]
+    ) -> None:
+        """Detect database clients."""
+        for dep, db_type in self.DATABASE_INDICATORS.items():
+            if dep in all_deps:
+                databases.append({"type": db_type, "client": dep})
+
+    def _detect_cache(self, all_deps: set[str], cache: list[dict[str, str]]) -> None:
+        """Detect cache services."""
+        for indicator in self.CACHE_INDICATORS:
+            if indicator in all_deps:
+                cache.append({"type": indicator})
+
+    def _detect_message_queues(
+        self, all_deps: set[str], queues: list[dict[str, str]]
+    ) -> None:
+        """Detect message queue systems."""
+        for dep, queue_type in self.QUEUE_INDICATORS.items():
+            if dep in all_deps:
+                queues.append({"type": queue_type, "client": dep})
+
+    def _detect_email(self, all_deps: set[str], email: list[dict[str, str]]) -> None:
+        """Detect email service providers."""
+        for dep, email_type in self.EMAIL_INDICATORS.items():
+            if dep in all_deps:
+                email.append({"provider": email_type, "client": dep})
+
+    def _detect_payments(
+        self, all_deps: set[str], payments: list[dict[str, str]]
+    ) -> None:
+        """Detect payment processors."""
+        for dep, payment_type in self.PAYMENT_INDICATORS.items():
+            if dep in all_deps:
+                payments.append({"provider": payment_type, "client": dep})
+
+    def _detect_storage(
+        self, all_deps: set[str], storage: list[dict[str, str]]
+    ) -> None:
+        """Detect storage services."""
+        for dep, storage_type in self.STORAGE_INDICATORS.items():
+            if dep in all_deps:
+                storage.append({"provider": storage_type, "client": dep})
+
+    def _detect_auth_providers(
+        self, all_deps: set[str], auth: list[dict[str, str]]
+    ) -> None:
+        """Detect authentication providers."""
+        for dep, auth_type in self.AUTH_INDICATORS.items():
+            if dep in all_deps:
+                auth.append({"type": auth_type, "client": dep})
+
+    def _detect_monitoring(
+        self, all_deps: set[str], monitoring: list[dict[str, str]]
+    ) -> None:
+        """Detect monitoring and observability tools."""
+        for dep, monitoring_type in self.MONITORING_INDICATORS.items():
+            if dep in all_deps:
+                monitoring.append({"type": monitoring_type, "client": dep})
@@ -0,0 +1,102 @@
+"""
+Context Analyzer Module
+=======================
+
+Orchestrates comprehensive project context analysis including:
+- Environment variables and configuration
+- External service integrations
+- Authentication patterns
+- Database migrations
+- Background jobs/task queues
+- API documentation
+- Monitoring and observability
+
+This module delegates to specialized detectors for clean separation of concerns.
+"""
+
+from __future__ import annotations
+
+from pathlib import Path
+from typing import Any
+
+from .base import BaseAnalyzer
+from .context import (
+    ApiDocsDetector,
+    AuthDetector,
+    EnvironmentDetector,
+    JobsDetector,
+    MigrationsDetector,
+    MonitoringDetector,
+    ServicesDetector,
+)
+
+
+class ContextAnalyzer(BaseAnalyzer):
+    """Orchestrates project context and configuration analysis."""
+
+    def __init__(self, path: Path, analysis: dict[str, Any]):
+        super().__init__(path)
+        self.analysis = analysis
+
+    def detect_environment_variables(self) -> None:
+        """
+        Discover all environment variables from multiple sources.
+
+        Delegates to EnvironmentDetector for actual detection logic.
+        """
+        detector = EnvironmentDetector(self.path, self.analysis)
+        detector.detect()
+
+    def detect_external_services(self) -> None:
+        """
+        Detect external service integrations.
+
+        Delegates to ServicesDetector for actual detection logic.
+        """
+        detector = ServicesDetector(self.path, self.analysis)
+        detector.detect()
+
+    def detect_auth_patterns(self) -> None:
+        """
+        Detect authentication and authorization patterns.
+
+        Delegates to AuthDetector for actual detection logic.
+        """
+        detector = AuthDetector(self.path, self.analysis)
+        detector.detect()
+
+    def detect_migrations(self) -> None:
+        """
+        Detect database migration setup.
+
+        Delegates to MigrationsDetector for actual detection logic.
+        """
+        detector = MigrationsDetector(self.path, self.analysis)
+        detector.detect()
+
+    def detect_background_jobs(self) -> None:
+        """
+        Detect background job/task queue systems.
+
+        Delegates to JobsDetector for actual detection logic.
+        """
+        detector = JobsDetector(self.path, self.analysis)
+        detector.detect()
+
+    def detect_api_documentation(self) -> None:
+        """
+        Detect API documentation setup.
+
+        Delegates to ApiDocsDetector for actual detection logic.
+        """
+        detector = ApiDocsDetector(self.path, self.analysis)
+        detector.detect()
+
+    def detect_monitoring(self) -> None:
+        """
+        Detect monitoring and observability setup.
+
+        Delegates to MonitoringDetector for actual detection logic.
+        """
+        detector = MonitoringDetector(self.path, self.analysis)
+        detector.detect()
@@ -0,0 +1,316 @@
+"""
+Database Detector Module
+========================
+
+Detects database models and schemas across different ORMs:
+- Python: SQLAlchemy, Django ORM
+- JavaScript/TypeScript: Prisma, TypeORM, Drizzle, Mongoose
+"""
+
+from __future__ import annotations
+
+import re
+from pathlib import Path
+
+from .base import BaseAnalyzer
+
+
+class DatabaseDetector(BaseAnalyzer):
+    """Detects database models across multiple ORMs."""
+
+    def __init__(self, path: Path):
+        super().__init__(path)
+
+    def detect_all_models(self) -> dict:
+        """Detect all database models across different ORMs."""
+        models = {}
+
+        # Python SQLAlchemy
+        models.update(self._detect_sqlalchemy_models())
+
+        # Python Django
+        models.update(self._detect_django_models())
+
+        # Prisma schema
+        models.update(self._detect_prisma_models())
+
+        # TypeORM entities
+        models.update(self._detect_typeorm_models())
+
+        # Drizzle schema
+        models.update(self._detect_drizzle_models())
+
+        # Mongoose models
+        models.update(self._detect_mongoose_models())
+
+        return models
+
+    def _detect_sqlalchemy_models(self) -> dict:
+        """Detect SQLAlchemy models."""
+        models = {}
+        py_files = list(self.path.glob("**/*.py"))
+
+        for file_path in py_files:
+            try:
+                content = file_path.read_text(encoding="utf-8")
+            except (OSError, UnicodeDecodeError):
+                continue
+
+            # Find class definitions that inherit from Base or db.Model
+            class_pattern = (
+                r"class\s+(\w+)\([^)]*(?:Base|db\.Model|DeclarativeBase)[^)]*\):"
+            )
+            matches = re.finditer(class_pattern, content)
+
+            for match in matches:
+                model_name = match.group(1)
+
+                # Extract table name if defined
+                table_match = re.search(r'__tablename__\s*=\s*["\'](\w+)["\']', content)
+                table_name = (
+                    table_match.group(1) if table_match else model_name.lower() + "s"
+                )
+
+                # Extract columns
+                fields = {}
+                column_pattern = r"(\w+)\s*=\s*Column\((.*?)\)"
+                column_matches = re.finditer(
+                    column_pattern, content[match.end() : match.end() + 2000]
+                )
+
+                for col_match in column_matches:
+                    field_name = col_match.group(1)
+                    field_def = col_match.group(2)
+
+                    # Detect field properties
+                    is_primary = "primary_key=True" in field_def
+                    is_unique = "unique=True" in field_def
+                    is_nullable = "nullable=False" not in field_def
+
+                    # Extract type
+                    type_match = re.search(
+                        r"(Integer|String|Text|Boolean|DateTime|Float|JSON)", field_def
+                    )
+                    field_type = type_match.group(1) if type_match else "Unknown"
+
+                    fields[field_name] = {
+                        "type": field_type,
+                        "primary_key": is_primary,
+                        "unique": is_unique,
+                        "nullable": is_nullable,
+                    }
+
+                if fields:  # Only add if we found fields
+                    models[model_name] = {
+                        "table": table_name,
+                        "fields": fields,
+                        "file": str(file_path.relative_to(self.path)),
+                        "orm": "SQLAlchemy",
+                    }
+
+        return models
+
+    def _detect_django_models(self) -> dict:
+        """Detect Django models."""
+        models = {}
+        model_files = list(self.path.glob("**/models.py")) + list(
+            self.path.glob("**/models/*.py")
+        )
+
+        for file_path in model_files:
+            try:
+                content = file_path.read_text(encoding="utf-8")
+            except (OSError, UnicodeDecodeError):
+                continue
+
+            # Find class definitions that inherit from models.Model
+            class_pattern = r"class\s+(\w+)\(models\.Model\):"
+            matches = re.finditer(class_pattern, content)
+
+            for match in matches:
+                model_name = match.group(1)
+                table_name = model_name.lower()
+
+                # Extract fields
+                fields = {}
+                field_pattern = r"(\w+)\s*=\s*models\.(\w+Field)\((.*?)\)"
+                field_matches = re.finditer(
+                    field_pattern, content[match.end() : match.end() + 2000]
+                )
+
+                for field_match in field_matches:
+                    field_name = field_match.group(1)
+                    field_type = field_match.group(2)
+                    field_args = field_match.group(3)
+
+                    fields[field_name] = {
+                        "type": field_type,
+                        "unique": "unique=True" in field_args,
+                        "nullable": "null=True" in field_args,
+                    }
+
+                if fields:
+                    models[model_name] = {
+                        "table": table_name,
+                        "fields": fields,
+                        "file": str(file_path.relative_to(self.path)),
+                        "orm": "Django",
+                    }
+
+        return models
+
+    def _detect_prisma_models(self) -> dict:
+        """Detect Prisma models from schema.prisma."""
+        models = {}
+        schema_file = self.path / "prisma" / "schema.prisma"
+
+        if not schema_file.exists():
+            return models
+
+        try:
+            content = schema_file.read_text(encoding="utf-8")
+        except (OSError, UnicodeDecodeError):
+            return models
+
+        # Find model definitions
+        model_pattern = r"model\s+(\w+)\s*\{([^}]+)\}"
+        matches = re.finditer(model_pattern, content, re.MULTILINE)
+
+        for match in matches:
+            model_name = match.group(1)
+            model_body = match.group(2)
+
+            fields = {}
+            # Parse fields: id Int @id @default(autoincrement())
+            field_pattern = r"(\w+)\s+(\w+)([^/\n]*)"
+            field_matches = re.finditer(field_pattern, model_body)
+
+            for field_match in field_matches:
+                field_name = field_match.group(1)
+                field_type = field_match.group(2)
+                field_attrs = field_match.group(3)
+
+                fields[field_name] = {
+                    "type": field_type,
+                    "primary_key": "@id" in field_attrs,
+                    "unique": "@unique" in field_attrs,
+                    "nullable": "?" in field_type,
+                }
+
+            if fields:
+                models[model_name] = {
+                    "table": model_name.lower(),
+                    "fields": fields,
+                    "file": "prisma/schema.prisma",
+                    "orm": "Prisma",
+                }
+
+        return models
+
+    def _detect_typeorm_models(self) -> dict:
+        """Detect TypeORM entities."""
+        models = {}
+        ts_files = list(self.path.glob("**/*.entity.ts")) + list(
+            self.path.glob("**/entities/*.ts")
+        )
+
+        for file_path in ts_files:
+            try:
+                content = file_path.read_text(encoding="utf-8")
+            except (OSError, UnicodeDecodeError):
+                continue
+
+            # Find @Entity() class declarations
+            entity_pattern = r"@Entity\([^)]*\)\s*(?:export\s+)?class\s+(\w+)"
+            matches = re.finditer(entity_pattern, content)
+
+            for match in matches:
+                model_name = match.group(1)
+
+                # Extract columns
+                fields = {}
+                column_pattern = (
+                    r"@(PrimaryGeneratedColumn|Column)\(([^)]*)\)\s+(\w+):\s*(\w+)"
+                )
+                column_matches = re.finditer(column_pattern, content)
+
+                for col_match in column_matches:
+                    decorator = col_match.group(1)
+                    options = col_match.group(2)
+                    field_name = col_match.group(3)
+                    field_type = col_match.group(4)
+
+                    fields[field_name] = {
+                        "type": field_type,
+                        "primary_key": decorator == "PrimaryGeneratedColumn",
+                        "unique": "unique: true" in options,
+                    }
+
+                if fields:
+                    models[model_name] = {
+                        "table": model_name.lower(),
+                        "fields": fields,
+                        "file": str(file_path.relative_to(self.path)),
+                        "orm": "TypeORM",
+                    }
+
+        return models
+
+    def _detect_drizzle_models(self) -> dict:
+        """Detect Drizzle ORM schemas."""
+        models = {}
+        schema_files = list(self.path.glob("**/schema.ts")) + list(
+            self.path.glob("**/db/schema.ts")
+        )
+
+        for file_path in schema_files:
+            try:
+                content = file_path.read_text(encoding="utf-8")
+            except (OSError, UnicodeDecodeError):
+                continue
+
+            # Find table definitions: export const users = pgTable('users', {...})
+            table_pattern = r'export\s+const\s+(\w+)\s*=\s*(?:pg|mysql|sqlite)Table\(["\'](\w+)["\']'
+            matches = re.finditer(table_pattern, content)
+
+            for match in matches:
+                const_name = match.group(1)
+                table_name = match.group(2)
+
+                models[const_name] = {
+                    "table": table_name,
+                    "fields": {},  # Would need more parsing for fields
+                    "file": str(file_path.relative_to(self.path)),
+                    "orm": "Drizzle",
+                }
+
+        return models
+
+    def _detect_mongoose_models(self) -> dict:
+        """Detect Mongoose models."""
+        models = {}
+        model_files = list(self.path.glob("**/models/*.js")) + list(
+            self.path.glob("**/models/*.ts")
+        )
+
+        for file_path in model_files:
+            try:
+                content = file_path.read_text(encoding="utf-8")
+            except (OSError, UnicodeDecodeError):
+                continue
+
+            # Find mongoose.model() or new Schema()
+            model_pattern = r'mongoose\.model\(["\'](\w+)["\']'
+            matches = re.finditer(model_pattern, content)
+
+            for match in matches:
+                model_name = match.group(1)
+
+                models[model_name] = {
+                    "table": model_name.lower(),
+                    "fields": {},
+                    "file": str(file_path.relative_to(self.path)),
+                    "orm": "Mongoose",
+                }
+
+        return models
@@ -0,0 +1,418 @@
+"""
+Framework Analyzer Module
+=========================
+
+Detects programming languages, frameworks, and related technologies across different ecosystems.
+Supports Python, Node.js/TypeScript, Go, Rust, and Ruby frameworks.
+"""
+
+from __future__ import annotations
+
+from pathlib import Path
+from typing import Any
+
+from .base import BaseAnalyzer
+
+
+class FrameworkAnalyzer(BaseAnalyzer):
+    """Analyzes and detects programming languages and frameworks."""
+
+    def __init__(self, path: Path, analysis: dict[str, Any]):
+        super().__init__(path)
+        self.analysis = analysis
+
+    def detect_language_and_framework(self) -> None:
+        """Detect primary language and framework."""
+        # Python detection
+        if self._exists("requirements.txt"):
+            self.analysis["language"] = "Python"
+            self.analysis["package_manager"] = "pip"
+            deps = self._read_file("requirements.txt")
+            self._detect_python_framework(deps)
+
+        elif self._exists("pyproject.toml"):
+            self.analysis["language"] = "Python"
+            content = self._read_file("pyproject.toml")
+            if "[tool.poetry]" in content:
+                self.analysis["package_manager"] = "poetry"
+            elif "[tool.uv]" in content:
+                self.analysis["package_manager"] = "uv"
+            else:
+                self.analysis["package_manager"] = "pip"
+            self._detect_python_framework(content)
+
+        elif self._exists("Pipfile"):
+            self.analysis["language"] = "Python"
+            self.analysis["package_manager"] = "pipenv"
+            content = self._read_file("Pipfile")
+            self._detect_python_framework(content)
+
+        # Node.js/TypeScript detection
+        elif self._exists("package.json"):
+            pkg = self._read_json("package.json")
+            if pkg:
+                # Check if TypeScript
+                deps = {**pkg.get("dependencies", {}), **pkg.get("devDependencies", {})}
+                if "typescript" in deps:
+                    self.analysis["language"] = "TypeScript"
+                else:
+                    self.analysis["language"] = "JavaScript"
+
+                self.analysis["package_manager"] = self._detect_node_package_manager()
+                self._detect_node_framework(pkg)
+
+        # Go detection
+        elif self._exists("go.mod"):
+            self.analysis["language"] = "Go"
+            self.analysis["package_manager"] = "go mod"
+            content = self._read_file("go.mod")
+            self._detect_go_framework(content)
+
+        # Rust detection
+        elif self._exists("Cargo.toml"):
+            self.analysis["language"] = "Rust"
+            self.analysis["package_manager"] = "cargo"
+            content = self._read_file("Cargo.toml")
+            self._detect_rust_framework(content)
+
+        # Swift/iOS detection (check BEFORE Ruby - iOS projects often have Gemfile for CocoaPods/Fastlane)
+        elif self._exists("Package.swift") or any(self.path.glob("*.xcodeproj")):
+            self.analysis["language"] = "Swift"
+            if self._exists("Package.swift"):
+                self.analysis["package_manager"] = "Swift Package Manager"
+            else:
+                self.analysis["package_manager"] = "Xcode"
+            self._detect_swift_framework()
+
+        # Ruby detection
+        elif self._exists("Gemfile"):
+            self.analysis["language"] = "Ruby"
+            self.analysis["package_manager"] = "bundler"
+            content = self._read_file("Gemfile")
+            self._detect_ruby_framework(content)
+
+    def _detect_python_framework(self, content: str) -> None:
+        """Detect Python framework."""
+        from .port_detector import PortDetector
+
+        content_lower = content.lower()
+
+        # Web frameworks (with conventional defaults)
+        frameworks = {
+            "fastapi": {"name": "FastAPI", "type": "backend", "port": 8000},
+            "flask": {"name": "Flask", "type": "backend", "port": 5000},
+            "django": {"name": "Django", "type": "backend", "port": 8000},
+            "starlette": {"name": "Starlette", "type": "backend", "port": 8000},
+            "litestar": {"name": "Litestar", "type": "backend", "port": 8000},
+        }
+
+        for key, info in frameworks.items():
+            if key in content_lower:
+                self.analysis["framework"] = info["name"]
+                self.analysis["type"] = info["type"]
+                # Try to detect actual port, fall back to default
+                port_detector = PortDetector(self.path, self.analysis)
+                detected_port = port_detector.detect_port_from_sources(info["port"])
+                self.analysis["default_port"] = detected_port
+                break
+
+        # Task queues
+        if "celery" in content_lower:
+            self.analysis["task_queue"] = "Celery"
+            if not self.analysis.get("type"):
+                self.analysis["type"] = "worker"
+        elif "dramatiq" in content_lower:
+            self.analysis["task_queue"] = "Dramatiq"
+        elif "huey" in content_lower:
+            self.analysis["task_queue"] = "Huey"
+
+        # ORM
+        if "sqlalchemy" in content_lower:
+            self.analysis["orm"] = "SQLAlchemy"
+        elif "tortoise" in content_lower:
+            self.analysis["orm"] = "Tortoise ORM"
+        elif "prisma" in content_lower:
+            self.analysis["orm"] = "Prisma"
+
+    def _detect_node_framework(self, pkg: dict) -> None:
+        """Detect Node.js/TypeScript framework."""
+        from .port_detector import PortDetector
+
+        deps = {**pkg.get("dependencies", {}), **pkg.get("devDependencies", {})}
+        deps_lower = {k.lower(): k for k in deps.keys()}
+
+        # Frontend frameworks
+        frontend_frameworks = {
+            "next": {"name": "Next.js", "type": "frontend", "port": 3000},
+            "nuxt": {"name": "Nuxt", "type": "frontend", "port": 3000},
+            "react": {"name": "React", "type": "frontend", "port": 3000},
+            "vue": {"name": "Vue", "type": "frontend", "port": 5173},
+            "svelte": {"name": "Svelte", "type": "frontend", "port": 5173},
+            "@sveltejs/kit": {"name": "SvelteKit", "type": "frontend", "port": 5173},
+            "angular": {"name": "Angular", "type": "frontend", "port": 4200},
+            "@angular/core": {"name": "Angular", "type": "frontend", "port": 4200},
+            "solid-js": {"name": "SolidJS", "type": "frontend", "port": 3000},
+            "astro": {"name": "Astro", "type": "frontend", "port": 4321},
+        }
+
+        # Backend frameworks
+        backend_frameworks = {
+            "express": {"name": "Express", "type": "backend", "port": 3000},
+            "fastify": {"name": "Fastify", "type": "backend", "port": 3000},
+            "koa": {"name": "Koa", "type": "backend", "port": 3000},
+            "hono": {"name": "Hono", "type": "backend", "port": 3000},
+            "elysia": {"name": "Elysia", "type": "backend", "port": 3000},
+            "@nestjs/core": {"name": "NestJS", "type": "backend", "port": 3000},
+        }
+
+        port_detector = PortDetector(self.path, self.analysis)
+
+        # Check frontend first (Next.js includes React, etc.)
+        for key, info in frontend_frameworks.items():
+            if key in deps_lower:
+                self.analysis["framework"] = info["name"]
+                self.analysis["type"] = info["type"]
+                detected_port = port_detector.detect_port_from_sources(info["port"])
+                self.analysis["default_port"] = detected_port
+                break
+
+        # If no frontend, check backend
+        if not self.analysis.get("framework"):
+            for key, info in backend_frameworks.items():
+                if key in deps_lower:
+                    self.analysis["framework"] = info["name"]
+                    self.analysis["type"] = info["type"]
+                    detected_port = port_detector.detect_port_from_sources(info["port"])
+                    self.analysis["default_port"] = detected_port
+                    break
+
+        # Build tool
+        if "vite" in deps_lower:
+            self.analysis["build_tool"] = "Vite"
+            if not self.analysis.get("default_port"):
+                detected_port = port_detector.detect_port_from_sources(5173)
+                self.analysis["default_port"] = detected_port
+        elif "webpack" in deps_lower:
+            self.analysis["build_tool"] = "Webpack"
+        elif "esbuild" in deps_lower:
+            self.analysis["build_tool"] = "esbuild"
+        elif "turbopack" in deps_lower:
+            self.analysis["build_tool"] = "Turbopack"
+
+        # Styling
+        if "tailwindcss" in deps_lower:
+            self.analysis["styling"] = "Tailwind CSS"
+        elif "styled-components" in deps_lower:
+            self.analysis["styling"] = "styled-components"
+        elif "@emotion/react" in deps_lower:
+            self.analysis["styling"] = "Emotion"
+
+        # State management
+        if "zustand" in deps_lower:
+            self.analysis["state_management"] = "Zustand"
+        elif "@reduxjs/toolkit" in deps_lower or "redux" in deps_lower:
+            self.analysis["state_management"] = "Redux"
+        elif "jotai" in deps_lower:
+            self.analysis["state_management"] = "Jotai"
+        elif "pinia" in deps_lower:
+            self.analysis["state_management"] = "Pinia"
+
+        # Task queues
+        if "bullmq" in deps_lower or "bull" in deps_lower:
+            self.analysis["task_queue"] = "BullMQ"
+            if not self.analysis.get("type"):
+                self.analysis["type"] = "worker"
+
+        # ORM
+        if "@prisma/client" in deps_lower or "prisma" in deps_lower:
+            self.analysis["orm"] = "Prisma"
+        elif "typeorm" in deps_lower:
+            self.analysis["orm"] = "TypeORM"
+        elif "drizzle-orm" in deps_lower:
+            self.analysis["orm"] = "Drizzle"
+        elif "mongoose" in deps_lower:
+            self.analysis["orm"] = "Mongoose"
+
+        # Scripts
+        scripts = pkg.get("scripts", {})
+        pkg_mgr = self.analysis.get("package_manager", "npm")
+        if "dev" in scripts:
+            self.analysis["dev_command"] = f"{pkg_mgr} run dev"
+        elif "start" in scripts:
+            self.analysis["dev_command"] = f"{pkg_mgr} run start"
+
+        # Capture available scripts for downstream consumers (QA agents, init.sh)
+        if scripts:
+            self.analysis["scripts"] = dict(scripts)
+
+    def _detect_go_framework(self, content: str) -> None:
+        """Detect Go framework."""
+        from .port_detector import PortDetector
+
+        frameworks = {
+            "gin-gonic/gin": {"name": "Gin", "port": 8080},
+            "labstack/echo": {"name": "Echo", "port": 8080},
+            "gofiber/fiber": {"name": "Fiber", "port": 3000},
+            "go-chi/chi": {"name": "Chi", "port": 8080},
+        }
+
+        for key, info in frameworks.items():
+            if key in content:
+                self.analysis["framework"] = info["name"]
+                self.analysis["type"] = "backend"
+                port_detector = PortDetector(self.path, self.analysis)
+                detected_port = port_detector.detect_port_from_sources(info["port"])
+                self.analysis["default_port"] = detected_port
+                break
+
+    def _detect_rust_framework(self, content: str) -> None:
+        """Detect Rust framework."""
+        from .port_detector import PortDetector
+
+        frameworks = {
+            "actix-web": {"name": "Actix Web", "port": 8080},
+            "axum": {"name": "Axum", "port": 3000},
+            "rocket": {"name": "Rocket", "port": 8000},
+        }
+
+        for key, info in frameworks.items():
+            if key in content:
+                self.analysis["framework"] = info["name"]
+                self.analysis["type"] = "backend"
+                port_detector = PortDetector(self.path, self.analysis)
+                detected_port = port_detector.detect_port_from_sources(info["port"])
+                self.analysis["default_port"] = detected_port
+                break
+
+    def _detect_ruby_framework(self, content: str) -> None:
+        """Detect Ruby framework."""
+        from .port_detector import PortDetector
+
+        port_detector = PortDetector(self.path, self.analysis)
+
+        if "rails" in content.lower():
+            self.analysis["framework"] = "Ruby on Rails"
+            self.analysis["type"] = "backend"
+            detected_port = port_detector.detect_port_from_sources(3000)
+            self.analysis["default_port"] = detected_port
+        elif "sinatra" in content.lower():
+            self.analysis["framework"] = "Sinatra"
+            self.analysis["type"] = "backend"
+            detected_port = port_detector.detect_port_from_sources(4567)
+            self.analysis["default_port"] = detected_port
+
+        if "sidekiq" in content.lower():
+            self.analysis["task_queue"] = "Sidekiq"
+
+    def _detect_swift_framework(self) -> None:
+        """Detect Swift/iOS framework and dependencies."""
+        try:
+            # Scan Swift files for imports, excluding hidden/vendor dirs
+            swift_files = []
+            for swift_file in self.path.rglob("*.swift"):
+                # Skip hidden directories, node_modules, .worktrees, etc.
+                if any(
+                    part.startswith(".") or part in ("node_modules", "Pods", "Carthage")
+                    for part in swift_file.parts
+                ):
+                    continue
+                swift_files.append(swift_file)
+                if len(swift_files) >= 50:  # Limit for performance
+                    break
+
+            imports = set()
+            for swift_file in swift_files:
+                try:
+                    content = swift_file.read_text(encoding="utf-8", errors="ignore")
+                    for line in content.split("\n"):
+                        line = line.strip()
+                        if line.startswith("import "):
+                            module = line.replace("import ", "").split()[0]
+                            imports.add(module)
+                except Exception:
+                    continue
+
+            # Detect UI framework
+            if "SwiftUI" in imports:
+                self.analysis["framework"] = "SwiftUI"
+                self.analysis["type"] = "mobile"
+            elif "UIKit" in imports:
+                self.analysis["framework"] = "UIKit"
+                self.analysis["type"] = "mobile"
+            elif "AppKit" in imports:
+                self.analysis["framework"] = "AppKit"
+                self.analysis["type"] = "desktop"
+
+            # Detect iOS/Apple frameworks
+            apple_frameworks = []
+            framework_map = {
+                "Combine": "Combine",
+                "CoreData": "CoreData",
+                "MapKit": "MapKit",
+                "WidgetKit": "WidgetKit",
+                "CoreLocation": "CoreLocation",
+                "StoreKit": "StoreKit",
+                "CloudKit": "CloudKit",
+                "ActivityKit": "ActivityKit",
+                "UserNotifications": "UserNotifications",
+            }
+            for key, name in framework_map.items():
+                if key in imports:
+                    apple_frameworks.append(name)
+
+            if apple_frameworks:
+                self.analysis["apple_frameworks"] = apple_frameworks
+
+            # Detect SPM dependencies from Package.swift or xcodeproj
+            dependencies = self._detect_spm_dependencies()
+            if dependencies:
+                self.analysis["spm_dependencies"] = dependencies
+        except Exception:
+            # Silently fail if Swift detection has issues
+            pass
+
+    def _detect_spm_dependencies(self) -> list[str]:
+        """Detect Swift Package Manager dependencies."""
+        dependencies = []
+
+        # Try Package.swift first
+        if self._exists("Package.swift"):
+            content = self._read_file("Package.swift")
+            # Look for .package(url: "...", patterns
+            import re
+
+            urls = re.findall(r'\.package\s*\([^)]*url:\s*"([^"]+)"', content)
+            for url in urls:
+                # Extract package name from URL
+                name = url.rstrip("/").split("/")[-1].replace(".git", "")
+                if name:
+                    dependencies.append(name)
+
+        # Also check xcodeproj for XCRemoteSwiftPackageReference
+        for xcodeproj in self.path.glob("*.xcodeproj"):
+            pbxproj = xcodeproj / "project.pbxproj"
+            if pbxproj.exists():
+                try:
+                    content = pbxproj.read_text(encoding="utf-8", errors="ignore")
+                    import re
+
+                    # Match repositoryURL patterns
+                    urls = re.findall(r'repositoryURL\s*=\s*"([^"]+)"', content)
+                    for url in urls:
+                        name = url.rstrip("/").split("/")[-1].replace(".git", "")
+                        if name and name not in dependencies:
+                            dependencies.append(name)
+                except Exception:
+                    continue
+
+        return dependencies
+
+    def _detect_node_package_manager(self) -> str:
+        """Detect Node.js package manager."""
+        if self._exists("pnpm-lock.yaml"):
+            return "pnpm"
+        elif self._exists("yarn.lock"):
+            return "yarn"
+        elif self._exists("bun.lockb") or self._exists("bun.lock"):
+            return "bun"
+        return "npm"
@@ -0,0 +1,337 @@
+"""
+Port Detector Module
+====================
+
+Detects application ports from multiple sources including entry points,
+environment files, Docker Compose, configuration files, and scripts.
+"""
+
+from __future__ import annotations
+
+import re
+from pathlib import Path
+from typing import Any
+
+from .base import BaseAnalyzer
+
+
+class PortDetector(BaseAnalyzer):
+    """Detects application ports from various configuration sources."""
+
+    def __init__(self, path: Path, analysis: dict[str, Any]):
+        super().__init__(path)
+        self.analysis = analysis
+
+    def detect_port_from_sources(self, default_port: int) -> int:
+        """
+        Robustly detect the actual port by checking multiple sources.
+
+        Checks in order of priority:
+        1. Entry point files (app.py, main.py, etc.) for uvicorn.run(), app.run(), etc.
+        2. Environment files (.env, .env.local, .env.development)
+        3. Docker Compose port mappings
+        4. Configuration files (config.py, settings.py, etc.)
+        5. Package.json scripts (for Node.js)
+        6. Makefile/shell scripts
+        7. Falls back to default_port if nothing found
+
+        Args:
+            default_port: The framework's conventional default port
+
+        Returns:
+            Detected port or default_port if not found
+        """
+        # 1. Check entry point files for explicit port definitions
+        port = self._detect_port_in_entry_points()
+        if port:
+            return port
+
+        # 2. Check environment files
+        port = self._detect_port_in_env_files()
+        if port:
+            return port
+
+        # 3. Check Docker Compose
+        port = self._detect_port_in_docker_compose()
+        if port:
+            return port
+
+        # 4. Check configuration files
+        port = self._detect_port_in_config_files()
+        if port:
+            return port
+
+        # 5. Check package.json scripts (for Node.js)
+        if self.analysis.get("language") in ["JavaScript", "TypeScript"]:
+            port = self._detect_port_in_package_scripts()
+            if port:
+                return port
+
+        # 6. Check Makefile/shell scripts
+        port = self._detect_port_in_scripts()
+        if port:
+            return port
+
+        # Fall back to default
+        return default_port
+
+    def _detect_port_in_entry_points(self) -> int | None:
+        """Detect port in entry point files."""
+        entry_files = [
+            "app.py",
+            "main.py",
+            "server.py",
+            "__main__.py",
+            "asgi.py",
+            "wsgi.py",
+            "src/app.py",
+            "src/main.py",
+            "src/server.py",
+            "index.js",
+            "index.ts",
+            "server.js",
+            "server.ts",
+            "main.js",
+            "main.ts",
+            "src/index.js",
+            "src/index.ts",
+            "src/server.js",
+            "src/server.ts",
+            "main.go",
+            "cmd/main.go",
+            "src/main.rs",
+        ]
+
+        # Patterns to search for ports
+        patterns = [
+            # Python: uvicorn.run(app, host="0.0.0.0", port=8050)
+            r"uvicorn\.run\([^)]*port\s*=\s*(\d+)",
+            # Python: app.run(port=8050, host="0.0.0.0")
+            r"\.run\([^)]*port\s*=\s*(\d+)",
+            # Python: port = 8050 or PORT = 8050
+            r"^\s*[Pp][Oo][Rr][Tt]\s*=\s*(\d+)",
+            # Python: os.getenv("PORT", 8050) or os.environ.get("PORT", 8050)
+            r'getenv\(\s*["\']PORT["\']\s*,\s*(\d+)',
+            r'environ\.get\(\s*["\']PORT["\']\s*,\s*(\d+)',
+            # JavaScript/TypeScript: app.listen(8050)
+            r"\.listen\(\s*(\d+)",
+            # JavaScript/TypeScript: const PORT = 8050 or let port = 8050
+            r"(?:const|let|var)\s+[Pp][Oo][Rr][Tt]\s*=\s*(\d+)",
+            # JavaScript/TypeScript: process.env.PORT || 8050
+            r"process\.env\.PORT\s*\|\|\s*(\d+)",
+            # JavaScript/TypeScript: Number(process.env.PORT) || 8050
+            r"Number\(process\.env\.PORT\)\s*\|\|\s*(\d+)",
+            # Go: :8050 or ":8050"
+            r':\s*(\d+)(?:["\s]|$)',
+            # Rust: .bind("127.0.0.1:8050")
+            r'\.bind\(["\'][\d.]+:(\d+)',
+        ]
+
+        for entry_file in entry_files:
+            content = self._read_file(entry_file)
+            if not content:
+                continue
+
+            for pattern in patterns:
+                matches = re.findall(pattern, content, re.MULTILINE)
+                if matches:
+                    # Return the first valid port found
+                    for match in matches:
+                        try:
+                            port = int(match)
+                            if 1000 <= port <= 65535:  # Valid port range
+                                return port
+                        except ValueError:
+                            continue
+
+        return None
+
+    def _detect_port_in_env_files(self) -> int | None:
+        """Detect port in environment files."""
+        env_files = [
+            ".env",
+            ".env.local",
+            ".env.development",
+            ".env.dev",
+            "config/.env",
+            "config/.env.local",
+            "../.env",
+        ]
+
+        patterns = [
+            r"^\s*PORT\s*=\s*(\d+)",
+            r"^\s*API_PORT\s*=\s*(\d+)",
+            r"^\s*SERVER_PORT\s*=\s*(\d+)",
+            r"^\s*APP_PORT\s*=\s*(\d+)",
+        ]
+
+        for env_file in env_files:
+            content = self._read_file(env_file)
+            if not content:
+                continue
+
+            for pattern in patterns:
+                matches = re.findall(pattern, content, re.MULTILINE)
+                if matches:
+                    try:
+                        port = int(matches[0])
+                        if 1000 <= port <= 65535:
+                            return port
+                    except ValueError:
+                        continue
+
+        return None
+
+    def _detect_port_in_docker_compose(self) -> int | None:
+        """Detect port from docker-compose.yml mappings."""
+        compose_files = [
+            "docker-compose.yml",
+            "docker-compose.yaml",
+            "../docker-compose.yml",
+            "../docker-compose.yaml",
+        ]
+
+        service_name = self.path.name.lower()
+
+        for compose_file in compose_files:
+            content = self._read_file(compose_file)
+            if not content:
+                continue
+
+            # Look for port mappings like "8050:8000" or "8050:8050"
+            # Match the service name if possible
+            pattern = r'^\s*-\s*["\']?(\d+):\d+["\']?'
+
+            in_service = False
+            in_ports = False
+
+            for line in content.split("\n"):
+                # Check if we're in the right service block
+                if re.match(rf"^\s*{re.escape(service_name)}\s*:", line):
+                    in_service = True
+                    continue
+
+                # Check if we hit another service
+                if (
+                    in_service
+                    and re.match(r"^\s*\w+\s*:", line)
+                    and "ports:" not in line
+                ):
+                    in_service = False
+                    in_ports = False
+                    continue
+
+                # Check if we're in the ports section
+                if in_service and "ports:" in line:
+                    in_ports = True
+                    continue
+
+                # Extract port mapping
+                if in_ports:
+                    match = re.match(pattern, line)
+                    if match:
+                        try:
+                            port = int(match.group(1))
+                            if 1000 <= port <= 65535:
+                                return port
+                        except ValueError:
+                            continue
+
+        return None
+
+    def _detect_port_in_config_files(self) -> int | None:
+        """Detect port in configuration files."""
+        config_files = [
+            "config.py",
+            "settings.py",
+            "config/settings.py",
+            "src/config.py",
+            "config.json",
+            "settings.json",
+            "config/config.json",
+            "config.toml",
+            "settings.toml",
+        ]
+
+        for config_file in config_files:
+            content = self._read_file(config_file)
+            if not content:
+                continue
+
+            # Python config patterns
+            patterns = [
+                r"[Pp][Oo][Rr][Tt]\s*=\s*(\d+)",
+                r'["\']port["\']\s*:\s*(\d+)',
+            ]
+
+            for pattern in patterns:
+                matches = re.findall(pattern, content)
+                if matches:
+                    try:
+                        port = int(matches[0])
+                        if 1000 <= port <= 65535:
+                            return port
+                    except ValueError:
+                        continue
+
+        return None
+
+    def _detect_port_in_package_scripts(self) -> int | None:
+        """Detect port in package.json scripts."""
+        pkg = self._read_json("package.json")
+        if not pkg:
+            return None
+
+        scripts = pkg.get("scripts", {})
+
+        # Look for port specifications in scripts
+        # e.g., "dev": "next dev -p 3001"
+        # e.g., "start": "node server.js --port 8050"
+        patterns = [
+            r"-p\s+(\d+)",
+            r"--port\s+(\d+)",
+            r"PORT=(\d+)",
+        ]
+
+        for script in scripts.values():
+            if not isinstance(script, str):
+                continue
+
+            for pattern in patterns:
+                matches = re.findall(pattern, script)
+                if matches:
+                    try:
+                        port = int(matches[0])
+                        if 1000 <= port <= 65535:
+                            return port
+                    except ValueError:
+                        continue
+
+        return None
+
+    def _detect_port_in_scripts(self) -> int | None:
+        """Detect port in Makefile or shell scripts."""
+        script_files = ["Makefile", "start.sh", "run.sh", "dev.sh"]
+
+        patterns = [
+            r"PORT=(\d+)",
+            r"--port\s+(\d+)",
+            r"-p\s+(\d+)",
+        ]
+
+        for script_file in script_files:
+            content = self._read_file(script_file)
+            if not content:
+                continue
+
+            for pattern in patterns:
+                matches = re.findall(pattern, content)
+                if matches:
+                    try:
+                        port = int(matches[0])
+                        if 1000 <= port <= 65535:
+                            return port
+                    except ValueError:
+                        continue
+
+        return None
@@ -0,0 +1,350 @@
+"""
+Project Analyzer Module
+=======================
+
+Analyzes entire projects, detecting monorepo structures, services, infrastructure, and conventions.
+"""
+
+from __future__ import annotations
+
+from pathlib import Path
+from typing import Any
+
+from .base import SERVICE_INDICATORS, SERVICE_ROOT_FILES, SKIP_DIRS
+from .service_analyzer import ServiceAnalyzer
+
+
+class ProjectAnalyzer:
+    """Analyzes an entire project, detecting monorepo structure and all services."""
+
+    def __init__(self, project_dir: Path):
+        self.project_dir = project_dir.resolve()
+        self.index = {
+            "project_root": str(self.project_dir),
+            "project_type": "single",  # or "monorepo"
+            "services": {},
+            "infrastructure": {},
+            "conventions": {},
+        }
+
+    def analyze(self) -> dict[str, Any]:
+        """Run full project analysis."""
+        self._detect_project_type()
+        self._find_and_analyze_services()
+        self._aggregate_dependency_locations()
+        self._analyze_infrastructure()
+        self._detect_conventions()
+        self._map_dependencies()
+        return self.index
+
+    def _detect_project_type(self) -> None:
+        """Detect if this is a monorepo or single project."""
+        monorepo_indicators = [
+            "pnpm-workspace.yaml",
+            "lerna.json",
+            "nx.json",
+            "turbo.json",
+            "rush.json",
+        ]
+
+        for indicator in monorepo_indicators:
+            if (self.project_dir / indicator).exists():
+                self.index["project_type"] = "monorepo"
+                self.index["monorepo_tool"] = indicator.replace(".json", "").replace(
+                    ".yaml", ""
+                )
+                return
+
+        # Check for packages/apps directories
+        if (self.project_dir / "packages").exists() or (
+            self.project_dir / "apps"
+        ).exists():
+            self.index["project_type"] = "monorepo"
+            return
+
+        # Check for multiple service directories
+        service_dirs_found = 0
+        for item in self.project_dir.iterdir():
+            if not item.is_dir():
+                continue
+            if item.name in SKIP_DIRS or item.name.startswith("."):
+                continue
+
+            # Check if this directory has service root files
+            if any((item / f).exists() for f in SERVICE_ROOT_FILES):
+                service_dirs_found += 1
+
+        # If we have 2+ directories with service root files, it's likely a monorepo
+        if service_dirs_found >= 2:
+            self.index["project_type"] = "monorepo"
+
+    def _find_and_analyze_services(self) -> None:
+        """Find all services and analyze each."""
+        services = {}
+
+        if self.index["project_type"] == "monorepo":
+            # Look for services in common locations
+            service_locations = [
+                self.project_dir,
+                self.project_dir / "packages",
+                self.project_dir / "apps",
+                self.project_dir / "services",
+            ]
+
+            for location in service_locations:
+                if not location.exists():
+                    continue
+
+                for item in location.iterdir():
+                    if not item.is_dir():
+                        continue
+                    if item.name in SKIP_DIRS:
+                        continue
+                    if item.name.startswith("."):
+                        continue
+
+                    # Check if this looks like a service
+                    has_root_file = any((item / f).exists() for f in SERVICE_ROOT_FILES)
+                    is_service_name = item.name.lower() in SERVICE_INDICATORS
+
+                    if has_root_file or (
+                        location == self.project_dir and is_service_name
+                    ):
+                        analyzer = ServiceAnalyzer(item, item.name)
+                        service_info = analyzer.analyze()
+                        if service_info.get(
+                            "language"
+                        ):  # Only include if we detected something
+                            services[item.name] = service_info
+        else:
+            # Single project - analyze root
+            analyzer = ServiceAnalyzer(self.project_dir, "main")
+            service_info = analyzer.analyze()
+            if service_info.get("language"):
+                services["main"] = service_info
+
+        self.index["services"] = services
+
+    def _aggregate_dependency_locations(self) -> None:
+        """Aggregate dependency location metadata from all services.
+
+        Collects dependency_locations from each service and stores them as
+        paths relative to the project root (e.g., 'apps/backend/.venv'
+        instead of just '.venv').
+        """
+        aggregated: list[dict[str, Any]] = []
+
+        for service_name, service_info in self.index.get("services", {}).items():
+            service_deps = service_info.get("dependency_locations", [])
+            service_path = service_info.get("path", "")
+
+            # Compute service-relative prefix once per service
+            service_rel: Path | None = None
+            if service_path:
+                try:
+                    service_rel = Path(service_path).relative_to(self.project_dir)
+                except ValueError:
+                    # Service path is outside the project root — skip its deps
+                    # to avoid producing absolute paths that bypass containment
+                    continue
+
+            for dep in service_deps:
+                dep_path = dep.get("path")
+                if not dep_path:
+                    continue
+
+                # Build project-relative path from service path + dep path
+                if service_rel is not None:
+                    project_relative = str(service_rel / dep_path)
+                else:
+                    project_relative = dep_path
+
+                entry: dict[str, Any] = {
+                    "type": dep.get("type", "unknown"),
+                    "path": project_relative,
+                    "exists": dep.get("exists", False),
+                    "service": service_name,
+                }
+                if dep.get("requirements_file"):
+                    # Convert to project-relative path like we do for "path"
+                    if service_rel is not None:
+                        entry["requirements_file"] = str(
+                            service_rel / dep["requirements_file"]
+                        )
+                    else:
+                        entry["requirements_file"] = dep["requirements_file"]
+                pkg_mgr = dep.get("package_manager") or service_info.get(
+                    "package_manager"
+                )
+                if pkg_mgr:
+                    entry["package_manager"] = pkg_mgr
+                aggregated.append(entry)
+
+        self.index["dependency_locations"] = aggregated
+
+    def _analyze_infrastructure(self) -> None:
+        """Analyze infrastructure configuration."""
+        infra = {}
+
+        # Docker
+        if (self.project_dir / "docker-compose.yml").exists():
+            infra["docker_compose"] = "docker-compose.yml"
+            compose_content = self._read_file("docker-compose.yml")
+            infra["docker_services"] = self._parse_compose_services(compose_content)
+        elif (self.project_dir / "docker-compose.yaml").exists():
+            infra["docker_compose"] = "docker-compose.yaml"
+            compose_content = self._read_file("docker-compose.yaml")
+            infra["docker_services"] = self._parse_compose_services(compose_content)
+
+        if (self.project_dir / "Dockerfile").exists():
+            infra["dockerfile"] = "Dockerfile"
+
+        # Docker directory
+        docker_dir = self.project_dir / "docker"
+        if docker_dir.exists():
+            dockerfiles = list(docker_dir.glob("Dockerfile*")) + list(
+                docker_dir.glob("*.Dockerfile")
+            )
+            if dockerfiles:
+                infra["docker_directory"] = "docker/"
+                infra["dockerfiles"] = [
+                    str(f.relative_to(self.project_dir)) for f in dockerfiles
+                ]
+
+        # CI/CD
+        if (self.project_dir / ".github" / "workflows").exists():
+            infra["ci"] = "GitHub Actions"
+            workflows = list((self.project_dir / ".github" / "workflows").glob("*.yml"))
+            infra["ci_workflows"] = [f.name for f in workflows]
+        elif (self.project_dir / ".gitlab-ci.yml").exists():
+            infra["ci"] = "GitLab CI"
+        elif (self.project_dir / ".circleci").exists():
+            infra["ci"] = "CircleCI"
+
+        # Deployment
+        deployment_files = {
+            "vercel.json": "Vercel",
+            "netlify.toml": "Netlify",
+            "fly.toml": "Fly.io",
+            "render.yaml": "Render",
+            "railway.json": "Railway",
+            "Procfile": "Heroku",
+            "app.yaml": "Google App Engine",
+            "serverless.yml": "Serverless Framework",
+        }
+
+        for file, platform in deployment_files.items():
+            if (self.project_dir / file).exists():
+                infra["deployment"] = platform
+                break
+
+        self.index["infrastructure"] = infra
+
+    def _parse_compose_services(self, content: str) -> list[str]:
+        """Extract service names from docker-compose content."""
+        services = []
+        in_services = False
+        for line in content.split("\n"):
+            if line.strip() == "services:":
+                in_services = True
+                continue
+            if in_services:
+                # Service names are at 2-space indent
+                if (
+                    line.startswith("  ")
+                    and not line.startswith("    ")
+                    and line.strip().endswith(":")
+                ):
+                    service_name = line.strip().rstrip(":")
+                    services.append(service_name)
+                elif line and not line.startswith(" "):
+                    break  # End of services section
+        return services
+
+    def _detect_conventions(self) -> None:
+        """Detect project-wide conventions."""
+        conventions = {}
+
+        # Python linting
+        if (self.project_dir / "ruff.toml").exists() or self._has_in_pyproject("ruff"):
+            conventions["python_linting"] = "Ruff"
+        elif (self.project_dir / ".flake8").exists():
+            conventions["python_linting"] = "Flake8"
+        elif (self.project_dir / "pylintrc").exists():
+            conventions["python_linting"] = "Pylint"
+
+        # Python formatting
+        if (self.project_dir / "pyproject.toml").exists():
+            content = self._read_file("pyproject.toml")
+            if "[tool.black]" in content:
+                conventions["python_formatting"] = "Black"
+
+        # JavaScript/TypeScript linting
+        eslint_files = [
+            ".eslintrc",
+            ".eslintrc.js",
+            ".eslintrc.json",
+            ".eslintrc.yml",
+            "eslint.config.js",
+        ]
+        if any((self.project_dir / f).exists() for f in eslint_files):
+            conventions["js_linting"] = "ESLint"
+
+        # Prettier
+        prettier_files = [
+            ".prettierrc",
+            ".prettierrc.js",
+            ".prettierrc.json",
+            "prettier.config.js",
+        ]
+        if any((self.project_dir / f).exists() for f in prettier_files):
+            conventions["formatting"] = "Prettier"
+
+        # TypeScript
+        if (self.project_dir / "tsconfig.json").exists():
+            conventions["typescript"] = True
+
+        # Git hooks
+        if (self.project_dir / ".husky").exists():
+            conventions["git_hooks"] = "Husky"
+        elif (self.project_dir / ".pre-commit-config.yaml").exists():
+            conventions["git_hooks"] = "pre-commit"
+
+        self.index["conventions"] = conventions
+
+    def _map_dependencies(self) -> None:
+        """Map dependencies between services."""
+        services = self.index.get("services", {})
+
+        for service_name, service_info in services.items():
+            consumes = []
+
+            # Check for API client patterns
+            if service_info.get("type") == "frontend":
+                # Frontend typically consumes backend
+                for other_name, other_info in services.items():
+                    if other_info.get("type") == "backend":
+                        consumes.append(f"{other_name}.api")
+
+            # Check for shared libraries
+            if service_info.get("dependencies"):
+                deps = service_info["dependencies"]
+                for other_name in services.keys():
+                    if other_name in deps or f"@{other_name}" in str(deps):
+                        consumes.append(other_name)
+
+            if consumes:
+                service_info["consumes"] = consumes
+
+    def _has_in_pyproject(self, tool: str) -> bool:
+        """Check if a tool is configured in pyproject.toml."""
+        if (self.project_dir / "pyproject.toml").exists():
+            content = self._read_file("pyproject.toml")
+            return f"[tool.{tool}]" in content
+        return False
+
+    def _read_file(self, path: str) -> str:
+        try:
+            return (self.project_dir / path).read_text(encoding="utf-8")
+        except (OSError, UnicodeDecodeError):
+            return ""
@@ -0,0 +1,418 @@
+"""
+Route Detector Module
+=====================
+
+Detects API routes and endpoints across different frameworks:
+- Python: FastAPI, Flask, Django
+- Node.js: Express, Next.js
+- Go: Gin, Echo, Chi, Fiber
+- Rust: Axum, Actix
+"""
+
+from __future__ import annotations
+
+import re
+from pathlib import Path
+
+from .base import BaseAnalyzer
+
+
+class RouteDetector(BaseAnalyzer):
+    """Detects API routes across multiple web frameworks."""
+
+    # Directories to exclude from route detection
+    EXCLUDED_DIRS = {"node_modules", ".venv", "venv", "__pycache__", ".git"}
+
+    def __init__(self, path: Path):
+        super().__init__(path)
+
+    def _should_include_file(self, file_path: Path) -> bool:
+        """Check if file should be included (not in excluded directories)."""
+        return not any(part in self.EXCLUDED_DIRS for part in file_path.parts)
+
+    def detect_all_routes(self) -> list[dict]:
+        """Detect all API routes across different frameworks."""
+        routes = []
+
+        # Python FastAPI
+        routes.extend(self._detect_fastapi_routes())
+
+        # Python Flask
+        routes.extend(self._detect_flask_routes())
+
+        # Python Django
+        routes.extend(self._detect_django_routes())
+
+        # Node.js Express/Fastify/Koa
+        routes.extend(self._detect_express_routes())
+
+        # Next.js (file-based routing)
+        routes.extend(self._detect_nextjs_routes())
+
+        # Go Gin/Echo/Chi
+        routes.extend(self._detect_go_routes())
+
+        # Rust Axum/Actix
+        routes.extend(self._detect_rust_routes())
+
+        return routes
+
+    def _detect_fastapi_routes(self) -> list[dict]:
+        """Detect FastAPI routes."""
+        routes = []
+        files_to_check = [
+            f for f in self.path.glob("**/*.py") if self._should_include_file(f)
+        ]
+
+        for file_path in files_to_check:
+            try:
+                content = file_path.read_text(encoding="utf-8")
+            except (OSError, UnicodeDecodeError):
+                continue
+
+            # Pattern: @app.get("/path") or @router.post("/path", dependencies=[...])
+            patterns = [
+                (
+                    r'@(?:app|router)\.(get|post|put|delete|patch)\(["\']([^"\']+)["\']',
+                    "decorator",
+                ),
+                (
+                    r'@(?:app|router)\.api_route\(["\']([^"\']+)["\'][^)]*methods\s*=\s*\[([^\]]+)\]',
+                    "api_route",
+                ),
+            ]
+
+            for pattern, pattern_type in patterns:
+                matches = re.finditer(pattern, content, re.MULTILINE)
+                for match in matches:
+                    if pattern_type == "decorator":
+                        method = match.group(1).upper()
+                        path = match.group(2)
+                        methods = [method]
+                    else:
+                        path = match.group(1)
+                        methods_str = match.group(2)
+                        methods = [
+                            m.strip().strip('"').strip("'").upper()
+                            for m in methods_str.split(",")
+                        ]
+
+                    # Check if route requires auth (has Depends in the decorator)
+                    line_start = content.rfind("\n", 0, match.start()) + 1
+                    line_end = content.find("\n", match.end())
+                    route_definition = content[
+                        line_start : line_end if line_end != -1 else len(content)
+                    ]
+
+                    requires_auth = (
+                        "Depends" in route_definition
+                        or "require" in route_definition.lower()
+                    )
+
+                    routes.append(
+                        {
+                            "path": path,
+                            "methods": methods,
+                            "file": str(file_path.relative_to(self.path)),
+                            "framework": "FastAPI",
+                            "requires_auth": requires_auth,
+                        }
+                    )
+
+        return routes
+
+    def _detect_flask_routes(self) -> list[dict]:
+        """Detect Flask routes."""
+        routes = []
+        files_to_check = [
+            f for f in self.path.glob("**/*.py") if self._should_include_file(f)
+        ]
+
+        for file_path in files_to_check:
+            try:
+                content = file_path.read_text(encoding="utf-8")
+            except (OSError, UnicodeDecodeError):
+                continue
+
+            # Pattern: @app.route("/path", methods=["GET", "POST"])
+            pattern = r'@(?:app|bp|blueprint)\.route\(["\']([^"\']+)["\'](?:[^)]*methods\s*=\s*\[([^\]]+)\])?'
+            matches = re.finditer(pattern, content, re.MULTILINE)
+
+            for match in matches:
+                path = match.group(1)
+                methods_str = match.group(2)
+
+                if methods_str:
+                    methods = [
+                        m.strip().strip('"').strip("'").upper()
+                        for m in methods_str.split(",")
+                    ]
+                else:
+                    methods = ["GET"]  # Flask default
+
+                # Check for @login_required decorator
+                decorator_start = content.rfind("@", 0, match.start())
+                decorator_section = content[decorator_start : match.end()]
+                requires_auth = (
+                    "login_required" in decorator_section
+                    or "require" in decorator_section.lower()
+                )
+
+                routes.append(
+                    {
+                        "path": path,
+                        "methods": methods,
+                        "file": str(file_path.relative_to(self.path)),
+                        "framework": "Flask",
+                        "requires_auth": requires_auth,
+                    }
+                )
+
+        return routes
+
+    def _detect_django_routes(self) -> list[dict]:
+        """Detect Django routes from urls.py files."""
+        routes = []
+        url_files = [
+            f for f in self.path.glob("**/urls.py") if self._should_include_file(f)
+        ]
+
+        for file_path in url_files:
+            try:
+                content = file_path.read_text(encoding="utf-8")
+            except (OSError, UnicodeDecodeError):
+                continue
+
+            # Pattern: path('users/<int:id>/', views.user_detail)
+            patterns = [
+                r'path\(["\']([^"\']+)["\']',
+                r're_path\([r]?["\']([^"\']+)["\']',
+            ]
+
+            for pattern in patterns:
+                matches = re.finditer(pattern, content)
+                for match in matches:
+                    path = match.group(1)
+
+                    routes.append(
+                        {
+                            "path": f"/{path}" if not path.startswith("/") else path,
+                            "methods": ["GET", "POST"],  # Django allows both by default
+                            "file": str(file_path.relative_to(self.path)),
+                            "framework": "Django",
+                            "requires_auth": False,  # Can't easily detect without middleware analysis
+                        }
+                    )
+
+        return routes
+
+    def _detect_express_routes(self) -> list[dict]:
+        """Detect Express/Fastify/Koa routes."""
+        routes = []
+        js_files = [
+            f for f in self.path.glob("**/*.js") if self._should_include_file(f)
+        ]
+        ts_files = [
+            f for f in self.path.glob("**/*.ts") if self._should_include_file(f)
+        ]
+        files_to_check = js_files + ts_files
+        for file_path in files_to_check:
+            try:
+                content = file_path.read_text(encoding="utf-8")
+            except (OSError, UnicodeDecodeError):
+                continue
+
+            # Pattern: app.get('/path', handler) or router.post('/path', middleware, handler)
+            pattern = (
+                r'(?:app|router)\.(get|post|put|delete|patch|use)\(["\']([^"\']+)["\']'
+            )
+            matches = re.finditer(pattern, content)
+
+            for match in matches:
+                method = match.group(1).upper()
+                path = match.group(2)
+
+                if method == "USE":
+                    # .use() is middleware, might be a route prefix
+                    continue
+
+                # Check for auth middleware in the route definition
+                line_start = content.rfind("\n", 0, match.start()) + 1
+                line_end = content.find("\n", match.end())
+                route_line = content[
+                    line_start : line_end if line_end != -1 else len(content)
+                ]
+
+                requires_auth = any(
+                    keyword in route_line.lower()
+                    for keyword in ["auth", "authenticate", "protect", "require"]
+                )
+
+                routes.append(
+                    {
+                        "path": path,
+                        "methods": [method],
+                        "file": str(file_path.relative_to(self.path)),
+                        "framework": "Express",
+                        "requires_auth": requires_auth,
+                    }
+                )
+
+        return routes
+
+    def _detect_nextjs_routes(self) -> list[dict]:
+        """Detect Next.js file-based routes."""
+        routes = []
+
+        # Next.js App Router (app directory)
+        app_dir = self.path / "app"
+        if app_dir.exists():
+            # Find all route.ts/js files
+            route_files = [
+                f
+                for f in app_dir.glob("**/route.{ts,js,tsx,jsx}")
+                if self._should_include_file(f)
+            ]
+            for route_file in route_files:
+                # Convert file path to route path
+                # app/api/users/[id]/route.ts -> /api/users/:id
+                relative_path = route_file.parent.relative_to(app_dir)
+                route_path = "/" + str(relative_path).replace("\\", "/")
+
+                # Convert [id] to :id
+                route_path = re.sub(r"\[([^\]]+)\]", r":\1", route_path)
+
+                try:
+                    content = route_file.read_text(encoding="utf-8")
+                    # Detect exported methods: export async function GET(request)
+                    methods = re.findall(
+                        r"export\s+(?:async\s+)?function\s+(GET|POST|PUT|DELETE|PATCH)",
+                        content,
+                    )
+
+                    if methods:
+                        routes.append(
+                            {
+                                "path": route_path,
+                                "methods": methods,
+                                "file": str(route_file.relative_to(self.path)),
+                                "framework": "Next.js",
+                                "requires_auth": "auth" in content.lower(),
+                            }
+                        )
+                except (OSError, UnicodeDecodeError):
+                    continue
+
+        # Next.js Pages Router (pages/api directory)
+        pages_api = self.path / "pages" / "api"
+        if pages_api.exists():
+            api_files = [
+                f
+                for f in pages_api.glob("**/*.{ts,js,tsx,jsx}")
+                if self._should_include_file(f)
+            ]
+            for api_file in api_files:
+                if api_file.name.startswith("_"):
+                    continue
+
+                # Convert file path to route
+                relative_path = api_file.relative_to(pages_api)
+                route_path = "/api/" + str(relative_path.with_suffix("")).replace(
+                    "\\", "/"
+                )
+
+                # Convert [id] to :id
+                route_path = re.sub(r"\[([^\]]+)\]", r":\1", route_path)
+
+                routes.append(
+                    {
+                        "path": route_path,
+                        "methods": [
+                            "GET",
+                            "POST",
+                        ],  # Next.js API routes handle all methods
+                        "file": str(api_file.relative_to(self.path)),
+                        "framework": "Next.js",
+                        "requires_auth": False,
+                    }
+                )
+
+        return routes
+
+    def _detect_go_routes(self) -> list[dict]:
+        """Detect Go framework routes (Gin, Echo, Chi, Fiber)."""
+        routes = []
+        go_files = [
+            f for f in self.path.glob("**/*.go") if self._should_include_file(f)
+        ]
+
+        for file_path in go_files:
+            try:
+                content = file_path.read_text(encoding="utf-8")
+            except (OSError, UnicodeDecodeError):
+                continue
+
+            # Gin: r.GET("/path", handler)
+            # Echo: e.POST("/path", handler)
+            # Chi: r.Get("/path", handler)
+            # Fiber: app.Get("/path", handler)
+            pattern = r'(?:r|e|app|router)\.(GET|POST|PUT|DELETE|PATCH|Get|Post|Put|Delete|Patch)\(["\']([^"\']+)["\']'
+            matches = re.finditer(pattern, content)
+
+            for match in matches:
+                method = match.group(1).upper()
+                path = match.group(2)
+
+                routes.append(
+                    {
+                        "path": path,
+                        "methods": [method],
+                        "file": str(file_path.relative_to(self.path)),
+                        "framework": "Go",
+                        "requires_auth": False,
+                    }
+                )
+
+        return routes
+
+    def _detect_rust_routes(self) -> list[dict]:
+        """Detect Rust framework routes (Axum, Actix)."""
+        routes = []
+        rust_files = [
+            f for f in self.path.glob("**/*.rs") if self._should_include_file(f)
+        ]
+
+        for file_path in rust_files:
+            try:
+                content = file_path.read_text(encoding="utf-8")
+            except (OSError, UnicodeDecodeError):
+                continue
+
+            # Axum: .route("/path", get(handler))
+            # Actix: web::get().to(handler)
+            patterns = [
+                r'\.route\(["\']([^"\']+)["\'],\s*(get|post|put|delete|patch)',
+                r"web::(get|post|put|delete|patch)\(\)",
+            ]
+
+            for pattern in patterns:
+                matches = re.finditer(pattern, content)
+                for match in matches:
+                    if len(match.groups()) == 2:
+                        path = match.group(1)
+                        method = match.group(2).upper()
+                    else:
+                        path = "/"  # Can't determine path from web:: syntax
+                        method = match.group(1).upper()
+
+                    routes.append(
+                        {
+                            "path": path,
+                            "methods": [method],
+                            "file": str(file_path.relative_to(self.path)),
+                            "framework": "Rust",
+                            "requires_auth": False,
+                        }
+                    )
+
+        return routes
@@ -0,0 +1,430 @@
+"""
+Service Analyzer Module
+=======================
+
+Main ServiceAnalyzer class that coordinates all analysis for a single service/package.
+Integrates framework detection, route analysis, database models, and context extraction.
+"""
+
+from __future__ import annotations
+
+import re
+from pathlib import Path
+from typing import Any
+
+from .base import BaseAnalyzer
+from .context_analyzer import ContextAnalyzer
+from .database_detector import DatabaseDetector
+from .framework_analyzer import FrameworkAnalyzer
+from .route_detector import RouteDetector
+
+
+class ServiceAnalyzer(BaseAnalyzer):
+    """Analyzes a single service/package within a project."""
+
+    def __init__(self, service_path: Path, service_name: str):
+        super().__init__(service_path)
+        self.name = service_name
+        self.analysis = {
+            "name": service_name,
+            "path": str(service_path),
+            "language": None,
+            "framework": None,
+            "type": None,  # backend, frontend, worker, library, etc.
+        }
+
+    def analyze(self) -> dict[str, Any]:
+        """Run full analysis on this service."""
+        self._detect_language_and_framework()
+        self._detect_service_type()
+        self._find_key_directories()
+        self._find_entry_points()
+        self._detect_dependencies()
+        self._detect_dependency_locations()
+        self._detect_package_manager()
+        self._detect_testing()
+        self._find_dockerfile()
+
+        # Comprehensive context extraction
+        self._detect_environment_variables()
+        self._detect_api_routes()
+        self._detect_database_models()
+        self._detect_external_services()
+        self._detect_auth_patterns()
+        self._detect_migrations()
+        self._detect_background_jobs()
+        self._detect_api_documentation()
+        self._detect_monitoring()
+
+        return self.analysis
+
+    def _detect_language_and_framework(self) -> None:
+        """Detect primary language and framework."""
+        framework_analyzer = FrameworkAnalyzer(self.path, self.analysis)
+        framework_analyzer.detect_language_and_framework()
+
+    def _detect_service_type(self) -> None:
+        """Infer service type from name and content if not already set."""
+        if self.analysis.get("type"):
+            return
+
+        name_lower = self.name.lower()
+
+        # Infer from name
+        if any(kw in name_lower for kw in ["frontend", "client", "web", "ui", "app"]):
+            self.analysis["type"] = "frontend"
+        elif any(kw in name_lower for kw in ["backend", "api", "server", "service"]):
+            self.analysis["type"] = "backend"
+        elif any(
+            kw in name_lower for kw in ["worker", "job", "queue", "task", "celery"]
+        ):
+            self.analysis["type"] = "worker"
+        elif any(kw in name_lower for kw in ["scraper", "crawler", "spider"]):
+            self.analysis["type"] = "scraper"
+        elif any(kw in name_lower for kw in ["proxy", "gateway", "router"]):
+            self.analysis["type"] = "proxy"
+        elif any(
+            kw in name_lower for kw in ["lib", "shared", "common", "core", "utils"]
+        ):
+            self.analysis["type"] = "library"
+        else:
+            # Try to infer from language and content if name doesn't match
+            language = self.analysis.get("language")
+
+            if language == "Python":
+                # Check if it's a CLI tool, framework, or backend service
+                has_run_py = (self.path / "run.py").exists()
+                has_main_py = (self.path / "main.py").exists()
+                has_main_module = (self.path / "__main__.py").exists()
+
+                # Check for agent/automation framework patterns
+                has_agent_files = any(
+                    (self.path / f).exists()
+                    for f in ["agent.py", "agents", "runner.py", "runners"]
+                )
+
+                if has_run_py or has_main_py or has_main_module or has_agent_files:
+                    # It's a backend tool/framework/CLI
+                    self.analysis["type"] = "backend"
+                    return
+
+            # Default to unknown if no clear indicators
+            self.analysis["type"] = "unknown"
+
+    def _find_key_directories(self) -> None:
+        """Find important directories within this service."""
+        key_dirs = {}
+
+        # Common directory patterns
+        patterns = {
+            "src": "Source code",
+            "lib": "Library code",
+            "app": "Application code",
+            "api": "API endpoints",
+            "routes": "Route handlers",
+            "controllers": "Controllers",
+            "models": "Data models",
+            "schemas": "Schemas/DTOs",
+            "services": "Business logic",
+            "components": "UI components",
+            "pages": "Page components",
+            "views": "Views/templates",
+            "hooks": "Custom hooks",
+            "utils": "Utilities",
+            "helpers": "Helper functions",
+            "middleware": "Middleware",
+            "tests": "Tests",
+            "test": "Tests",
+            "__tests__": "Tests",
+            "config": "Configuration",
+            "tasks": "Background tasks",
+            "jobs": "Background jobs",
+            "workers": "Worker processes",
+        }
+
+        for dir_name, purpose in patterns.items():
+            dir_path = self.path / dir_name
+            if dir_path.exists() and dir_path.is_dir():
+                key_dirs[dir_name] = {
+                    "path": str(dir_path.relative_to(self.path)),
+                    "purpose": purpose,
+                }
+
+        if key_dirs:
+            self.analysis["key_directories"] = key_dirs
+
+    def _find_entry_points(self) -> None:
+        """Find main entry point files."""
+        entry_patterns = [
+            "main.py",
+            "app.py",
+            "__main__.py",
+            "server.py",
+            "wsgi.py",
+            "asgi.py",
+            "index.ts",
+            "index.js",
+            "main.ts",
+            "main.js",
+            "server.ts",
+            "server.js",
+            "app.ts",
+            "app.js",
+            "src/index.ts",
+            "src/index.js",
+            "src/main.ts",
+            "src/app.ts",
+            "src/server.ts",
+            "src/App.tsx",
+            "src/App.jsx",
+            "pages/_app.tsx",
+            "pages/_app.js",  # Next.js
+            "main.go",
+            "cmd/main.go",
+            "src/main.rs",
+            "src/lib.rs",
+        ]
+
+        for pattern in entry_patterns:
+            if self._exists(pattern):
+                self.analysis["entry_point"] = pattern
+                break
+
+    def _detect_dependencies(self) -> None:
+        """Extract key dependencies."""
+        if self._exists("package.json"):
+            pkg = self._read_json("package.json")
+            if pkg:
+                deps = pkg.get("dependencies", {})
+                dev_deps = pkg.get("devDependencies", {})
+                self.analysis["dependencies"] = list(deps.keys())[:20]  # Top 20
+                self.analysis["dev_dependencies"] = list(dev_deps.keys())[:10]
+
+        elif self._exists("requirements.txt"):
+            content = self._read_file("requirements.txt")
+            deps = []
+            for line in content.split("\n"):
+                line = line.strip()
+                if line and not line.startswith("#") and not line.startswith("-"):
+                    match = re.match(r"^([a-zA-Z0-9_-]+)", line)
+                    if match:
+                        deps.append(match.group(1))
+            self.analysis["dependencies"] = deps[:20]
+
+    def _detect_dependency_locations(self) -> None:
+        """Detect where dependencies live on disk for this service."""
+        locations: list[dict[str, Any]] = []
+
+        # Node.js: node_modules (only if package.json exists)
+        if self._exists("package.json"):
+            node_modules = self.path / "node_modules"
+            locations.append(
+                {
+                    "type": "node_modules",
+                    "path": "node_modules",
+                    "exists": node_modules.exists() and node_modules.is_dir(),
+                }
+            )
+
+        # Python: .venv or venv
+        for venv_dir in [".venv", "venv"]:
+            venv_path = self.path / venv_dir
+            if venv_path.exists() and venv_path.is_dir():
+                entry: dict[str, Any] = {
+                    "type": "venv",
+                    "path": venv_dir,
+                    "exists": True,
+                }
+                # Find requirements file
+                for req_file in ["requirements.txt", "pyproject.toml", "Pipfile"]:
+                    if self._exists(req_file):
+                        entry["requirements_file"] = req_file
+                        break
+                locations.append(entry)
+                break
+        else:
+            # No venv found, still record requirements file if present
+            for req_file in ["requirements.txt", "pyproject.toml", "Pipfile"]:
+                if self._exists(req_file):
+                    locations.append(
+                        {
+                            "type": "venv",
+                            "path": ".venv",
+                            "exists": False,
+                            "requirements_file": req_file,
+                        }
+                    )
+                    break
+
+        # PHP: vendor
+        vendor_path = self.path / "vendor"
+        if vendor_path.exists() and vendor_path.is_dir():
+            locations.append(
+                {
+                    "type": "vendor_php",
+                    "path": "vendor",
+                    "exists": True,
+                }
+            )
+
+        # Rust: target
+        target_path = self.path / "target"
+        if target_path.exists() and target_path.is_dir():
+            locations.append(
+                {
+                    "type": "cargo_target",
+                    "path": "target",
+                    "exists": True,
+                }
+            )
+
+        # Ruby: vendor/bundle
+        bundle_path = self.path / "vendor" / "bundle"
+        if bundle_path.exists() and bundle_path.is_dir():
+            locations.append(
+                {
+                    "type": "vendor_bundle",
+                    "path": "vendor/bundle",
+                    "exists": True,
+                }
+            )
+
+        self.analysis["dependency_locations"] = locations
+
+    def _detect_package_manager(self) -> None:
+        """Detect the package manager used by this service."""
+        # Node.js package managers
+        if self._exists("package-lock.json"):
+            self.analysis["package_manager"] = "npm"
+        elif self._exists("yarn.lock"):
+            self.analysis["package_manager"] = "yarn"
+        elif self._exists("pnpm-lock.yaml"):
+            self.analysis["package_manager"] = "pnpm"
+        elif self._exists("bun.lockb") or self._exists("bun.lock"):
+            self.analysis["package_manager"] = "bun"
+        # Python package managers
+        elif self._exists("Pipfile"):
+            self.analysis["package_manager"] = "pipenv"
+        elif self._exists("pyproject.toml"):
+            if self._exists("uv.lock"):
+                self.analysis["package_manager"] = "uv"
+            elif self._exists("poetry.lock"):
+                self.analysis["package_manager"] = "poetry"
+            else:
+                self.analysis["package_manager"] = "pip"
+        elif self._exists("requirements.txt"):
+            self.analysis["package_manager"] = "pip"
+        # Other
+        elif self._exists("Cargo.toml"):
+            self.analysis["package_manager"] = "cargo"
+        elif self._exists("go.mod"):
+            self.analysis["package_manager"] = "go_mod"
+        elif self._exists("Gemfile"):
+            self.analysis["package_manager"] = "gem"
+        elif self._exists("composer.json"):
+            self.analysis["package_manager"] = "composer"
+        else:
+            self.analysis["package_manager"] = None
+
+    def _detect_testing(self) -> None:
+        """Detect testing framework and configuration."""
+        if self._exists("package.json"):
+            pkg = self._read_json("package.json")
+            if pkg:
+                deps = {**pkg.get("dependencies", {}), **pkg.get("devDependencies", {})}
+                if "vitest" in deps:
+                    self.analysis["testing"] = "Vitest"
+                elif "jest" in deps:
+                    self.analysis["testing"] = "Jest"
+                if "@playwright/test" in deps:
+                    self.analysis["e2e_testing"] = "Playwright"
+                elif "cypress" in deps:
+                    self.analysis["e2e_testing"] = "Cypress"
+
+        elif self._exists("pytest.ini") or self._exists("pyproject.toml"):
+            self.analysis["testing"] = "pytest"
+
+        # Find test directory
+        for test_dir in ["tests", "test", "__tests__", "spec"]:
+            if self._exists(test_dir):
+                self.analysis["test_directory"] = test_dir
+                break
+
+    def _find_dockerfile(self) -> None:
+        """Find Dockerfile for this service."""
+        dockerfile_patterns = [
+            "Dockerfile",
+            f"Dockerfile.{self.name}",
+            f"docker/{self.name}.Dockerfile",
+            f"docker/Dockerfile.{self.name}",
+            "../docker/Dockerfile." + self.name,
+        ]
+
+        for pattern in dockerfile_patterns:
+            if self._exists(pattern):
+                self.analysis["dockerfile"] = pattern
+                break
+
+    def _detect_environment_variables(self) -> None:
+        """Detect environment variables."""
+        context = ContextAnalyzer(self.path, self.analysis)
+        context.detect_environment_variables()
+
+    def _detect_api_routes(self) -> None:
+        """Detect API routes."""
+        route_detector = RouteDetector(self.path)
+        routes = route_detector.detect_all_routes()
+
+        if routes:
+            self.analysis["api"] = {
+                "routes": routes,
+                "total_routes": len(routes),
+                "methods": list(
+                    set(method for r in routes for method in r.get("methods", []))
+                ),
+                "protected_routes": [
+                    r["path"] for r in routes if r.get("requires_auth")
+                ],
+            }
+
+    def _detect_database_models(self) -> None:
+        """Detect database models."""
+        db_detector = DatabaseDetector(self.path)
+        models = db_detector.detect_all_models()
+
+        if models:
+            self.analysis["database"] = {
+                "models": models,
+                "total_models": len(models),
+                "model_names": list(models.keys()),
+            }
+
+    def _detect_external_services(self) -> None:
+        """Detect external services."""
+        context = ContextAnalyzer(self.path, self.analysis)
+        context.detect_external_services()
+
+    def _detect_auth_patterns(self) -> None:
+        """Detect authentication patterns."""
+        context = ContextAnalyzer(self.path, self.analysis)
+        context.detect_auth_patterns()
+
+    def _detect_migrations(self) -> None:
+        """Detect database migrations."""
+        context = ContextAnalyzer(self.path, self.analysis)
+        context.detect_migrations()
+
+    def _detect_background_jobs(self) -> None:
+        """Detect background jobs."""
+        context = ContextAnalyzer(self.path, self.analysis)
+        context.detect_background_jobs()
+
+    def _detect_api_documentation(self) -> None:
+        """Detect API documentation."""
+        context = ContextAnalyzer(self.path, self.analysis)
+        context.detect_api_documentation()
+
+    def _detect_monitoring(self) -> None:
+        """Detect monitoring setup."""
+        context = ContextAnalyzer(self.path, self.analysis)
+        context.detect_monitoring()
@@ -0,0 +1,589 @@
+#!/usr/bin/env python3
+"""
+CI Discovery Module
+===================
+
+Parses CI/CD configuration files to extract test commands and workflows.
+Supports GitHub Actions, GitLab CI, CircleCI, and Jenkins.
+
+The CI discovery results are used by:
+- QA Agent: To understand existing CI test patterns
+- Validation Strategy: To match CI commands
+- Planner: To align verification with CI
+
+Usage:
+    from ci_discovery import CIDiscovery
+
+    discovery = CIDiscovery()
+    result = discovery.discover(project_dir)
+
+    if result:
+        print(f"CI System: {result.ci_system}")
+        print(f"Test Commands: {result.test_commands}")
+"""
+
+from __future__ import annotations
+
+import json
+import re
+from dataclasses import dataclass, field
+from pathlib import Path
+from typing import Any
+
+# Try to import yaml, fall back gracefully
+try:
+    import yaml
+
+    HAS_YAML = True
+except ImportError:
+    HAS_YAML = False
+
+
+# =============================================================================
+# DATA CLASSES
+# =============================================================================
+
+
+@dataclass
+class CIWorkflow:
+    """
+    Represents a CI workflow or job.
+
+    Attributes:
+        name: Name of the workflow/job
+        trigger: What triggers this workflow (push, pull_request, etc.)
+        steps: List of step names or commands
+        test_related: Whether this appears to be test-related
+    """
+
+    name: str
+    trigger: list[str] = field(default_factory=list)
+    steps: list[str] = field(default_factory=list)
+    test_related: bool = False
+
+
+@dataclass
+class CIConfig:
+    """
+    Result of CI configuration discovery.
+
+    Attributes:
+        ci_system: Name of CI system (github_actions, gitlab, circleci, jenkins)
+        config_files: List of CI config files found
+        test_commands: Extracted test commands by type
+        coverage_command: Coverage command if found
+        workflows: List of discovered workflows
+        environment_variables: Environment variables used
+    """
+
+    ci_system: str
+    config_files: list[str] = field(default_factory=list)
+    test_commands: dict[str, str] = field(default_factory=dict)
+    coverage_command: str | None = None
+    workflows: list[CIWorkflow] = field(default_factory=list)
+    environment_variables: list[str] = field(default_factory=list)
+
+
+# =============================================================================
+# CI PARSERS
+# =============================================================================
+
+
+class CIDiscovery:
+    """
+    Discovers CI/CD configurations in a project.
+
+    Analyzes:
+    - GitHub Actions (.github/workflows/*.yml)
+    - GitLab CI (.gitlab-ci.yml)
+    - CircleCI (.circleci/config.yml)
+    - Jenkins (Jenkinsfile)
+    """
+
+    def __init__(self) -> None:
+        """Initialize CI discovery."""
+        self._cache: dict[str, CIConfig | None] = {}
+
+    def discover(self, project_dir: Path) -> CIConfig | None:
+        """
+        Discover CI configuration in the project.
+
+        Args:
+            project_dir: Path to the project root
+
+        Returns:
+            CIConfig if CI found, None otherwise
+        """
+        project_dir = Path(project_dir)
+        cache_key = str(project_dir.resolve())
+
+        if cache_key in self._cache:
+            return self._cache[cache_key]
+
+        # Try each CI system
+        result = None
+
+        # GitHub Actions
+        github_workflows = project_dir / ".github" / "workflows"
+        if github_workflows.exists():
+            result = self._parse_github_actions(github_workflows)
+
+        # GitLab CI
+        if not result:
+            gitlab_ci = project_dir / ".gitlab-ci.yml"
+            if gitlab_ci.exists():
+                result = self._parse_gitlab_ci(gitlab_ci)
+
+        # CircleCI
+        if not result:
+            circleci = project_dir / ".circleci" / "config.yml"
+            if circleci.exists():
+                result = self._parse_circleci(circleci)
+
+        # Jenkins
+        if not result:
+            jenkinsfile = project_dir / "Jenkinsfile"
+            if jenkinsfile.exists():
+                result = self._parse_jenkinsfile(jenkinsfile)
+
+        self._cache[cache_key] = result
+        return result
+
+    def _parse_github_actions(self, workflows_dir: Path) -> CIConfig:
+        """Parse GitHub Actions workflow files."""
+        result = CIConfig(ci_system="github_actions")
+
+        workflow_files = list(workflows_dir.glob("*.yml")) + list(
+            workflows_dir.glob("*.yaml")
+        )
+
+        for wf_file in workflow_files:
+            result.config_files.append(
+                str(wf_file.relative_to(workflows_dir.parent.parent))
+            )
+
+            try:
+                content = wf_file.read_text(encoding="utf-8")
+                workflow_data = self._parse_yaml(content)
+
+                if not workflow_data:
+                    continue
+
+                # Get workflow name
+                wf_name = workflow_data.get("name", wf_file.stem)
+
+                # Get triggers
+                triggers = []
+                on_trigger = workflow_data.get("on", {})
+                if isinstance(on_trigger, str):
+                    triggers = [on_trigger]
+                elif isinstance(on_trigger, list):
+                    triggers = on_trigger
+                elif isinstance(on_trigger, dict):
+                    triggers = list(on_trigger.keys())
+
+                # Parse jobs
+                jobs = workflow_data.get("jobs", {})
+                for job_name, job_config in jobs.items():
+                    if not isinstance(job_config, dict):
+                        continue
+
+                    steps = job_config.get("steps", [])
+                    step_commands = []
+                    test_related = False
+
+                    for step in steps:
+                        if not isinstance(step, dict):
+                            continue
+
+                        # Get step name or command
+                        step_name = step.get("name", "")
+                        run_cmd = step.get("run", "")
+                        uses = step.get("uses", "")
+
+                        if step_name:
+                            step_commands.append(step_name)
+                        if run_cmd:
+                            step_commands.append(run_cmd)
+                            # Extract test commands
+                            self._extract_test_commands(run_cmd, result)
+                        if uses:
+                            step_commands.append(f"uses: {uses}")
+
+                        # Check if test-related
+                        test_keywords = ["test", "pytest", "jest", "vitest", "coverage"]
+                        if any(kw in str(step).lower() for kw in test_keywords):
+                            test_related = True
+
+                    result.workflows.append(
+                        CIWorkflow(
+                            name=f"{wf_name}/{job_name}",
+                            trigger=triggers,
+                            steps=step_commands,
+                            test_related=test_related,
+                        )
+                    )
+
+                # Extract environment variables
+                env = workflow_data.get("env", {})
+                if isinstance(env, dict):
+                    result.environment_variables.extend(env.keys())
+
+            except Exception:
+                continue
+
+        return result
+
+    def _parse_gitlab_ci(self, config_file: Path) -> CIConfig:
+        """Parse GitLab CI configuration."""
+        result = CIConfig(
+            ci_system="gitlab",
+            config_files=[".gitlab-ci.yml"],
+        )
+
+        try:
+            content = config_file.read_text(encoding="utf-8")
+            data = self._parse_yaml(content)
+
+            if not data:
+                return result
+
+            # Parse jobs (top-level keys that aren't special keywords)
+            special_keys = {
+                "stages",
+                "variables",
+                "image",
+                "services",
+                "before_script",
+                "after_script",
+                "cache",
+                "include",
+                "default",
+                "workflow",
+            }
+
+            for key, value in data.items():
+                if key.startswith(".") or key in special_keys:
+                    continue
+
+                if not isinstance(value, dict):
+                    continue
+
+                job_config = value
+                script = job_config.get("script", [])
+                if isinstance(script, str):
+                    script = [script]
+
+                test_related = any(
+                    kw in str(script).lower()
+                    for kw in ["test", "pytest", "jest", "vitest", "coverage"]
+                )
+
+                result.workflows.append(
+                    CIWorkflow(
+                        name=key,
+                        trigger=job_config.get("only", [])
+                        or job_config.get("rules", []),
+                        steps=script,
+                        test_related=test_related,
+                    )
+                )
+
+                # Extract test commands
+                for cmd in script:
+                    if isinstance(cmd, str):
+                        self._extract_test_commands(cmd, result)
+
+            # Extract variables
+            variables = data.get("variables", {})
+            if isinstance(variables, dict):
+                result.environment_variables.extend(variables.keys())
+
+        except Exception:
+            pass
+
+        return result
+
+    def _parse_circleci(self, config_file: Path) -> CIConfig:
+        """Parse CircleCI configuration."""
+        result = CIConfig(
+            ci_system="circleci",
+            config_files=[".circleci/config.yml"],
+        )
+
+        try:
+            content = config_file.read_text(encoding="utf-8")
+            data = self._parse_yaml(content)
+
+            if not data:
+                return result
+
+            # Parse jobs
+            jobs = data.get("jobs", {})
+            for job_name, job_config in jobs.items():
+                if not isinstance(job_config, dict):
+                    continue
+
+                steps = job_config.get("steps", [])
+                step_commands = []
+                test_related = False
+
+                for step in steps:
+                    if isinstance(step, str):
+                        step_commands.append(step)
+                    elif isinstance(step, dict):
+                        if "run" in step:
+                            run = step["run"]
+                            if isinstance(run, str):
+                                step_commands.append(run)
+                                self._extract_test_commands(run, result)
+                            elif isinstance(run, dict):
+                                cmd = run.get("command", "")
+                                step_commands.append(cmd)
+                                self._extract_test_commands(cmd, result)
+
+                        if any(
+                            kw in str(step).lower()
+                            for kw in ["test", "pytest", "jest", "coverage"]
+                        ):
+                            test_related = True
+
+                result.workflows.append(
+                    CIWorkflow(
+                        name=job_name,
+                        trigger=[],
+                        steps=step_commands,
+                        test_related=test_related,
+                    )
+                )
+
+        except Exception:
+            pass
+
+        return result
+
+    def _parse_jenkinsfile(self, jenkinsfile: Path) -> CIConfig:
+        """Parse Jenkinsfile (basic extraction)."""
+        result = CIConfig(
+            ci_system="jenkins",
+            config_files=["Jenkinsfile"],
+        )
+
+        try:
+            content = jenkinsfile.read_text(encoding="utf-8")
+
+            # Extract sh commands using regex
+            sh_pattern = re.compile(r'sh\s+[\'"]([^\'"]+)[\'"]')
+            matches = sh_pattern.findall(content)
+
+            steps = []
+            test_related = False
+
+            for cmd in matches:
+                steps.append(cmd)
+                self._extract_test_commands(cmd, result)
+
+                if any(
+                    kw in cmd.lower() for kw in ["test", "pytest", "jest", "coverage"]
+                ):
+                    test_related = True
+
+            # Extract stage names
+            stage_pattern = re.compile(r'stage\s*\([\'"]([^\'"]+)[\'"]\)')
+            stages = stage_pattern.findall(content)
+
+            for stage in stages:
+                result.workflows.append(
+                    CIWorkflow(
+                        name=stage,
+                        trigger=[],
+                        steps=steps if "test" in stage.lower() else [],
+                        test_related="test" in stage.lower(),
+                    )
+                )
+
+        except Exception:
+            pass
+
+        return result
+
+    def _parse_yaml(self, content: str) -> dict | None:
+        """Parse YAML content, with fallback to basic parsing if yaml not available."""
+        if HAS_YAML:
+            try:
+                return yaml.safe_load(content)
+            except Exception:
+                return None
+
+        # Basic fallback for simple YAML (very limited)
+        # This won't work for complex structures
+        return None
+
+    def _extract_test_commands(self, cmd: str, result: CIConfig) -> None:
+        """Extract test commands from a command string."""
+        cmd_lower = cmd.lower()
+
+        # Python pytest
+        if "pytest" in cmd_lower:
+            if "pytest" not in result.test_commands:
+                result.test_commands["unit"] = cmd.strip()
+            if "--cov" in cmd_lower:
+                result.coverage_command = cmd.strip()
+
+        # Node.js test commands
+        if (
+            "npm test" in cmd_lower
+            or "yarn test" in cmd_lower
+            or "pnpm test" in cmd_lower
+        ):
+            if "unit" not in result.test_commands:
+                result.test_commands["unit"] = cmd.strip()
+
+        # Jest/Vitest
+        if "jest" in cmd_lower or "vitest" in cmd_lower:
+            if "unit" not in result.test_commands:
+                result.test_commands["unit"] = cmd.strip()
+            if "--coverage" in cmd_lower:
+                result.coverage_command = cmd.strip()
+
+        # E2E testing
+        if "playwright" in cmd_lower:
+            result.test_commands["e2e"] = cmd.strip()
+        if "cypress" in cmd_lower:
+            result.test_commands["e2e"] = cmd.strip()
+
+        # Integration tests
+        if "integration" in cmd_lower:
+            result.test_commands["integration"] = cmd.strip()
+
+        # Go tests
+        if "go test" in cmd_lower:
+            if "unit" not in result.test_commands:
+                result.test_commands["unit"] = cmd.strip()
+
+        # Rust tests
+        if "cargo test" in cmd_lower:
+            if "unit" not in result.test_commands:
+                result.test_commands["unit"] = cmd.strip()
+
+    def to_dict(self, result: CIConfig) -> dict[str, Any]:
+        """Convert result to dictionary for JSON serialization."""
+        return {
+            "ci_system": result.ci_system,
+            "config_files": result.config_files,
+            "test_commands": result.test_commands,
+            "coverage_command": result.coverage_command,
+            "workflows": [
+                {
+                    "name": w.name,
+                    "trigger": w.trigger,
+                    "steps": w.steps,
+                    "test_related": w.test_related,
+                }
+                for w in result.workflows
+            ],
+            "environment_variables": result.environment_variables,
+        }
+
+    def clear_cache(self) -> None:
+        """Clear the internal cache."""
+        self._cache.clear()
+
+
+# =============================================================================
+# CONVENIENCE FUNCTIONS
+# =============================================================================
+
+
+def discover_ci(project_dir: Path) -> CIConfig | None:
+    """
+    Convenience function to discover CI configuration.
+
+    Args:
+        project_dir: Path to project root
+
+    Returns:
+        CIConfig if found, None otherwise
+    """
+    discovery = CIDiscovery()
+    return discovery.discover(project_dir)
+
+
+def get_ci_test_commands(project_dir: Path) -> dict[str, str]:
+    """
+    Get test commands from CI configuration.
+
+    Args:
+        project_dir: Path to project root
+
+    Returns:
+        Dictionary of test type to command
+    """
+    discovery = CIDiscovery()
+    result = discovery.discover(project_dir)
+    if result:
+        return result.test_commands
+    return {}
+
+
+def get_ci_system(project_dir: Path) -> str | None:
+    """
+    Get the CI system name if configured.
+
+    Args:
+        project_dir: Path to project root
+
+    Returns:
+        CI system name or None
+    """
+    discovery = CIDiscovery()
+    result = discovery.discover(project_dir)
+    if result:
+        return result.ci_system
+    return None
+
+
+# =============================================================================
+# CLI
+# =============================================================================
+
+
+def main() -> None:
+    """CLI entry point for testing."""
+    import argparse
+
+    parser = argparse.ArgumentParser(description="Discover CI configuration")
+    parser.add_argument("project_dir", type=Path, help="Path to project root")
+    parser.add_argument("--json", action="store_true", help="Output as JSON")
+
+    args = parser.parse_args()
+
+    discovery = CIDiscovery()
+    result = discovery.discover(args.project_dir)
+
+    if not result:
+        print("No CI configuration found")
+        return
+
+    if args.json:
+        print(json.dumps(discovery.to_dict(result), indent=2))
+    else:
+        print(f"CI System: {result.ci_system}")
+        print(f"Config Files: {', '.join(result.config_files)}")
+        print("\nTest Commands:")
+        for test_type, cmd in result.test_commands.items():
+            print(f"  {test_type}: {cmd}")
+        if result.coverage_command:
+            print(f"\nCoverage Command: {result.coverage_command}")
+        print(f"\nWorkflows ({len(result.workflows)}):")
+        for w in result.workflows:
+            marker = "[TEST]" if w.test_related else ""
+            print(f"  - {w.name} {marker}")
+            if w.trigger:
+                print(f"    Triggers: {', '.join(str(t) for t in w.trigger)}")
+        if result.environment_variables:
+            print(f"\nEnvironment Variables: {', '.join(result.environment_variables)}")
+
+
+if __name__ == "__main__":
+    main()
@@ -0,0 +1,643 @@
+"""
+Insight Extractor
+=================
+
+Automatically extracts structured insights from completed coding sessions.
+Runs after each session to capture rich, actionable knowledge for Graphiti memory.
+
+Uses the Claude Agent SDK (same as the rest of the system) for extraction.
+Falls back to generic insights if extraction fails (never blocks the build).
+"""
+
+from __future__ import annotations
+
+import json
+import logging
+import os
+import subprocess
+from pathlib import Path
+from typing import Any
+
+logger = logging.getLogger(__name__)
+
+# Check for Claude SDK availability
+try:
+    from claude_agent_sdk import ClaudeAgentOptions, ClaudeSDKClient
+
+    SDK_AVAILABLE = True
+except ImportError:
+    SDK_AVAILABLE = False
+    ClaudeAgentOptions = None
+    ClaudeSDKClient = None
+
+from core.auth import ensure_claude_code_oauth_token, get_auth_token
+
+# Default model for insight extraction (fast and cheap)
+# Note: Using Haiku 4.5 for fast, cheap extraction. Haiku does not support
+# extended thinking, so thinking_default is set to "none" in models.py
+DEFAULT_EXTRACTION_MODEL = "claude-haiku-4-5-20251001"
+
+# Maximum diff size to send to the LLM (avoid context limits)
+MAX_DIFF_CHARS = 15000
+
+# Maximum attempt history entries to include
+MAX_ATTEMPTS_TO_INCLUDE = 3
+
+
+def is_extraction_enabled() -> bool:
+    """Check if insight extraction is enabled."""
+    # Extraction requires Claude SDK and authentication token
+    if not SDK_AVAILABLE:
+        return False
+    if not get_auth_token():
+        return False
+    enabled_str = os.environ.get("INSIGHT_EXTRACTION_ENABLED", "true").lower()
+    return enabled_str in ("true", "1", "yes")
+
+
+def get_extraction_model() -> str:
+    """Get the model to use for insight extraction."""
+    return os.environ.get("INSIGHT_EXTRACTOR_MODEL", DEFAULT_EXTRACTION_MODEL)
+
+
+# =============================================================================
+# Git Helpers
+# =============================================================================
+
+
+def get_session_diff(
+    project_dir: Path,
+    commit_before: str | None,
+    commit_after: str | None,
+) -> str:
+    """
+    Get the git diff between two commits.
+
+    Args:
+        project_dir: Project root directory
+        commit_before: Commit hash before session (or None)
+        commit_after: Commit hash after session (or None)
+
+    Returns:
+        Diff text (truncated if too large)
+    """
+    if not commit_before or not commit_after:
+        return "(No commits to diff)"
+
+    if commit_before == commit_after:
+        return "(No changes - same commit)"
+
+    try:
+        result = subprocess.run(
+            ["git", "diff", commit_before, commit_after],
+            cwd=project_dir,
+            capture_output=True,
+            text=True,
+            timeout=30,
+        )
+        diff = result.stdout
+
+        if len(diff) > MAX_DIFF_CHARS:
+            # Truncate and add note
+            diff = (
+                diff[:MAX_DIFF_CHARS] + f"\n\n... (truncated, {len(diff)} chars total)"
+            )
+
+        return diff if diff else "(Empty diff)"
+
+    except subprocess.TimeoutExpired:
+        logger.warning("Git diff timed out")
+        return "(Git diff timed out)"
+    except Exception as e:
+        logger.warning(f"Failed to get git diff: {e}")
+        return f"(Failed to get diff: {e})"
+
+
+def get_changed_files(
+    project_dir: Path,
+    commit_before: str | None,
+    commit_after: str | None,
+) -> list[str]:
+    """
+    Get list of files changed between two commits.
+
+    Args:
+        project_dir: Project root directory
+        commit_before: Commit hash before session
+        commit_after: Commit hash after session
+
+    Returns:
+        List of changed file paths
+    """
+    if not commit_before or not commit_after or commit_before == commit_after:
+        return []
+
+    try:
+        result = subprocess.run(
+            ["git", "diff", "--name-only", commit_before, commit_after],
+            cwd=project_dir,
+            capture_output=True,
+            text=True,
+            timeout=10,
+        )
+        files = [f.strip() for f in result.stdout.strip().split("\n") if f.strip()]
+        return files
+
+    except Exception as e:
+        logger.warning(f"Failed to get changed files: {e}")
+        return []
+
+
+def get_commit_messages(
+    project_dir: Path,
+    commit_before: str | None,
+    commit_after: str | None,
+) -> str:
+    """Get commit messages between two commits."""
+    if not commit_before or not commit_after or commit_before == commit_after:
+        return "(No commits)"
+
+    try:
+        result = subprocess.run(
+            ["git", "log", "--oneline", f"{commit_before}..{commit_after}"],
+            cwd=project_dir,
+            capture_output=True,
+            text=True,
+            timeout=10,
+        )
+        return result.stdout.strip() if result.stdout.strip() else "(No commits)"
+
+    except Exception as e:
+        logger.warning(f"Failed to get commit messages: {e}")
+        return f"(Failed: {e})"
+
+
+# =============================================================================
+# Input Gathering
+# =============================================================================
+
+
+def gather_extraction_inputs(
+    spec_dir: Path,
+    project_dir: Path,
+    subtask_id: str,
+    session_num: int,
+    commit_before: str | None,
+    commit_after: str | None,
+    success: bool,
+    recovery_manager: Any,
+) -> dict:
+    """
+    Gather all inputs needed for insight extraction.
+
+    Args:
+        spec_dir: Spec directory
+        project_dir: Project root
+        subtask_id: The subtask that was worked on
+        session_num: Session number
+        commit_before: Commit before session
+        commit_after: Commit after session
+        success: Whether session succeeded
+        recovery_manager: Recovery manager with attempt history
+
+    Returns:
+        Dict with all inputs for the extractor
+    """
+    # Get subtask description from implementation plan
+    subtask_description = _get_subtask_description(spec_dir, subtask_id)
+
+    # Get git diff
+    diff = get_session_diff(project_dir, commit_before, commit_after)
+
+    # Get changed files
+    changed_files = get_changed_files(project_dir, commit_before, commit_after)
+
+    # Get commit messages
+    commit_messages = get_commit_messages(project_dir, commit_before, commit_after)
+
+    # Get attempt history
+    attempt_history = _get_attempt_history(recovery_manager, subtask_id)
+
+    return {
+        "subtask_id": subtask_id,
+        "subtask_description": subtask_description,
+        "session_num": session_num,
+        "success": success,
+        "diff": diff,
+        "changed_files": changed_files,
+        "commit_messages": commit_messages,
+        "attempt_history": attempt_history,
+    }
+
+
+def _get_subtask_description(spec_dir: Path, subtask_id: str) -> str:
+    """Get subtask description from implementation plan."""
+    plan_file = spec_dir / "implementation_plan.json"
+    if not plan_file.exists():
+        return f"Subtask: {subtask_id}"
+
+    try:
+        with open(plan_file, encoding="utf-8") as f:
+            plan = json.load(f)
+
+        # Search through phases for the subtask
+        for phase in plan.get("phases", []):
+            for subtask in phase.get("subtasks", []):
+                if subtask.get("id") == subtask_id:
+                    return subtask.get("description", f"Subtask: {subtask_id}")
+
+        return f"Subtask: {subtask_id}"
+
+    except Exception as e:
+        logger.warning(f"Failed to load subtask description: {e}")
+        return f"Subtask: {subtask_id}"
+
+
+def _get_attempt_history(recovery_manager: Any, subtask_id: str) -> list[dict]:
+    """Get previous attempt history for this subtask."""
+    if not recovery_manager:
+        return []
+
+    try:
+        history = recovery_manager.get_subtask_history(subtask_id)
+        attempts = history.get("attempts", [])
+
+        # Limit to recent attempts
+        return attempts[-MAX_ATTEMPTS_TO_INCLUDE:]
+
+    except Exception as e:
+        logger.warning(f"Failed to get attempt history: {e}")
+        return []
+
+
+# =============================================================================
+# LLM Extraction
+# =============================================================================
+
+
+def _build_extraction_prompt(inputs: dict) -> str:
+    """Build the prompt for insight extraction."""
+    prompt_file = Path(__file__).parent / "prompts" / "insight_extractor.md"
+
+    if prompt_file.exists():
+        base_prompt = prompt_file.read_text(encoding="utf-8")
+    else:
+        # Fallback if prompt file missing
+        base_prompt = """Extract structured insights from this coding session.
+Output ONLY valid JSON with: file_insights, patterns_discovered, gotchas_discovered, approach_outcome, recommendations"""
+
+    # Build session context
+    session_context = f"""
+---
+
+## SESSION DATA
+
+### Subtask
+- **ID**: {inputs["subtask_id"]}
+- **Description**: {inputs["subtask_description"]}
+- **Session Number**: {inputs["session_num"]}
+- **Outcome**: {"SUCCESS" if inputs["success"] else "FAILED"}
+
+### Files Changed
+{chr(10).join(f"- {f}" for f in inputs["changed_files"]) if inputs["changed_files"] else "(No files changed)"}
+
+### Commit Messages
+{inputs["commit_messages"]}
+
+### Git Diff
+```diff
+{inputs["diff"]}
+```
+
+### Previous Attempts
+{_format_attempt_history(inputs["attempt_history"])}
+
+---
+
+Now analyze this session and output ONLY the JSON object.
+"""
+
+    return base_prompt + session_context
+
+
+def _format_attempt_history(attempts: list[dict]) -> str:
+    """Format attempt history for the prompt."""
+    if not attempts:
+        return "(First attempt - no previous history)"
+
+    lines = []
+    for i, attempt in enumerate(attempts, 1):
+        success = "SUCCESS" if attempt.get("success") else "FAILED"
+        approach = attempt.get("approach", "Unknown approach")
+        error = attempt.get("error", "")
+        lines.append(f"**Attempt {i}** ({success}): {approach}")
+        if error:
+            lines.append(f"  Error: {error}")
+
+    return "\n".join(lines)
+
+
+async def run_insight_extraction(
+    inputs: dict, project_dir: Path | None = None
+) -> dict | None:
+    """
+    Run the insight extraction using Claude Agent SDK.
+
+    Args:
+        inputs: Gathered session inputs
+        project_dir: Project directory for SDK context (optional)
+
+    Returns:
+        Extracted insights dict or None if failed
+    """
+    if not SDK_AVAILABLE:
+        logger.warning("Claude SDK not available, skipping insight extraction")
+        return None
+
+    if not get_auth_token():
+        logger.warning("No authentication token found, skipping insight extraction")
+        return None
+
+    # Ensure SDK can find the token
+    ensure_claude_code_oauth_token()
+
+    model = get_extraction_model()
+    prompt = _build_extraction_prompt(inputs)
+
+    # Use current directory if project_dir not specified
+    cwd = str(project_dir.resolve()) if project_dir else os.getcwd()
+
+    try:
+        # Use simple_client for insight extraction
+        from pathlib import Path
+
+        from core.simple_client import create_simple_client
+
+        client = create_simple_client(
+            agent_type="insights",
+            model=model,
+            system_prompt=(
+                "You are an expert code analyst. You extract structured insights from coding sessions. "
+                "Always respond with valid JSON only, no markdown formatting or explanations."
+            ),
+            cwd=Path(cwd) if cwd else None,
+        )
+
+        # Use async context manager
+        async with client:
+            await client.query(prompt)
+
+            # Collect the response
+            response_text = ""
+            message_count = 0
+            text_blocks_found = 0
+
+            async for msg in client.receive_response():
+                msg_type = type(msg).__name__
+                message_count += 1
+
+                if msg_type == "AssistantMessage" and hasattr(msg, "content"):
+                    for block in msg.content:
+                        # Must check block type - only TextBlock has .text attribute
+                        block_type = type(block).__name__
+                        if block_type == "TextBlock" and hasattr(block, "text"):
+                            text_blocks_found += 1
+                            if block.text:  # Only add non-empty text
+                                response_text += block.text
+                            else:
+                                logger.debug(
+                                    f"Found empty TextBlock in response (block #{text_blocks_found})"
+                                )
+
+            # Log response collection summary
+            logger.debug(
+                f"Insight extraction response: {message_count} messages, "
+                f"{text_blocks_found} text blocks, {len(response_text)} chars collected"
+            )
+
+            # Validate we received content before parsing
+            if not response_text.strip():
+                logger.warning(
+                    f"Insight extraction returned empty response. "
+                    f"Messages received: {message_count}, TextBlocks found: {text_blocks_found}. "
+                    f"This may indicate the AI model did not respond with text content."
+                )
+                return None
+
+        # Parse JSON from response
+        return parse_insights(response_text)
+
+    except Exception as e:
+        logger.warning(f"Insight extraction failed: {e}")
+        return None
+
+
+def parse_insights(response_text: str) -> dict | None:
+    """
+    Parse the LLM response into structured insights.
+
+    Args:
+        response_text: Raw LLM response
+
+    Returns:
+        Parsed insights dict or None if parsing failed
+    """
+    # Try to extract JSON from the response
+    text = response_text.strip()
+
+    # Early validation - check for empty response
+    if not text:
+        logger.warning("Cannot parse insights: response text is empty")
+        return None
+
+    # Handle markdown code blocks
+    if text.startswith("```"):
+        # Remove code block markers
+        lines = text.split("\n")
+        # Remove first line (```json or ```)
+        if lines[0].startswith("```"):
+            lines = lines[1:]
+        # Remove last line if it's ```
+        if lines and lines[-1].strip() == "```":
+            lines = lines[:-1]
+        text = "\n".join(lines).strip()
+
+        # Check again after removing code blocks
+        if not text:
+            logger.warning(
+                "Cannot parse insights: response contained only markdown code block markers with no content"
+            )
+            return None
+
+    try:
+        insights = json.loads(text)
+
+        # Validate structure
+        if not isinstance(insights, dict):
+            logger.warning(
+                f"Insights is not a dict, got type: {type(insights).__name__}"
+            )
+            return None
+
+        # Ensure required keys exist with defaults
+        insights.setdefault("file_insights", [])
+        insights.setdefault("patterns_discovered", [])
+        insights.setdefault("gotchas_discovered", [])
+        insights.setdefault("approach_outcome", {})
+        insights.setdefault("recommendations", [])
+
+        return insights
+
+    except json.JSONDecodeError as e:
+        logger.warning(f"Failed to parse insights JSON: {e}")
+        # Show more context in the error message
+        preview_length = min(500, len(text))
+        logger.warning(
+            f"Response text preview (first {preview_length} chars): {text[:preview_length]}"
+        )
+        if len(text) > preview_length:
+            logger.warning(f"... (total length: {len(text)} chars)")
+        return None
+
+
+# =============================================================================
+# Main Entry Point
+# =============================================================================
+
+
+async def extract_session_insights(
+    spec_dir: Path,
+    project_dir: Path,
+    subtask_id: str,
+    session_num: int,
+    commit_before: str | None,
+    commit_after: str | None,
+    success: bool,
+    recovery_manager: Any,
+) -> dict:
+    """
+    Extract insights from a completed coding session.
+
+    This is the main entry point called from post_session_processing().
+    Falls back to generic insights if extraction fails.
+
+    Args:
+        spec_dir: Spec directory
+        project_dir: Project root
+        subtask_id: Subtask that was worked on
+        session_num: Session number
+        commit_before: Commit before session
+        commit_after: Commit after session
+        success: Whether session succeeded
+        recovery_manager: Recovery manager with attempt history
+
+    Returns:
+        Insights dict (rich if extraction succeeded, generic if failed)
+    """
+    # Check if extraction is enabled
+    if not is_extraction_enabled():
+        logger.info("Insight extraction disabled")
+        return _get_generic_insights(subtask_id, success)
+
+    # Check for no changes
+    if commit_before == commit_after:
+        logger.info("No changes to extract insights from")
+        return _get_generic_insights(subtask_id, success)
+
+    try:
+        # Gather inputs
+        inputs = gather_extraction_inputs(
+            spec_dir=spec_dir,
+            project_dir=project_dir,
+            subtask_id=subtask_id,
+            session_num=session_num,
+            commit_before=commit_before,
+            commit_after=commit_after,
+            success=success,
+            recovery_manager=recovery_manager,
+        )
+
+        # Run extraction
+        extracted = await run_insight_extraction(inputs, project_dir=project_dir)
+
+        if extracted:
+            # Add metadata
+            extracted["subtask_id"] = subtask_id
+            extracted["session_num"] = session_num
+            extracted["success"] = success
+            extracted["changed_files"] = inputs["changed_files"]
+
+            logger.info(
+                f"Extracted insights: {len(extracted.get('file_insights', []))} file insights, "
+                f"{len(extracted.get('patterns_discovered', []))} patterns, "
+                f"{len(extracted.get('gotchas_discovered', []))} gotchas"
+            )
+            return extracted
+        else:
+            logger.warning("Extraction returned no results, using generic insights")
+            return _get_generic_insights(subtask_id, success)
+
+    except Exception as e:
+        logger.warning(f"Insight extraction failed: {e}, using generic insights")
+        return _get_generic_insights(subtask_id, success)
+
+
+def _get_generic_insights(subtask_id: str, success: bool) -> dict:
+    """Return generic insights when extraction fails or is disabled."""
+    return {
+        "file_insights": [],
+        "patterns_discovered": [],
+        "gotchas_discovered": [],
+        "approach_outcome": {
+            "success": success,
+            "approach_used": f"Implemented subtask: {subtask_id}",
+            "why_it_worked": None,
+            "why_it_failed": None,
+            "alternatives_tried": [],
+        },
+        "recommendations": [],
+        "subtask_id": subtask_id,
+        "success": success,
+        "changed_files": [],
+    }
+
+
+# =============================================================================
+# CLI for Testing
+# =============================================================================
+
+if __name__ == "__main__":
+    import argparse
+    import asyncio
+
+    parser = argparse.ArgumentParser(description="Test insight extraction")
+    parser.add_argument("--spec-dir", type=Path, required=True, help="Spec directory")
+    parser.add_argument(
+        "--project-dir", type=Path, required=True, help="Project directory"
+    )
+    parser.add_argument(
+        "--commit-before", type=str, required=True, help="Commit before session"
+    )
+    parser.add_argument(
+        "--commit-after", type=str, required=True, help="Commit after session"
+    )
+    parser.add_argument(
+        "--subtask-id", type=str, default="test-subtask", help="Subtask ID"
+    )
+
+    args = parser.parse_args()
+
+    async def main():
+        insights = await extract_session_insights(
+            spec_dir=args.spec_dir,
+            project_dir=args.project_dir,
+            subtask_id=args.subtask_id,
+            session_num=1,
+            commit_before=args.commit_before,
+            commit_after=args.commit_after,
+            success=True,
+            recovery_manager=None,
+        )
+        print(json.dumps(insights, indent=2))
+
+    asyncio.run(main())
@@ -0,0 +1,109 @@
+"""
+Smart Project Analyzer for Dynamic Security Profiles
+=====================================================
+
+FACADE MODULE: This module re-exports all functionality from the
+auto-claude/project/ package for backward compatibility.
+
+The implementation has been refactored into focused modules:
+- project/command_registry.py - Command registries
+- project/models.py - Data structures
+- project/config_parser.py - Config file parsing
+- project/stack_detector.py - Stack detection
+- project/framework_detector.py - Framework detection
+- project/structure_analyzer.py - Project structure analysis
+- project/analyzer.py - Main orchestration
+
+This file maintains the original API so existing imports continue to work.
+
+This system:
+1. Detects languages, frameworks, databases, and infrastructure
+2. Parses package.json scripts, Makefile targets, pyproject.toml scripts
+3. Builds a tailored security profile for the specific project
+4. Caches the profile for subsequent runs
+5. Can re-analyze when project structure changes
+
+The goal: Allow an AI developer to run any command that's legitimately
+needed for the detected tech stack, while blocking dangerous operations.
+"""
+
+# Re-export all public API from the project module
+
+from __future__ import annotations
+
+from project import (
+    # Command registries
+    BASE_COMMANDS,
+    VALIDATED_COMMANDS,
+    CustomScripts,
+    # Main classes
+    ProjectAnalyzer,
+    SecurityProfile,
+    TechnologyStack,
+    # Utility functions
+    get_or_create_profile,
+    is_command_allowed,
+    needs_validation,
+)
+
+# Also re-export command registries for backward compatibility
+from project.command_registry import (
+    CLOUD_COMMANDS,
+    CODE_QUALITY_COMMANDS,
+    DATABASE_COMMANDS,
+    FRAMEWORK_COMMANDS,
+    INFRASTRUCTURE_COMMANDS,
+    LANGUAGE_COMMANDS,
+    PACKAGE_MANAGER_COMMANDS,
+    VERSION_MANAGER_COMMANDS,
+)
+
+__all__ = [
+    # Main classes
+    "ProjectAnalyzer",
+    "SecurityProfile",
+    "TechnologyStack",
+    "CustomScripts",
+    # Utility functions
+    "get_or_create_profile",
+    "is_command_allowed",
+    "needs_validation",
+    # Base command sets
+    "BASE_COMMANDS",
+    "VALIDATED_COMMANDS",
+    # Technology-specific command sets
+    "LANGUAGE_COMMANDS",
+    "PACKAGE_MANAGER_COMMANDS",
+    "FRAMEWORK_COMMANDS",
+    "DATABASE_COMMANDS",
+    "INFRASTRUCTURE_COMMANDS",
+    "CLOUD_COMMANDS",
+    "CODE_QUALITY_COMMANDS",
+    "VERSION_MANAGER_COMMANDS",
+]
+
+
+# =============================================================================
+# CLI for testing
+# =============================================================================
+
+if __name__ == "__main__":
+    import sys
+    from pathlib import Path
+
+    if len(sys.argv) < 2:
+        print("Usage: python project_analyzer.py <project_dir> [--force]")
+        sys.exit(1)
+
+    project_dir = Path(sys.argv[1])
+    force = "--force" in sys.argv
+
+    if not project_dir.exists():
+        print(f"Error: {project_dir} does not exist")
+        sys.exit(1)
+
+    profile = get_or_create_profile(project_dir, force_reanalyze=force)
+
+    print("\nAllowed commands:")
+    for cmd in sorted(profile.get_all_allowed_commands()):
+        print(f"  {cmd}")
@@ -0,0 +1,591 @@
+#!/usr/bin/env python3
+"""
+Risk Classifier Module
+======================
+
+Reads the AI-generated complexity_assessment.json and provides programmatic
+access to risk classification and validation recommendations.
+
+This module serves as the bridge between the AI complexity assessor prompt
+and the rest of the validation system.
+
+Usage:
+    from risk_classifier import RiskClassifier
+
+    classifier = RiskClassifier()
+    assessment = classifier.load_assessment(spec_dir)
+
+    if classifier.should_skip_validation(spec_dir):
+        print("Validation can be skipped for this task")
+
+    test_types = classifier.get_required_test_types(spec_dir)
+"""
+
+from __future__ import annotations
+
+import json
+from dataclasses import dataclass, field
+from pathlib import Path
+from typing import Any
+
+# =============================================================================
+# DATA CLASSES
+# =============================================================================
+
+
+@dataclass
+class ScopeAnalysis:
+    """Analysis of task scope."""
+
+    estimated_files: int = 0
+    estimated_services: int = 0
+    is_cross_cutting: bool = False
+    notes: str = ""
+
+
+@dataclass
+class IntegrationAnalysis:
+    """Analysis of external integrations."""
+
+    external_services: list[str] = field(default_factory=list)
+    new_dependencies: list[str] = field(default_factory=list)
+    research_needed: bool = False
+    notes: str = ""
+
+
+@dataclass
+class InfrastructureAnalysis:
+    """Analysis of infrastructure requirements."""
+
+    docker_changes: bool = False
+    database_changes: bool = False
+    config_changes: bool = False
+    notes: str = ""
+
+
+@dataclass
+class KnowledgeAnalysis:
+    """Analysis of knowledge requirements."""
+
+    patterns_exist: bool = True
+    research_required: bool = False
+    unfamiliar_tech: list[str] = field(default_factory=list)
+    notes: str = ""
+
+
+@dataclass
+class RiskAnalysis:
+    """Analysis of task risk."""
+
+    level: str = "low"  # low, medium, high
+    concerns: list[str] = field(default_factory=list)
+    notes: str = ""
+
+
+@dataclass
+class ComplexityAnalysis:
+    """Full complexity analysis from the AI assessor."""
+
+    scope: ScopeAnalysis = field(default_factory=ScopeAnalysis)
+    integrations: IntegrationAnalysis = field(default_factory=IntegrationAnalysis)
+    infrastructure: InfrastructureAnalysis = field(
+        default_factory=InfrastructureAnalysis
+    )
+    knowledge: KnowledgeAnalysis = field(default_factory=KnowledgeAnalysis)
+    risk: RiskAnalysis = field(default_factory=RiskAnalysis)
+
+
+@dataclass
+class ValidationRecommendations:
+    """Validation recommendations from the AI assessor."""
+
+    risk_level: str = "medium"  # trivial, low, medium, high, critical
+    skip_validation: bool = False
+    minimal_mode: bool = False
+    test_types_required: list[str] = field(default_factory=lambda: ["unit"])
+    security_scan_required: bool = False
+    staging_deployment_required: bool = False
+    reasoning: str = ""
+
+
+@dataclass
+class AssessmentFlags:
+    """Flags indicating special requirements."""
+
+    needs_research: bool = False
+    needs_self_critique: bool = False
+    needs_infrastructure_setup: bool = False
+
+
+@dataclass
+class RiskAssessment:
+    """Complete risk assessment from complexity_assessment.json."""
+
+    complexity: str  # simple, standard, complex
+    workflow_type: str  # feature, refactor, investigation, migration, simple
+    confidence: float
+    reasoning: str
+    analysis: ComplexityAnalysis
+    recommended_phases: list[str]
+    flags: AssessmentFlags
+    validation: ValidationRecommendations
+    created_at: str | None = None
+
+    @property
+    def risk_level(self) -> str:
+        """Get the risk level from validation recommendations."""
+        return self.validation.risk_level
+
+
+# =============================================================================
+# RISK CLASSIFIER
+# =============================================================================
+
+
+class RiskClassifier:
+    """
+    Reads AI-generated complexity_assessment.json and provides risk classification.
+
+    The complexity_assessment.json is generated by the AI complexity assessor
+    agent using the complexity_assessor.md prompt. This module parses that output
+    and provides programmatic access to the risk classification.
+    """
+
+    def __init__(self) -> None:
+        """Initialize the risk classifier."""
+        self._cache: dict[str, RiskAssessment] = {}
+
+    def load_assessment(self, spec_dir: Path) -> RiskAssessment | None:
+        """
+        Load complexity_assessment.json from spec directory.
+
+        Args:
+            spec_dir: Path to the spec directory containing complexity_assessment.json
+
+        Returns:
+            RiskAssessment object if file exists and is valid, None otherwise
+        """
+        spec_dir = Path(spec_dir)
+        cache_key = str(spec_dir.resolve())
+
+        # Return cached result if available
+        if cache_key in self._cache:
+            return self._cache[cache_key]
+
+        assessment_file = spec_dir / "complexity_assessment.json"
+        if not assessment_file.exists():
+            return None
+
+        try:
+            with open(assessment_file, encoding="utf-8") as f:
+                data = json.load(f)
+
+            assessment = self._parse_assessment(data)
+            self._cache[cache_key] = assessment
+            return assessment
+
+        except (json.JSONDecodeError, KeyError, TypeError) as e:
+            # Log error but don't crash - return None to allow fallback behavior
+            print(f"Warning: Failed to parse complexity_assessment.json: {e}")
+            return None
+
+    def _parse_assessment(self, data: dict[str, Any]) -> RiskAssessment:
+        """Parse raw JSON data into a RiskAssessment object."""
+        # Parse analysis sections
+        analysis_data = data.get("analysis", {})
+        analysis = ComplexityAnalysis(
+            scope=self._parse_scope(analysis_data.get("scope", {})),
+            integrations=self._parse_integrations(
+                analysis_data.get("integrations", {})
+            ),
+            infrastructure=self._parse_infrastructure(
+                analysis_data.get("infrastructure", {})
+            ),
+            knowledge=self._parse_knowledge(analysis_data.get("knowledge", {})),
+            risk=self._parse_risk(analysis_data.get("risk", {})),
+        )
+
+        # Parse flags
+        flags_data = data.get("flags", {})
+        flags = AssessmentFlags(
+            needs_research=flags_data.get("needs_research", False),
+            needs_self_critique=flags_data.get("needs_self_critique", False),
+            needs_infrastructure_setup=flags_data.get(
+                "needs_infrastructure_setup", False
+            ),
+        )
+
+        # Parse validation recommendations
+        validation_data = data.get("validation_recommendations", {})
+        validation = self._parse_validation_recommendations(validation_data, analysis)
+
+        return RiskAssessment(
+            complexity=data.get("complexity", "standard"),
+            workflow_type=data.get("workflow_type", "feature"),
+            confidence=float(data.get("confidence", 0.5)),
+            reasoning=data.get("reasoning", ""),
+            analysis=analysis,
+            recommended_phases=data.get("recommended_phases", []),
+            flags=flags,
+            validation=validation,
+            created_at=data.get("created_at"),
+        )
+
+    def _parse_scope(self, data: dict[str, Any]) -> ScopeAnalysis:
+        """Parse scope analysis section."""
+        return ScopeAnalysis(
+            estimated_files=int(data.get("estimated_files", 0)),
+            estimated_services=int(data.get("estimated_services", 0)),
+            is_cross_cutting=bool(data.get("is_cross_cutting", False)),
+            notes=str(data.get("notes", "")),
+        )
+
+    def _parse_integrations(self, data: dict[str, Any]) -> IntegrationAnalysis:
+        """Parse integrations analysis section."""
+        return IntegrationAnalysis(
+            external_services=list(data.get("external_services", [])),
+            new_dependencies=list(data.get("new_dependencies", [])),
+            research_needed=bool(data.get("research_needed", False)),
+            notes=str(data.get("notes", "")),
+        )
+
+    def _parse_infrastructure(self, data: dict[str, Any]) -> InfrastructureAnalysis:
+        """Parse infrastructure analysis section."""
+        return InfrastructureAnalysis(
+            docker_changes=bool(data.get("docker_changes", False)),
+            database_changes=bool(data.get("database_changes", False)),
+            config_changes=bool(data.get("config_changes", False)),
+            notes=str(data.get("notes", "")),
+        )
+
+    def _parse_knowledge(self, data: dict[str, Any]) -> KnowledgeAnalysis:
+        """Parse knowledge analysis section."""
+        return KnowledgeAnalysis(
+            patterns_exist=bool(data.get("patterns_exist", True)),
+            research_required=bool(data.get("research_required", False)),
+            unfamiliar_tech=list(data.get("unfamiliar_tech", [])),
+            notes=str(data.get("notes", "")),
+        )
+
+    def _parse_risk(self, data: dict[str, Any]) -> RiskAnalysis:
+        """Parse risk analysis section."""
+        return RiskAnalysis(
+            level=str(data.get("level", "low")),
+            concerns=list(data.get("concerns", [])),
+            notes=str(data.get("notes", "")),
+        )
+
+    def _parse_validation_recommendations(
+        self, data: dict[str, Any], analysis: ComplexityAnalysis
+    ) -> ValidationRecommendations:
+        """
+        Parse validation recommendations section.
+
+        If validation_recommendations is not present in the JSON (older assessments),
+        infer appropriate values from the analysis.
+        """
+        if data:
+            # New format with explicit validation recommendations
+            return ValidationRecommendations(
+                risk_level=str(data.get("risk_level", "medium")),
+                skip_validation=bool(data.get("skip_validation", False)),
+                minimal_mode=bool(data.get("minimal_mode", False)),
+                test_types_required=list(data.get("test_types_required", ["unit"])),
+                security_scan_required=bool(data.get("security_scan_required", False)),
+                staging_deployment_required=bool(
+                    data.get("staging_deployment_required", False)
+                ),
+                reasoning=str(data.get("reasoning", "")),
+            )
+        else:
+            # Infer from analysis (backward compatibility)
+            return self._infer_validation_recommendations(analysis)
+
+    def _infer_validation_recommendations(
+        self, analysis: ComplexityAnalysis
+    ) -> ValidationRecommendations:
+        """
+        Infer validation recommendations from analysis when not explicitly provided.
+
+        This provides backward compatibility with older complexity assessments
+        that don't have the validation_recommendations section.
+        """
+        risk_level = analysis.risk.level
+
+        # Map old risk levels to new ones
+        risk_mapping = {
+            "low": "low",
+            "medium": "medium",
+            "high": "high",
+        }
+        normalized_risk = risk_mapping.get(risk_level, "medium")
+
+        # Infer test types based on risk
+        test_types_map = {
+            "low": ["unit"],
+            "medium": ["unit", "integration"],
+            "high": ["unit", "integration", "e2e"],
+        }
+        test_types = test_types_map.get(normalized_risk, ["unit", "integration"])
+
+        # Security scan for high risk or security-related concerns
+        security_keywords = [
+            "security",
+            "auth",
+            "password",
+            "credential",
+            "token",
+            "api key",
+        ]
+        has_security_concerns = any(
+            kw in str(analysis.risk.concerns).lower() for kw in security_keywords
+        )
+        security_scan_required = normalized_risk == "high" or has_security_concerns
+
+        # Staging for database or infrastructure changes
+        staging_required = (
+            analysis.infrastructure.database_changes
+            and normalized_risk in ["medium", "high"]
+        )
+
+        # Minimal mode for simple changes
+        minimal_mode = (
+            analysis.scope.estimated_files <= 2
+            and analysis.scope.estimated_services <= 1
+            and not analysis.integrations.external_services
+        )
+
+        return ValidationRecommendations(
+            risk_level=normalized_risk,
+            skip_validation=False,  # Never skip by inference
+            minimal_mode=minimal_mode,
+            test_types_required=test_types,
+            security_scan_required=security_scan_required,
+            staging_deployment_required=staging_required,
+            reasoning="Inferred from complexity analysis (no explicit recommendations found)",
+        )
+
+    def should_skip_validation(self, spec_dir: Path) -> bool:
+        """
+        Quick check if validation can be skipped entirely.
+
+        Args:
+            spec_dir: Path to the spec directory
+
+        Returns:
+            True if validation can be skipped (trivial changes), False otherwise
+        """
+        assessment = self.load_assessment(spec_dir)
+        if not assessment:
+            return False  # When in doubt, don't skip
+
+        return assessment.validation.skip_validation
+
+    def should_use_minimal_mode(self, spec_dir: Path) -> bool:
+        """
+        Check if minimal validation mode should be used.
+
+        Args:
+            spec_dir: Path to the spec directory
+
+        Returns:
+            True if minimal mode is recommended, False otherwise
+        """
+        assessment = self.load_assessment(spec_dir)
+        if not assessment:
+            return False
+
+        return assessment.validation.minimal_mode
+
+    def get_required_test_types(self, spec_dir: Path) -> list[str]:
+        """
+        Get list of required test types based on risk.
+
+        Args:
+            spec_dir: Path to the spec directory
+
+        Returns:
+            List of test types (e.g., ["unit", "integration", "e2e"])
+        """
+        assessment = self.load_assessment(spec_dir)
+        if not assessment:
+            return ["unit"]  # Default to unit tests
+
+        return assessment.validation.test_types_required
+
+    def requires_security_scan(self, spec_dir: Path) -> bool:
+        """
+        Check if security scanning is required.
+
+        Args:
+            spec_dir: Path to the spec directory
+
+        Returns:
+            True if security scan is required, False otherwise
+        """
+        assessment = self.load_assessment(spec_dir)
+        if not assessment:
+            return False
+
+        return assessment.validation.security_scan_required
+
+    def requires_staging_deployment(self, spec_dir: Path) -> bool:
+        """
+        Check if staging deployment is required.
+
+        Args:
+            spec_dir: Path to the spec directory
+
+        Returns:
+            True if staging deployment is required, False otherwise
+        """
+        assessment = self.load_assessment(spec_dir)
+        if not assessment:
+            return False
+
+        return assessment.validation.staging_deployment_required
+
+    def get_risk_level(self, spec_dir: Path) -> str:
+        """
+        Get the risk level for the task.
+
+        Args:
+            spec_dir: Path to the spec directory
+
+        Returns:
+            Risk level string (trivial, low, medium, high, critical)
+        """
+        assessment = self.load_assessment(spec_dir)
+        if not assessment:
+            return "medium"  # Default to medium when unknown
+
+        return assessment.validation.risk_level
+
+    def get_complexity(self, spec_dir: Path) -> str:
+        """
+        Get the complexity level for the task.
+
+        Args:
+            spec_dir: Path to the spec directory
+
+        Returns:
+            Complexity level string (simple, standard, complex)
+        """
+        assessment = self.load_assessment(spec_dir)
+        if not assessment:
+            return "standard"  # Default to standard when unknown
+
+        return assessment.complexity
+
+    def get_validation_summary(self, spec_dir: Path) -> dict[str, Any]:
+        """
+        Get a summary of validation requirements.
+
+        Args:
+            spec_dir: Path to the spec directory
+
+        Returns:
+            Dictionary with validation summary
+        """
+        assessment = self.load_assessment(spec_dir)
+        if not assessment:
+            return {
+                "risk_level": "unknown",
+                "complexity": "unknown",
+                "skip_validation": False,
+                "minimal_mode": False,
+                "test_types": ["unit"],
+                "security_scan": False,
+                "staging_deployment": False,
+                "confidence": 0.0,
+            }
+
+        return {
+            "risk_level": assessment.validation.risk_level,
+            "complexity": assessment.complexity,
+            "skip_validation": assessment.validation.skip_validation,
+            "minimal_mode": assessment.validation.minimal_mode,
+            "test_types": assessment.validation.test_types_required,
+            "security_scan": assessment.validation.security_scan_required,
+            "staging_deployment": assessment.validation.staging_deployment_required,
+            "confidence": assessment.confidence,
+            "reasoning": assessment.validation.reasoning,
+        }
+
+    def clear_cache(self) -> None:
+        """Clear the internal cache of loaded assessments."""
+        self._cache.clear()
+
+
+# =============================================================================
+# CONVENIENCE FUNCTIONS
+# =============================================================================
+
+
+def load_risk_assessment(spec_dir: Path) -> RiskAssessment | None:
+    """
+    Convenience function to load a risk assessment.
+
+    Args:
+        spec_dir: Path to the spec directory
+
+    Returns:
+        RiskAssessment object or None
+    """
+    classifier = RiskClassifier()
+    return classifier.load_assessment(spec_dir)
+
+
+def get_validation_requirements(spec_dir: Path) -> dict[str, Any]:
+    """
+    Convenience function to get validation requirements.
+
+    Args:
+        spec_dir: Path to the spec directory
+
+    Returns:
+        Dictionary with validation requirements
+    """
+    classifier = RiskClassifier()
+    return classifier.get_validation_summary(spec_dir)
+
+
+# =============================================================================
+# CLI
+# =============================================================================
+
+
+def main() -> None:
+    """CLI entry point for testing."""
+    import argparse
+
+    parser = argparse.ArgumentParser(description="Load and display risk assessment")
+    parser.add_argument(
+        "spec_dir",
+        type=Path,
+        help="Path to spec directory with complexity_assessment.json",
+    )
+    parser.add_argument("--json", action="store_true", help="Output as JSON")
+
+    args = parser.parse_args()
+
+    classifier = RiskClassifier()
+    summary = classifier.get_validation_summary(args.spec_dir)
+
+    if args.json:
+        print(json.dumps(summary, indent=2))
+    else:
+        print(f"Risk Level: {summary['risk_level']}")
+        print(f"Complexity: {summary['complexity']}")
+        print(f"Skip Validation: {summary['skip_validation']}")
+        print(f"Minimal Mode: {summary['minimal_mode']}")
+        print(f"Test Types: {', '.join(summary['test_types'])}")
+        print(f"Security Scan: {summary['security_scan']}")
+        print(f"Staging Deployment: {summary['staging_deployment']}")
+        print(f"Confidence: {summary['confidence']:.2f}")
+        if summary.get("reasoning"):
+            print(f"Reasoning: {summary['reasoning']}")
+
+
+if __name__ == "__main__":
+    main()
@@ -0,0 +1,599 @@
+#!/usr/bin/env python3
+"""
+Security Scanner Module
+=======================
+
+Consolidates security scanning including secrets detection and SAST tools.
+This module integrates the existing scan_secrets.py and provides a unified
+interface for all security scanning.
+
+The security scanner is used by:
+- QA Agent: To verify no secrets are committed
+- Validation Strategy: To run security scans for high-risk changes
+
+Usage:
+    from analysis.security_scanner import SecurityScanner
+
+    scanner = SecurityScanner()
+    results = scanner.scan(project_dir, spec_dir)
+
+    if results.has_critical_issues:
+        print("Security issues found - blocking QA approval")
+"""
+
+from __future__ import annotations
+
+import json
+import subprocess
+from dataclasses import dataclass, field
+from pathlib import Path
+from typing import Any
+
+# Import the existing secrets scanner
+try:
+    from security.scan_secrets import SecretMatch, get_all_tracked_files, scan_files
+
+    HAS_SECRETS_SCANNER = True
+except ImportError:
+    HAS_SECRETS_SCANNER = False
+    SecretMatch = None
+
+
+# =============================================================================
+# DATA CLASSES
+# =============================================================================
+
+
+@dataclass
+class SecurityVulnerability:
+    """
+    Represents a security vulnerability found during scanning.
+
+    Attributes:
+        severity: Severity level (critical, high, medium, low, info)
+        source: Which scanner found this (secrets, bandit, npm_audit, etc.)
+        title: Short title of the vulnerability
+        description: Detailed description
+        file: File where vulnerability was found (if applicable)
+        line: Line number (if applicable)
+        cwe: CWE identifier if available
+    """
+
+    severity: str  # critical, high, medium, low, info
+    source: str  # secrets, bandit, npm_audit, semgrep, etc.
+    title: str
+    description: str
+    file: str | None = None
+    line: int | None = None
+    cwe: str | None = None
+
+
+@dataclass
+class SecurityScanResult:
+    """
+    Result of a security scan.
+
+    Attributes:
+        secrets: List of detected secrets
+        vulnerabilities: List of security vulnerabilities
+        scan_errors: List of errors during scanning
+        has_critical_issues: Whether any critical issues were found
+        should_block_qa: Whether these results should block QA approval
+    """
+
+    secrets: list[dict[str, Any]] = field(default_factory=list)
+    vulnerabilities: list[SecurityVulnerability] = field(default_factory=list)
+    scan_errors: list[str] = field(default_factory=list)
+    has_critical_issues: bool = False
+    should_block_qa: bool = False
+
+
+# =============================================================================
+# SECURITY SCANNER
+# =============================================================================
+
+
+class SecurityScanner:
+    """
+    Consolidates all security scanning operations.
+
+    Integrates:
+    - scan_secrets.py for secrets detection
+    - Bandit for Python SAST (if available)
+    - npm audit for JavaScript vulnerabilities (if applicable)
+    """
+
+    def __init__(self) -> None:
+        """Initialize the security scanner."""
+        self._bandit_available: bool | None = None
+        self._npm_available: bool | None = None
+
+    def scan(
+        self,
+        project_dir: Path,
+        spec_dir: Path | None = None,
+        changed_files: list[str] | None = None,
+        run_secrets: bool = True,
+        run_sast: bool = True,
+        run_dependency_audit: bool = True,
+    ) -> SecurityScanResult:
+        """
+        Run all applicable security scans.
+
+        Args:
+            project_dir: Path to the project root
+            spec_dir: Path to the spec directory (for storing results)
+            changed_files: Optional list of files to scan (if None, scans all)
+            run_secrets: Whether to run secrets scanning
+            run_sast: Whether to run SAST tools
+            run_dependency_audit: Whether to run dependency audits
+
+        Returns:
+            SecurityScanResult with all findings
+        """
+        project_dir = Path(project_dir)
+        result = SecurityScanResult()
+
+        # Run secrets scan
+        if run_secrets:
+            self._run_secrets_scan(project_dir, changed_files, result)
+
+        # Run SAST based on project type
+        if run_sast:
+            self._run_sast_scans(project_dir, result)
+
+        # Run dependency audits
+        if run_dependency_audit:
+            self._run_dependency_audits(project_dir, result)
+
+        # Determine if should block QA
+        result.has_critical_issues = (
+            any(v.severity in ["critical", "high"] for v in result.vulnerabilities)
+            or len(result.secrets) > 0
+        )
+
+        # Any secrets always block, critical vulnerabilities block
+        result.should_block_qa = len(result.secrets) > 0 or any(
+            v.severity == "critical" for v in result.vulnerabilities
+        )
+
+        # Save results if spec_dir provided
+        if spec_dir:
+            self._save_results(spec_dir, result)
+
+        return result
+
+    def _run_secrets_scan(
+        self,
+        project_dir: Path,
+        changed_files: list[str] | None,
+        result: SecurityScanResult,
+    ) -> None:
+        """Run secrets scanning using scan_secrets.py."""
+        if not HAS_SECRETS_SCANNER:
+            result.scan_errors.append("scan_secrets module not available")
+            return
+
+        try:
+            # Get files to scan
+            if changed_files:
+                files_to_scan = changed_files
+            else:
+                files_to_scan = get_all_tracked_files()
+
+            # Run scan
+            matches = scan_files(files_to_scan, project_dir)
+
+            # Convert matches to result format
+            for match in matches:
+                result.secrets.append(
+                    {
+                        "file": match.file_path,
+                        "line": match.line_number,
+                        "pattern": match.pattern_name,
+                        "matched_text": self._redact_secret(match.matched_text),
+                    }
+                )
+
+                # Also add as vulnerability
+                result.vulnerabilities.append(
+                    SecurityVulnerability(
+                        severity="critical",
+                        source="secrets",
+                        title=f"Potential secret: {match.pattern_name}",
+                        description=f"Found potential {match.pattern_name} in file",
+                        file=match.file_path,
+                        line=match.line_number,
+                    )
+                )
+
+        except Exception as e:
+            result.scan_errors.append(f"Secrets scan error: {str(e)}")
+
+    def _run_sast_scans(self, project_dir: Path, result: SecurityScanResult) -> None:
+        """Run SAST tools based on project type."""
+        # Python SAST with Bandit
+        if self._is_python_project(project_dir):
+            self._run_bandit(project_dir, result)
+
+        # JavaScript/Node.js - npm audit
+        # (handled in dependency audits for Node projects)
+
+    def _run_bandit(self, project_dir: Path, result: SecurityScanResult) -> None:
+        """Run Bandit security scanner for Python projects."""
+        if not self._check_bandit_available():
+            return
+
+        try:
+            # Find Python source directories
+            src_dirs = []
+            for candidate in ["src", "app", project_dir.name, "."]:
+                candidate_path = project_dir / candidate
+                if (
+                    candidate_path.exists()
+                    and (candidate_path / "__init__.py").exists()
+                ):
+                    src_dirs.append(str(candidate_path))
+
+            if not src_dirs:
+                # Try to find any Python files
+                py_files = list(project_dir.glob("**/*.py"))
+                if not py_files:
+                    return
+                src_dirs = ["."]
+
+            # Run bandit
+            cmd = [
+                "bandit",
+                "-r",
+                *src_dirs,
+                "-f",
+                "json",
+                "--exit-zero",  # Don't fail on findings
+            ]
+
+            proc = subprocess.run(
+                cmd,
+                cwd=project_dir,
+                capture_output=True,
+                text=True,
+                timeout=120,
+            )
+
+            if proc.stdout:
+                try:
+                    bandit_output = json.loads(proc.stdout)
+                    for finding in bandit_output.get("results", []):
+                        severity = finding.get("issue_severity", "MEDIUM").lower()
+                        if severity == "high":
+                            severity = "high"
+                        elif severity == "medium":
+                            severity = "medium"
+                        else:
+                            severity = "low"
+
+                        result.vulnerabilities.append(
+                            SecurityVulnerability(
+                                severity=severity,
+                                source="bandit",
+                                title=finding.get("issue_text", "Unknown issue"),
+                                description=finding.get("issue_text", ""),
+                                file=finding.get("filename"),
+                                line=finding.get("line_number"),
+                                cwe=finding.get("issue_cwe", {}).get("id"),
+                            )
+                        )
+                except json.JSONDecodeError:
+                    result.scan_errors.append("Failed to parse Bandit output")
+
+        except subprocess.TimeoutExpired:
+            result.scan_errors.append("Bandit scan timed out")
+        except FileNotFoundError:
+            result.scan_errors.append("Bandit not found")
+        except Exception as e:
+            result.scan_errors.append(f"Bandit error: {str(e)}")
+
+    def _run_dependency_audits(
+        self, project_dir: Path, result: SecurityScanResult
+    ) -> None:
+        """Run dependency vulnerability audits."""
+        # npm audit for JavaScript projects
+        if (project_dir / "package.json").exists():
+            self._run_npm_audit(project_dir, result)
+
+        # pip-audit for Python projects (if available)
+        if self._is_python_project(project_dir):
+            self._run_pip_audit(project_dir, result)
+
+    def _run_npm_audit(self, project_dir: Path, result: SecurityScanResult) -> None:
+        """Run npm audit for JavaScript projects."""
+        try:
+            cmd = ["npm", "audit", "--json"]
+
+            proc = subprocess.run(
+                cmd,
+                cwd=project_dir,
+                capture_output=True,
+                text=True,
+                timeout=120,
+            )
+
+            if proc.stdout:
+                try:
+                    audit_output = json.loads(proc.stdout)
+
+                    # npm audit v2+ format
+                    vulnerabilities = audit_output.get("vulnerabilities", {})
+                    for pkg_name, vuln_info in vulnerabilities.items():
+                        severity = vuln_info.get("severity", "moderate")
+                        if severity == "critical":
+                            severity = "critical"
+                        elif severity == "high":
+                            severity = "high"
+                        elif severity == "moderate":
+                            severity = "medium"
+                        else:
+                            severity = "low"
+
+                        result.vulnerabilities.append(
+                            SecurityVulnerability(
+                                severity=severity,
+                                source="npm_audit",
+                                title=f"Vulnerable dependency: {pkg_name}",
+                                description=vuln_info.get("via", [{}])[0].get(
+                                    "title", ""
+                                )
+                                if isinstance(vuln_info.get("via"), list)
+                                and vuln_info.get("via")
+                                else str(vuln_info.get("via", "")),
+                                file="package.json",
+                            )
+                        )
+                except json.JSONDecodeError:
+                    pass  # npm audit may return invalid JSON on no findings
+
+        except subprocess.TimeoutExpired:
+            result.scan_errors.append("npm audit timed out")
+        except FileNotFoundError:
+            pass  # npm not available
+        except Exception as e:
+            result.scan_errors.append(f"npm audit error: {str(e)}")
+
+    def _run_pip_audit(self, project_dir: Path, result: SecurityScanResult) -> None:
+        """Run pip-audit for Python projects (if available)."""
+        try:
+            cmd = ["pip-audit", "--format", "json"]
+
+            proc = subprocess.run(
+                cmd,
+                cwd=project_dir,
+                capture_output=True,
+                text=True,
+                timeout=120,
+            )
+
+            if proc.stdout:
+                try:
+                    audit_output = json.loads(proc.stdout)
+                    for vuln in audit_output:
+                        severity = "high" if vuln.get("fix_versions") else "medium"
+
+                        result.vulnerabilities.append(
+                            SecurityVulnerability(
+                                severity=severity,
+                                source="pip_audit",
+                                title=f"Vulnerable package: {vuln.get('name')}",
+                                description=vuln.get("description", ""),
+                                cwe=vuln.get("aliases", [""])[0]
+                                if vuln.get("aliases")
+                                else None,
+                            )
+                        )
+                except json.JSONDecodeError:
+                    pass
+
+        except FileNotFoundError:
+            pass  # pip-audit not available
+        except subprocess.TimeoutExpired:
+            pass
+        except Exception:
+            pass
+
+    def _is_python_project(self, project_dir: Path) -> bool:
+        """Check if this is a Python project."""
+        indicators = [
+            project_dir / "pyproject.toml",
+            project_dir / "requirements.txt",
+            project_dir / "setup.py",
+            project_dir / "setup.cfg",
+        ]
+        return any(p.exists() for p in indicators)
+
+    def _check_bandit_available(self) -> bool:
+        """Check if Bandit is available."""
+        if self._bandit_available is None:
+            try:
+                subprocess.run(
+                    ["bandit", "--version"],
+                    capture_output=True,
+                    timeout=5,
+                )
+                self._bandit_available = True
+            except (FileNotFoundError, subprocess.TimeoutExpired):
+                self._bandit_available = False
+        return self._bandit_available
+
+    def _redact_secret(self, text: str) -> str:
+        """Redact a secret for safe logging."""
+        if len(text) <= 8:
+            return "*" * len(text)
+        return text[:4] + "*" * (len(text) - 8) + text[-4:]
+
+    def _save_results(self, spec_dir: Path, result: SecurityScanResult) -> None:
+        """Save scan results to spec directory."""
+        spec_dir = Path(spec_dir)
+        spec_dir.mkdir(parents=True, exist_ok=True)
+
+        output_file = spec_dir / "security_scan_results.json"
+        output_data = self.to_dict(result)
+
+        with open(output_file, "w", encoding="utf-8") as f:
+            json.dump(output_data, f, indent=2)
+
+    def to_dict(self, result: SecurityScanResult) -> dict[str, Any]:
+        """Convert result to dictionary for JSON serialization."""
+        return {
+            "secrets": result.secrets,
+            "vulnerabilities": [
+                {
+                    "severity": v.severity,
+                    "source": v.source,
+                    "title": v.title,
+                    "description": v.description,
+                    "file": v.file,
+                    "line": v.line,
+                    "cwe": v.cwe,
+                }
+                for v in result.vulnerabilities
+            ],
+            "scan_errors": result.scan_errors,
+            "has_critical_issues": result.has_critical_issues,
+            "should_block_qa": result.should_block_qa,
+            "summary": {
+                "total_secrets": len(result.secrets),
+                "total_vulnerabilities": len(result.vulnerabilities),
+                "critical_count": sum(
+                    1 for v in result.vulnerabilities if v.severity == "critical"
+                ),
+                "high_count": sum(
+                    1 for v in result.vulnerabilities if v.severity == "high"
+                ),
+                "medium_count": sum(
+                    1 for v in result.vulnerabilities if v.severity == "medium"
+                ),
+                "low_count": sum(
+                    1 for v in result.vulnerabilities if v.severity == "low"
+                ),
+            },
+        }
+
+
+# =============================================================================
+# CONVENIENCE FUNCTIONS
+# =============================================================================
+
+
+def scan_for_security_issues(
+    project_dir: Path,
+    spec_dir: Path | None = None,
+    changed_files: list[str] | None = None,
+) -> SecurityScanResult:
+    """
+    Convenience function to run security scan.
+
+    Args:
+        project_dir: Path to project root
+        spec_dir: Optional spec directory to save results
+        changed_files: Optional list of files to scan
+
+    Returns:
+        SecurityScanResult with all findings
+    """
+    scanner = SecurityScanner()
+    return scanner.scan(project_dir, spec_dir, changed_files)
+
+
+def has_security_issues(project_dir: Path) -> bool:
+    """
+    Quick check if project has security issues.
+
+    Args:
+        project_dir: Path to project root
+
+    Returns:
+        True if any critical/high issues found
+    """
+    scanner = SecurityScanner()
+    result = scanner.scan(project_dir, run_sast=False, run_dependency_audit=False)
+    return result.has_critical_issues
+
+
+def scan_secrets_only(
+    project_dir: Path,
+    changed_files: list[str] | None = None,
+) -> list[dict[str, Any]]:
+    """
+    Scan only for secrets (quick scan).
+
+    Args:
+        project_dir: Path to project root
+        changed_files: Optional list of files to scan
+
+    Returns:
+        List of detected secrets
+    """
+    scanner = SecurityScanner()
+    result = scanner.scan(
+        project_dir,
+        changed_files=changed_files,
+        run_sast=False,
+        run_dependency_audit=False,
+    )
+    return result.secrets
+
+
+# =============================================================================
+# CLI
+# =============================================================================
+
+
+def main() -> None:
+    """CLI entry point for testing."""
+    import argparse
+
+    parser = argparse.ArgumentParser(description="Run security scans")
+    parser.add_argument("project_dir", type=Path, help="Path to project root")
+    parser.add_argument("--spec-dir", type=Path, help="Path to spec directory")
+    parser.add_argument(
+        "--secrets-only", action="store_true", help="Only scan for secrets"
+    )
+    parser.add_argument("--json", action="store_true", help="Output as JSON")
+
+    args = parser.parse_args()
+
+    scanner = SecurityScanner()
+    result = scanner.scan(
+        args.project_dir,
+        spec_dir=args.spec_dir,
+        run_sast=not args.secrets_only,
+        run_dependency_audit=not args.secrets_only,
+    )
+
+    if args.json:
+        print(json.dumps(scanner.to_dict(result), indent=2))
+    else:
+        print(f"Secrets Found: {len(result.secrets)}")
+        print(f"Vulnerabilities: {len(result.vulnerabilities)}")
+        print(f"Has Critical Issues: {result.has_critical_issues}")
+        print(f"Should Block QA: {result.should_block_qa}")
+
+        if result.secrets:
+            print("\nSecrets Detected:")
+            for secret in result.secrets:
+                print(f"  - {secret['pattern']} in {secret['file']}:{secret['line']}")
+
+        if result.vulnerabilities:
+            print(f"\nVulnerabilities ({len(result.vulnerabilities)}):")
+            for v in result.vulnerabilities:
+                print(f"  [{v.severity.upper()}] {v.title}")
+                if v.file:
+                    print(f"    File: {v.file}:{v.line or ''}")
+
+        if result.scan_errors:
+            print(f"\nScan Errors ({len(result.scan_errors)}):")
+            for error in result.scan_errors:
+                print(f"  - {error}")
+
+
+if __name__ == "__main__":
+    main()
@@ -0,0 +1,26 @@
+#!/usr/bin/env python3
+"""
+Analyzer facade module.
+
+Provides backward compatibility for scripts that import from analyzer.py at the root.
+Actual implementation is in analysis/analyzer.py.
+"""
+
+from analysis.analyzer import (
+    ProjectAnalyzer,
+    ServiceAnalyzer,
+    analyze_project,
+    analyze_service,
+    main,
+)
+
+__all__ = [
+    "ServiceAnalyzer",
+    "ProjectAnalyzer",
+    "analyze_project",
+    "analyze_service",
+    "main",
+]
+
+if __name__ == "__main__":
+    main()
@@ -0,0 +1,36 @@
+"""
+Auto Claude tools module facade.
+
+Provides MCP tools for agent operations.
+Re-exports from agents.tools_pkg for clean imports.
+"""
+
+from agents.tools_pkg.models import (  # noqa: F401
+    ELECTRON_TOOLS,
+    TOOL_GET_BUILD_PROGRESS,
+    TOOL_GET_SESSION_CONTEXT,
+    TOOL_RECORD_DISCOVERY,
+    TOOL_RECORD_GOTCHA,
+    TOOL_UPDATE_QA_STATUS,
+    TOOL_UPDATE_SUBTASK_STATUS,
+    is_electron_mcp_enabled,
+)
+from agents.tools_pkg.permissions import get_allowed_tools  # noqa: F401
+from agents.tools_pkg.registry import (  # noqa: F401
+    create_auto_claude_mcp_server,
+    is_tools_available,
+)
+
+__all__ = [
+    "create_auto_claude_mcp_server",
+    "get_allowed_tools",
+    "is_tools_available",
+    "TOOL_UPDATE_SUBTASK_STATUS",
+    "TOOL_GET_BUILD_PROGRESS",
+    "TOOL_RECORD_DISCOVERY",
+    "TOOL_RECORD_GOTCHA",
+    "TOOL_GET_SESSION_CONTEXT",
+    "TOOL_UPDATE_QA_STATUS",
+    "ELECTRON_TOOLS",
+    "is_electron_mcp_enabled",
+]
@@ -0,0 +1,21 @@
+"""Backward compatibility shim - import from analysis.ci_discovery instead."""
+
+from analysis.ci_discovery import (
+    HAS_YAML,
+    CIConfig,
+    CIDiscovery,
+    CIWorkflow,
+    discover_ci,
+    get_ci_system,
+    get_ci_test_commands,
+)
+
+__all__ = [
+    "CIConfig",
+    "CIWorkflow",
+    "CIDiscovery",
+    "discover_ci",
+    "get_ci_test_commands",
+    "get_ci_system",
+    "HAS_YAML",
+]
@@ -0,0 +1,18 @@
+"""
+Auto Claude CLI Package
+=======================
+
+Command-line interface for the Auto Claude autonomous coding framework.
+
+This package provides a modular CLI structure:
+- main.py: Argument parsing and command routing
+- spec_commands.py: Spec listing and management
+- build_commands.py: Build execution and follow-up tasks
+- workspace_commands.py: Workspace management (merge, review, discard)
+- qa_commands.py: QA validation commands
+- utils.py: Shared utilities and configuration
+"""
+
+from .main import main
+
+__all__ = ["main"]
@@ -0,0 +1,279 @@
+"""
+Batch Task Management Commands
+==============================
+
+Commands for creating and managing multiple tasks from batch files.
+"""
+
+import json
+import shutil
+import subprocess
+from pathlib import Path
+
+from qa.criteria import is_fixes_applied, is_qa_approved, is_qa_rejected
+from ui import highlight, print_status
+
+
+def handle_batch_create_command(batch_file: str, project_dir: str) -> bool:
+    """
+    Create multiple tasks from a batch JSON file.
+
+    Args:
+        batch_file: Path to JSON file with task definitions
+        project_dir: Project directory
+
+    Returns:
+        True if successful
+    """
+    batch_path = Path(batch_file)
+
+    if not batch_path.exists():
+        print_status(f"Batch file not found: {batch_file}", "error")
+        return False
+
+    try:
+        with open(batch_path, encoding="utf-8") as f:
+            batch_data = json.load(f)
+    except json.JSONDecodeError as e:
+        print_status(f"Invalid JSON in batch file: {e}", "error")
+        return False
+
+    tasks = batch_data.get("tasks", [])
+    if not tasks:
+        print_status("No tasks found in batch file", "warning")
+        return False
+
+    print_status(f"Creating {len(tasks)} tasks from batch file", "info")
+    print()
+
+    specs_dir = Path(project_dir) / ".auto-claude" / "specs"
+    specs_dir.mkdir(parents=True, exist_ok=True)
+
+    # Find next spec ID
+    existing_specs = [d.name for d in specs_dir.iterdir() if d.is_dir()]
+    next_id = (
+        max([int(s.split("-")[0]) for s in existing_specs if s[0].isdigit()] or [0]) + 1
+    )
+
+    created_specs = []
+
+    for idx, task in enumerate(tasks, 1):
+        spec_id = f"{next_id:03d}"
+        task_title = task.get("title", f"Task {idx}")
+        task_slug = task_title.lower().replace(" ", "-")[:50]
+        spec_name = f"{spec_id}-{task_slug}"
+        spec_dir = specs_dir / spec_name
+        spec_dir.mkdir(exist_ok=True)
+
+        # Create requirements.json
+        requirements = {
+            "task_description": task.get("description", task_title),
+            "description": task.get("description", task_title),
+            "workflow_type": task.get("workflow_type", "feature"),
+            "services_involved": task.get("services", ["frontend"]),
+            "priority": task.get("priority", 5),
+            "complexity_inferred": task.get("complexity", "standard"),
+            "inferred_from": {},
+            "created_at": Path(spec_dir).stat().st_mtime,
+            "estimate": {
+                "estimated_hours": task.get("estimated_hours", 4.0),
+                "estimated_days": task.get("estimated_days", 0.5),
+            },
+        }
+
+        req_file = spec_dir / "requirements.json"
+        with open(req_file, "w", encoding="utf-8") as f:
+            json.dump(requirements, f, indent=2, default=str)
+
+        created_specs.append(
+            {
+                "id": spec_id,
+                "name": spec_name,
+                "title": task_title,
+                "status": "pending_spec_creation",
+            }
+        )
+
+        print_status(
+            f"[{idx}/{len(tasks)}] Created {spec_id} - {task_title}", "success"
+        )
+        next_id += 1
+
+    print()
+    print_status(f"Created {len(created_specs)} spec(s) successfully", "success")
+    print()
+
+    # Show summary
+    print(highlight("Next steps:"))
+    print("  1. Generate specs: spec_runner.py --continue <spec_id>")
+    print("  2. Approve specs and build them")
+    print("  3. Run: python run.py --spec <id> to execute")
+
+    return True
+
+
+def handle_batch_status_command(project_dir: str) -> bool:
+    """
+    Show status of all specs in project.
+
+    Args:
+        project_dir: Project directory
+
+    Returns:
+        True if successful
+    """
+    specs_dir = Path(project_dir) / ".auto-claude" / "specs"
+
+    if not specs_dir.exists():
+        print_status("No specs found in project", "warning")
+        return True
+
+    specs = sorted([d for d in specs_dir.iterdir() if d.is_dir()])
+
+    if not specs:
+        print_status("No specs found", "warning")
+        return True
+
+    print_status(f"Found {len(specs)} spec(s)", "info")
+    print()
+
+    for spec_dir in specs:
+        spec_name = spec_dir.name
+        req_file = spec_dir / "requirements.json"
+
+        status = "unknown"
+        title = spec_name
+
+        if req_file.exists():
+            try:
+                with open(req_file, encoding="utf-8") as f:
+                    req = json.load(f)
+                    title = req.get("task_description", title)
+            except json.JSONDecodeError:
+                pass
+
+        # Determine status (highest priority first)
+        # Use authoritative QA status check, not just file existence
+        if is_qa_approved(spec_dir):
+            status = "qa_approved"
+        elif is_qa_rejected(spec_dir):
+            status = "qa_rejected"
+        elif is_fixes_applied(spec_dir):
+            status = "fixes_applied"
+        elif (spec_dir / "implementation_plan.json").exists():
+            # Check if there's a qa_report.md but no approval yet (QA in progress)
+            if (spec_dir / "qa_report.md").exists():
+                status = "qa_in_progress"
+            else:
+                status = "building"
+        elif (spec_dir / "spec.md").exists():
+            status = "spec_created"
+        else:
+            status = "pending_spec"
+
+        status_icon = {
+            "pending_spec": "⏳",
+            "spec_created": "📋",
+            "building": "⚙️",
+            "qa_in_progress": "🔍",
+            "qa_approved": "✅",
+            "qa_rejected": "❌",
+            "fixes_applied": "🔧",
+            "unknown": "❓",
+        }.get(status, "❓")
+
+        print(f"{status_icon} {spec_name:<40} {title}")
+
+    return True
+
+
+def handle_batch_cleanup_command(project_dir: str, dry_run: bool = True) -> bool:
+    """
+    Clean up completed specs and worktrees.
+
+    Args:
+        project_dir: Project directory
+        dry_run: If True, show what would be deleted
+
+    Returns:
+        True if successful
+    """
+    specs_dir = Path(project_dir) / ".auto-claude" / "specs"
+    worktrees_dir = Path(project_dir) / ".auto-claude" / "worktrees" / "tasks"
+
+    if not specs_dir.exists():
+        print_status("No specs directory found", "info")
+        return True
+
+    # Find completed specs (only QA-approved, matching status display logic)
+    completed = []
+    for spec_dir in specs_dir.iterdir():
+        if spec_dir.is_dir() and is_qa_approved(spec_dir):
+            completed.append(spec_dir.name)
+
+    if not completed:
+        print_status("No completed specs to clean up", "info")
+        return True
+
+    print_status(f"Found {len(completed)} completed spec(s)", "info")
+
+    if dry_run:
+        print()
+        print("Would remove:")
+        for spec_name in completed:
+            print(f"  - {spec_name}")
+            wt_path = worktrees_dir / spec_name
+            if wt_path.exists():
+                print(f"    └─ .auto-claude/worktrees/tasks/{spec_name}/")
+        print()
+        print("Run with --no-dry-run to actually delete")
+    else:
+        # Actually delete specs and worktrees
+        deleted_count = 0
+        for spec_name in completed:
+            spec_path = specs_dir / spec_name
+            wt_path = worktrees_dir / spec_name
+
+            # Remove worktree first (if exists)
+            if wt_path.exists():
+                try:
+                    result = subprocess.run(
+                        ["git", "worktree", "remove", "--force", str(wt_path)],
+                        cwd=project_dir,
+                        capture_output=True,
+                        text=True,
+                        timeout=30,
+                    )
+                    if result.returncode == 0:
+                        print_status(f"Removed worktree: {spec_name}", "success")
+                    else:
+                        # Fallback: remove directory manually if git fails
+                        shutil.rmtree(wt_path, ignore_errors=True)
+                        print_status(
+                            f"Removed worktree directory: {spec_name}", "success"
+                        )
+                except subprocess.TimeoutExpired:
+                    # Timeout: fall back to manual removal
+                    shutil.rmtree(wt_path, ignore_errors=True)
+                    print_status(
+                        f"Worktree removal timed out, removed directory: {spec_name}",
+                        "warning",
+                    )
+                except Exception as e:
+                    print_status(
+                        f"Failed to remove worktree {spec_name}: {e}", "warning"
+                    )
+
+            # Remove spec directory
+            if spec_path.exists():
+                try:
+                    shutil.rmtree(spec_path)
+                    print_status(f"Removed spec: {spec_name}", "success")
+                    deleted_count += 1
+                except Exception as e:
+                    print_status(f"Failed to remove spec {spec_name}: {e}", "error")
+
+        print()
+        print_status(f"Cleaned up {deleted_count} spec(s)", "info")
+
+    return True
@@ -0,0 +1,740 @@
+"""
+Build Commands
+==============
+
+CLI commands for building specs and handling the main build flow.
+"""
+
+import asyncio
+import sys
+from pathlib import Path
+
+# Ensure parent directory is in path for imports (before other imports)
+_PARENT_DIR = Path(__file__).parent.parent
+if str(_PARENT_DIR) not in sys.path:
+    sys.path.insert(0, str(_PARENT_DIR))
+
+# Import only what we need at module level
+# Heavy imports are lazy-loaded in functions to avoid import errors
+from progress import print_paused_banner
+from review import ReviewState
+from ui import (
+    BuildState,
+    Icons,
+    MenuOption,
+    StatusManager,
+    bold,
+    box,
+    highlight,
+    icon,
+    muted,
+    print_status,
+    select_menu,
+    success,
+    warning,
+)
+from workspace import (
+    WorkspaceMode,
+    check_existing_build,
+    choose_workspace,
+    finalize_workspace,
+    get_existing_build_worktree,
+    handle_workspace_choice,
+    setup_workspace,
+)
+
+from .input_handlers import (
+    read_from_file,
+    read_multiline_input,
+)
+
+
+def handle_build_command(
+    project_dir: Path,
+    spec_dir: Path,
+    model: str,
+    max_iterations: int | None,
+    verbose: bool,
+    force_isolated: bool,
+    force_direct: bool,
+    auto_continue: bool,
+    skip_qa: bool,
+    force_bypass_approval: bool,
+    base_branch: str | None = None,
+    issue_workflow: bool = False,
+    issue_number: int | None = None,
+) -> None:
+    """
+    Handle the main build command.
+
+    Args:
+        project_dir: Project root directory
+        spec_dir: Spec directory path
+        model: Model to use (used as default; may be overridden by task_metadata.json)
+        max_iterations: Maximum number of iterations (None for unlimited)
+        verbose: Enable verbose output
+        force_isolated: Force isolated workspace mode
+        force_direct: Force direct workspace mode
+        auto_continue: Auto-continue mode (non-interactive)
+        skip_qa: Skip automatic QA validation
+        force_bypass_approval: Force bypass approval check
+        base_branch: Base branch for worktree creation (default: current branch)
+        issue_workflow: If True, run from a GitHub issue investigation
+        issue_number: GitHub issue number (required when issue_workflow=True)
+    """
+    # Lazy imports to avoid loading heavy modules
+    from agent import run_autonomous_agent, sync_spec_to_source
+    from debug import (
+        debug,
+        debug_info,
+        debug_section,
+        debug_success,
+    )
+    from phase_config import get_phase_model
+    from prompts_pkg.prompts import (
+        get_base_branch_from_metadata,
+        get_use_local_branch_from_metadata,
+    )
+    from qa_loop import run_qa_validation_loop, should_run_qa
+
+    from .utils import print_banner, validate_environment
+
+    # Handle issue workflow: load investigation report and inject context
+    if issue_workflow:
+        if not issue_number:
+            print("\nError: --issue-number is required with --issue-workflow")
+            sys.exit(1)
+
+        pipeline_mode = _get_investigation_pipeline_mode(project_dir)
+        _inject_issue_workflow_context(project_dir, spec_dir, issue_number)
+
+        # pipelineMode controls which phases to skip:
+        # - "full": run complete spec + planning + coding + QA pipeline (default)
+        # - "skip_to_planning": skip spec creation, go to planning (investigation = spec)
+        # - "minimal": skip spec + planning, go straight to coding
+        if pipeline_mode == "skip_to_planning":
+            # Investigation report serves as the spec; bypass approval since
+            # the investigation was already reviewed.
+            force_bypass_approval = True
+        elif pipeline_mode == "minimal":
+            # Skip everything: create a minimal plan so the planner is bypassed
+            # and the coder starts immediately from the investigation context.
+            force_bypass_approval = True
+            skip_qa = True
+            _create_minimal_plan_for_issue(spec_dir, issue_number)
+
+    # Get the resolved model for the planning phase (first phase of build)
+    # This respects task_metadata.json phase configuration from the UI
+    planning_model = get_phase_model(spec_dir, "planning", model)
+    coding_model = get_phase_model(spec_dir, "coding", model)
+    qa_model = get_phase_model(spec_dir, "qa", model)
+
+    print_banner()
+    print(f"\nProject directory: {project_dir}")
+    print(f"Spec: {spec_dir.name}")
+    # Show phase-specific models if they differ
+    if planning_model != coding_model or coding_model != qa_model:
+        print(
+            f"Models: Planning={planning_model.split('-')[1] if '-' in planning_model else planning_model}, "
+            f"Coding={coding_model.split('-')[1] if '-' in coding_model else coding_model}, "
+            f"QA={qa_model.split('-')[1] if '-' in qa_model else qa_model}"
+        )
+    else:
+        print(f"Model: {planning_model}")
+
+    if max_iterations:
+        print(f"Max iterations: {max_iterations}")
+    else:
+        print("Max iterations: Unlimited (runs until all subtasks complete)")
+
+    print()
+
+    # Validate environment
+    if not validate_environment(spec_dir):
+        sys.exit(1)
+
+    # Check human review approval
+    review_state = ReviewState.load(spec_dir)
+    if not review_state.is_approval_valid(spec_dir):
+        if force_bypass_approval:
+            # User explicitly bypassed approval check
+            print()
+            print(
+                warning(
+                    f"{icon(Icons.WARNING)} WARNING: Bypassing approval check with --force"
+                )
+            )
+            print(muted("This spec has not been approved for building."))
+            print()
+        else:
+            print()
+            content = [
+                bold(f"{icon(Icons.WARNING)} BUILD BLOCKED - REVIEW REQUIRED"),
+                "",
+                "This spec requires human approval before building.",
+            ]
+
+            if review_state.approved and not review_state.is_approval_valid(spec_dir):
+                # Spec changed after approval
+                content.append("")
+                content.append(warning("The spec has been modified since approval."))
+                content.append("Please re-review and re-approve.")
+
+            content.extend(
+                [
+                    "",
+                    highlight("To review and approve:"),
+                    f"  python auto-claude/review.py --spec-dir {spec_dir}",
+                    "",
+                    muted("Or use --force to bypass this check (not recommended)."),
+                ]
+            )
+            print(box(content, width=70, style="heavy"))
+            print()
+            sys.exit(1)
+    else:
+        debug_success(
+            "run.py", "Review approval validated", approved_by=review_state.approved_by
+        )
+
+    # Check for existing build
+    if get_existing_build_worktree(project_dir, spec_dir.name):
+        if auto_continue:
+            # Non-interactive mode: auto-continue with existing build
+            debug("run.py", "Auto-continue mode: continuing with existing build")
+            print("Auto-continue: Resuming existing build...")
+        else:
+            continue_existing = check_existing_build(project_dir, spec_dir.name)
+            if continue_existing:
+                # Continue with existing worktree
+                pass
+            else:
+                # User chose to start fresh or merged existing
+                pass
+
+    # Choose workspace (skip for parallel mode - it always uses worktrees)
+    working_dir = project_dir
+    worktree_manager = None
+    source_spec_dir = None  # Track original spec dir for syncing back from worktree
+
+    # Let user choose workspace mode (or auto-select if --auto-continue)
+    workspace_mode = choose_workspace(
+        project_dir,
+        spec_dir.name,
+        force_isolated=force_isolated,
+        force_direct=force_direct,
+        auto_continue=auto_continue,
+    )
+
+    # If base_branch not provided via CLI, try to read from task_metadata.json
+    # This ensures the backend uses the branch configured in the frontend
+    if base_branch is None:
+        metadata_branch = get_base_branch_from_metadata(spec_dir)
+        if metadata_branch:
+            base_branch = metadata_branch
+            debug("run.py", f"Using base branch from task metadata: {base_branch}")
+
+    # Check if user requested local branch (preserves gitignored files like .env)
+    use_local_branch = get_use_local_branch_from_metadata(spec_dir)
+
+    if workspace_mode == WorkspaceMode.ISOLATED:
+        # Keep reference to original spec directory for syncing progress back
+        source_spec_dir = spec_dir
+
+        working_dir, worktree_manager, localized_spec_dir = setup_workspace(
+            project_dir,
+            spec_dir.name,
+            workspace_mode,
+            source_spec_dir=spec_dir,
+            base_branch=base_branch,
+            use_local_branch=use_local_branch,
+        )
+        # Use the localized spec directory (inside worktree) for AI access
+        if localized_spec_dir:
+            spec_dir = localized_spec_dir
+
+    # Run the autonomous agent
+    debug_section("run.py", "Starting Build Execution")
+    debug(
+        "run.py",
+        "Build configuration",
+        model=model,
+        workspace_mode=str(workspace_mode),
+        working_dir=str(working_dir),
+        spec_dir=str(spec_dir),
+    )
+
+    try:
+        debug("run.py", "Starting agent execution")
+
+        asyncio.run(
+            run_autonomous_agent(
+                project_dir=working_dir,  # Use worktree if isolated
+                spec_dir=spec_dir,
+                model=model,
+                max_iterations=max_iterations,
+                verbose=verbose,
+                source_spec_dir=source_spec_dir,  # For syncing progress back to main project
+            )
+        )
+        debug_success("run.py", "Agent execution completed")
+
+        # Run QA validation BEFORE finalization (while worktree still exists)
+        # QA must sign off before the build is considered complete
+        qa_approved = True  # Default to approved if QA is skipped
+        if not skip_qa and should_run_qa(spec_dir):
+            print("\n" + "=" * 70)
+            print("  SUBTASKS COMPLETE - STARTING QA VALIDATION")
+            print("=" * 70)
+            print("\nAll subtasks completed. Now running QA validation loop...")
+            print("This ensures production-quality output before sign-off.\n")
+
+            try:
+                qa_approved = asyncio.run(
+                    run_qa_validation_loop(
+                        project_dir=working_dir,
+                        spec_dir=spec_dir,
+                        model=model,
+                        verbose=verbose,
+                    )
+                )
+
+                if qa_approved:
+                    print("\n" + "=" * 70)
+                    print("  ✅ QA VALIDATION PASSED")
+                    print("=" * 70)
+                    print("\nAll acceptance criteria verified.")
+                    print("The implementation is production-ready.\n")
+                else:
+                    print("\n" + "=" * 70)
+                    print("  ⚠️  QA VALIDATION INCOMPLETE")
+                    print("=" * 70)
+                    print("\nSome issues require manual attention.")
+                    print(f"See: {spec_dir / 'qa_report.md'}")
+                    print(f"Or:  {spec_dir / 'QA_FIX_REQUEST.md'}")
+                    print(
+                        f"\nResume QA: python auto-claude/run.py --spec {spec_dir.name} --qa\n"
+                    )
+
+                # Sync implementation plan to main project after QA
+                # This ensures the main project has the latest status (human_review)
+                if sync_spec_to_source(spec_dir, source_spec_dir):
+                    debug_info(
+                        "run.py", "Implementation plan synced to main project after QA"
+                    )
+            except KeyboardInterrupt:
+                print("\n\nQA validation paused.")
+                print(f"Resume: python auto-claude/run.py --spec {spec_dir.name} --qa")
+                qa_approved = False
+
+        # Post-build finalization (only for isolated sequential mode)
+        # This happens AFTER QA validation so the worktree still exists
+        if worktree_manager:
+            choice = finalize_workspace(
+                project_dir,
+                spec_dir.name,
+                worktree_manager,
+                auto_continue=auto_continue,
+            )
+            handle_workspace_choice(
+                choice, project_dir, spec_dir.name, worktree_manager
+            )
+
+    except KeyboardInterrupt:
+        _handle_build_interrupt(
+            spec_dir=spec_dir,
+            project_dir=project_dir,
+            worktree_manager=worktree_manager,
+            working_dir=working_dir,
+            model=model,
+            max_iterations=max_iterations,
+            verbose=verbose,
+        )
+    except Exception as e:
+        print(f"\nFatal error: {e}")
+        if verbose:
+            import traceback
+
+            traceback.print_exc()
+        sys.exit(1)
+
+
+def _handle_build_interrupt(
+    spec_dir: Path,
+    project_dir: Path,
+    worktree_manager,
+    working_dir: Path,
+    model: str,
+    max_iterations: int | None,
+    verbose: bool,
+) -> None:
+    """
+    Handle keyboard interrupt during build.
+
+    Args:
+        spec_dir: Spec directory path
+        project_dir: Project root directory
+        worktree_manager: Worktree manager instance (if using isolated mode)
+        working_dir: Current working directory
+        model: Model being used
+        max_iterations: Maximum iterations
+        verbose: Verbose mode flag
+    """
+    from agent import run_autonomous_agent
+
+    # Print paused banner
+    print_paused_banner(spec_dir, spec_dir.name, has_worktree=bool(worktree_manager))
+
+    # Update status file
+    status_manager = StatusManager(project_dir)
+    status_manager.update(state=BuildState.PAUSED)
+
+    # Offer to add human input with enhanced menu
+    try:
+        options = [
+            MenuOption(
+                key="type",
+                label="Type instructions",
+                icon=Icons.EDIT,
+                description="Enter guidance for the agent's next session",
+            ),
+            MenuOption(
+                key="paste",
+                label="Paste from clipboard",
+                icon=Icons.CLIPBOARD,
+                description="Paste text you've copied (Cmd+V / Ctrl+Shift+V)",
+            ),
+            MenuOption(
+                key="file",
+                label="Read from file",
+                icon=Icons.DOCUMENT,
+                description="Load instructions from a text file",
+            ),
+            MenuOption(
+                key="skip",
+                label="Continue without instructions",
+                icon=Icons.SKIP,
+                description="Resume the build as-is",
+            ),
+            MenuOption(
+                key="quit",
+                label="Quit",
+                icon=Icons.DOOR,
+                description="Exit without resuming",
+            ),
+        ]
+
+        choice = select_menu(
+            title="What would you like to do?",
+            options=options,
+            subtitle="Progress saved. You can add instructions for the agent.",
+            allow_quit=False,  # We have explicit quit option
+        )
+
+        if choice == "quit" or choice is None:
+            print()
+            print_status("Exiting...", "info")
+            status_manager.set_inactive()
+            sys.exit(0)
+
+        human_input = ""
+
+        if choice == "file":
+            # Read from file
+            human_input = read_from_file()
+            if human_input is None:
+                human_input = ""
+
+        elif choice in ["type", "paste"]:
+            human_input = read_multiline_input("Enter/paste your instructions below.")
+            if human_input is None:
+                print()
+                print_status("Exiting without saving instructions...", "warning")
+                status_manager.set_inactive()
+                sys.exit(0)
+
+        if human_input:
+            # Save to HUMAN_INPUT.md
+            input_file = spec_dir / "HUMAN_INPUT.md"
+            input_file.write_text(human_input, encoding="utf-8")
+
+            content = [
+                success(f"{icon(Icons.SUCCESS)} INSTRUCTIONS SAVED"),
+                "",
+                f"Saved to: {highlight(str(input_file.name))}",
+                "",
+                muted(
+                    "The agent will read and follow these instructions when you resume."
+                ),
+            ]
+            print()
+            print(box(content, width=70, style="heavy"))
+        elif choice != "skip":
+            print()
+            print_status("No instructions provided.", "info")
+
+        # If 'skip' was selected, actually resume the build
+        if choice == "skip":
+            print()
+            print_status("Resuming build...", "info")
+            status_manager.update(state=BuildState.BUILDING)
+            asyncio.run(
+                run_autonomous_agent(
+                    project_dir=working_dir,
+                    spec_dir=spec_dir,
+                    model=model,
+                    max_iterations=max_iterations,
+                    verbose=verbose,
+                )
+            )
+            # Build completed or was interrupted again - exit
+            sys.exit(0)
+
+    except KeyboardInterrupt:
+        # User pressed Ctrl+C again during input prompt - exit immediately
+        print()
+        print_status("Exiting...", "warning")
+        status_manager = StatusManager(project_dir)
+        status_manager.set_inactive()
+        sys.exit(0)
+    except EOFError:
+        # stdin closed
+        pass
+
+    # Resume instructions (shown when user provided instructions or chose file/type/paste)
+    print()
+    content = [
+        bold(f"{icon(Icons.PLAY)} TO RESUME"),
+        "",
+        f"Run: {highlight(f'python auto-claude/run.py --spec {spec_dir.name}')}",
+    ]
+    if worktree_manager:
+        content.append("")
+        content.append(muted("Your build is in a separate workspace and is safe."))
+    print(box(content, width=70, style="light"))
+    print()
+
+
+def _create_minimal_plan_for_issue(spec_dir: Path, issue_number: int) -> None:
+    """Create a minimal implementation plan so the planner phase is skipped.
+
+    Used in "minimal" pipeline mode where the investigation context is
+    sufficient and we want the coder to start immediately.
+
+    Args:
+        spec_dir: Spec directory path
+        issue_number: GitHub issue number for context
+    """
+    from datetime import datetime
+
+    plan_file = spec_dir / "implementation_plan.json"
+    if plan_file.exists():
+        # Don't overwrite an existing plan (e.g. if resuming)
+        return
+
+    plan = {
+        "phases": [
+            {
+                "phase": 1,
+                "name": "Implementation",
+                "description": f"Implement fix for issue #{issue_number} based on investigation findings",
+                "depends_on": [],
+                "subtasks": [
+                    {
+                        "id": "subtask-1-1",
+                        "description": (
+                            f"Implement the fix for issue #{issue_number}. "
+                            "Follow the investigation context in HUMAN_INPUT.md "
+                            "for root cause analysis, recommended fix approach, "
+                            "and files to modify."
+                        ),
+                        "service": "main",
+                        "status": "pending",
+                        "files_to_create": [],
+                        "files_to_modify": [],
+                        "patterns_from": [],
+                        "verification": {
+                            "type": "manual",
+                            "run": "Verify the fix resolves the issue",
+                        },
+                    }
+                ],
+            }
+        ],
+        "metadata": {
+            "created_at": datetime.now().isoformat(),
+            "complexity": "simple",
+            "estimated_sessions": 1,
+            "pipeline_mode": "minimal",
+            "source_issue": issue_number,
+        },
+    }
+
+    from core.file_utils import write_json_atomic
+
+    write_json_atomic(plan_file, plan)
+    print("  Pipeline mode: minimal (skipping planner, created single-subtask plan)")
+
+
+def _get_investigation_pipeline_mode(project_dir: Path) -> str:
+    """Read the pipelineMode from investigation settings.
+
+    Reads from .auto-claude/github/config.json -> investigation_settings.pipelineMode.
+    Defaults to "full" if not configured.
+
+    Args:
+        project_dir: Project root directory
+
+    Returns:
+        Pipeline mode string: "full", "skip_to_planning", or "minimal"
+    """
+    import json
+
+    config_path = project_dir / ".auto-claude" / "github" / "config.json"
+    if not config_path.exists():
+        return "full"
+
+    try:
+        data = json.loads(config_path.read_text(encoding="utf-8"))
+        settings = data.get("investigation_settings", {})
+        mode = settings.get("pipelineMode", "full")
+        if mode in ("full", "skip_to_planning", "minimal"):
+            return mode
+        return "full"
+    except (json.JSONDecodeError, OSError):
+        return "full"
+
+
+def _inject_issue_workflow_context(
+    project_dir: Path,
+    spec_dir: Path,
+    issue_number: int,
+) -> None:
+    """Inject investigation context into the build workflow.
+
+    Loads the investigation report for the given issue and writes a
+    HUMAN_INPUT.md file in the spec directory with root cause analysis,
+    fix advice, and other investigation context. This is read by the
+    coder agent as additional guidance.
+
+    Also updates the investigation state to "building".
+
+    Args:
+        project_dir: Project root directory
+        spec_dir: Spec directory path
+        issue_number: GitHub issue number
+    """
+
+    # Use try/except for imports matching the codebase pattern
+    try:
+        from runners.github.services.investigation_persistence import (
+            load_investigation_report,
+            save_investigation_state,
+        )
+    except (ImportError, ValueError, SystemError):
+        # Add parent to path if needed
+        _backend = Path(__file__).parent.parent
+        if str(_backend) not in sys.path:
+            sys.path.insert(0, str(_backend))
+        from runners.github.services.investigation_persistence import (
+            load_investigation_report,
+            save_investigation_state,
+        )
+
+    report = load_investigation_report(project_dir, issue_number)
+    if report is None:
+        print(f"\nWarning: No investigation report found for issue #{issue_number}")
+        print("Proceeding without investigation context.")
+        return
+
+    # Build context string for the coder agent
+    context_parts: list[str] = []
+    context_parts.append(f"# Investigation Context for Issue #{issue_number}")
+    context_parts.append("")
+    context_parts.append(f"## {report.issue_title}")
+    context_parts.append("")
+    context_parts.append(f"**Severity:** {report.severity}")
+    context_parts.append("")
+
+    # AI summary
+    context_parts.append("## Summary")
+    context_parts.append(report.ai_summary)
+    context_parts.append("")
+
+    # Root cause
+    context_parts.append("## Root Cause")
+    context_parts.append(report.root_cause.identified_root_cause)
+    context_parts.append("")
+
+    if report.root_cause.code_paths:
+        context_parts.append("### Code Paths")
+        for cp in report.root_cause.code_paths:
+            end = cp.end_line if cp.end_line else cp.start_line
+            context_parts.append(
+                f"- `{cp.file}:{cp.start_line}-{end}`: {cp.description}"
+            )
+        context_parts.append("")
+
+    # Fix advice
+    if report.fix_advice.approaches:
+        rec_idx = report.fix_advice.recommended_approach
+        context_parts.append("## Recommended Fix")
+        if 0 <= rec_idx < len(report.fix_advice.approaches):
+            approach = report.fix_advice.approaches[rec_idx]
+            context_parts.append(approach.description)
+            context_parts.append("")
+            if approach.files_affected:
+                context_parts.append("**Files to modify:**")
+                for f in approach.files_affected:
+                    context_parts.append(f"- `{f}`")
+                context_parts.append("")
+
+    # Gotchas
+    if report.fix_advice.gotchas:
+        context_parts.append("## Gotchas")
+        for gotcha in report.fix_advice.gotchas:
+            context_parts.append(f"- {gotcha}")
+        context_parts.append("")
+
+    # Patterns
+    if report.fix_advice.patterns_to_follow:
+        context_parts.append("## Patterns to Follow")
+        for pat in report.fix_advice.patterns_to_follow:
+            context_parts.append(f"- `{pat.file}`: {pat.description}")
+        context_parts.append("")
+
+    context = "\n".join(context_parts)
+
+    # Write as HUMAN_INPUT.md (existing mechanism for agent guidance injection)
+    input_file = spec_dir / "HUMAN_INPUT.md"
+    existing = ""
+    if input_file.exists():
+        existing = input_file.read_text(encoding="utf-8")
+
+    if existing:
+        # Append investigation context to existing human input
+        combined = existing + "\n\n" + context
+    else:
+        combined = context
+
+    input_file.write_text(combined, encoding="utf-8")
+
+    # Update investigation state to "building" (note: we use the broader
+    # "task_created" status since "building" isn't a valid InvestigationState status)
+    from datetime import datetime, timezone
+
+    save_investigation_state(
+        project_dir,
+        issue_number,
+        {
+            "issue_number": issue_number,
+            "status": "task_created",
+            "started_at": datetime.now(timezone.utc).isoformat(),
+            "linked_spec_id": spec_dir.name,
+        },
+    )
+
+    print(f"\nIssue workflow: Injected investigation context for #{issue_number}")
+    print(f"  Root cause: {report.root_cause.identified_root_cause[:80]}...")
+    print(f"  Severity: {report.severity}")
+    print()
@@ -0,0 +1,375 @@
+"""
+Followup Commands
+=================
+
+CLI commands for adding follow-up tasks to completed specs.
+"""
+
+import asyncio
+import json
+import sys
+from pathlib import Path
+
+# Ensure parent directory is in path for imports (before other imports)
+_PARENT_DIR = Path(__file__).parent.parent
+if str(_PARENT_DIR) not in sys.path:
+    sys.path.insert(0, str(_PARENT_DIR))
+
+from progress import count_subtasks, is_build_complete
+from ui import (
+    Icons,
+    MenuOption,
+    bold,
+    box,
+    error,
+    highlight,
+    icon,
+    muted,
+    print_status,
+    select_menu,
+    success,
+    warning,
+)
+
+
+def collect_followup_task(spec_dir: Path, max_retries: int = 3) -> str | None:
+    """
+    Collect a follow-up task description from the user.
+
+    Provides multiple input methods (type, paste, file) similar to the
+    HUMAN_INPUT.md pattern used during build interrupts. Includes retry
+    logic for empty input.
+
+    Args:
+        spec_dir: The spec directory where FOLLOWUP_REQUEST.md will be saved
+        max_retries: Maximum number of times to prompt on empty input (default: 3)
+
+    Returns:
+        The collected task description, or None if cancelled
+    """
+    retry_count = 0
+
+    while retry_count < max_retries:
+        # Present options menu
+        options = [
+            MenuOption(
+                key="type",
+                label="Type follow-up task",
+                icon=Icons.EDIT,
+                description="Enter a description of additional work needed",
+            ),
+            MenuOption(
+                key="paste",
+                label="Paste from clipboard",
+                icon=Icons.CLIPBOARD,
+                description="Paste text you've copied (Cmd+V / Ctrl+Shift+V)",
+            ),
+            MenuOption(
+                key="file",
+                label="Read from file",
+                icon=Icons.DOCUMENT,
+                description="Load task description from a text file",
+            ),
+            MenuOption(
+                key="quit",
+                label="Cancel",
+                icon=Icons.DOOR,
+                description="Exit without adding follow-up",
+            ),
+        ]
+
+        # Show retry message if this is a retry
+        subtitle = "Describe the additional work you want to add to this spec."
+        if retry_count > 0:
+            subtitle = warning(
+                f"Empty input received. Please try again. ({max_retries - retry_count} attempts remaining)"
+            )
+
+        choice = select_menu(
+            title="How would you like to provide your follow-up task?",
+            options=options,
+            subtitle=subtitle,
+            allow_quit=False,  # We have explicit quit option
+        )
+
+        if choice == "quit" or choice is None:
+            return None
+
+        followup_task = ""
+
+        if choice == "file":
+            # Read from file
+            print()
+            print(
+                f"{icon(Icons.DOCUMENT)} Enter the path to your task description file:"
+            )
+            try:
+                file_path_str = input(f"  {icon(Icons.POINTER)} ").strip()
+            except (KeyboardInterrupt, EOFError):
+                print()
+                print_status("Cancelled.", "warning")
+                return None
+
+            # Handle empty file path
+            if not file_path_str:
+                print()
+                print_status("No file path provided.", "warning")
+                retry_count += 1
+                continue
+
+            try:
+                # Expand ~ and resolve path
+                file_path = Path(file_path_str).expanduser().resolve()
+                if file_path.exists():
+                    followup_task = file_path.read_text(encoding="utf-8").strip()
+                    if followup_task:
+                        print_status(
+                            f"Loaded {len(followup_task)} characters from file",
+                            "success",
+                        )
+                    else:
+                        print()
+                        print_status(
+                            "File is empty. Please provide a file with task description.",
+                            "error",
+                        )
+                        retry_count += 1
+                        continue
+                else:
+                    print_status(f"File not found: {file_path}", "error")
+                    print(
+                        muted("  Check that the path is correct and the file exists.")
+                    )
+                    retry_count += 1
+                    continue
+            except PermissionError:
+                print_status(f"Permission denied: cannot read {file_path_str}", "error")
+                print(muted("  Check file permissions and try again."))
+                retry_count += 1
+                continue
+            except Exception as e:
+                print_status(f"Error reading file: {e}", "error")
+                retry_count += 1
+                continue
+
+        elif choice in ["type", "paste"]:
+            print()
+            content = [
+                "Enter/paste your follow-up task description below.",
+                "",
+                muted("Describe what additional work you want to add."),
+                muted("The planner will create new subtasks based on this."),
+                "",
+                muted("Press Enter on an empty line when done."),
+            ]
+            print(box(content, width=60, style="light"))
+            print()
+
+            lines = []
+            empty_count = 0
+            while True:
+                try:
+                    line = input()
+                    if line == "":
+                        empty_count += 1
+                        if empty_count >= 1:  # Stop on first empty line
+                            break
+                    else:
+                        empty_count = 0
+                        lines.append(line)
+                except KeyboardInterrupt:
+                    print()
+                    print_status("Cancelled.", "warning")
+                    return None
+                except EOFError:
+                    break
+
+            followup_task = "\n".join(lines).strip()
+
+        # Validate that we have content
+        if not followup_task:
+            print()
+            print_status("No task description provided.", "warning")
+            retry_count += 1
+            continue
+
+        # Save to FOLLOWUP_REQUEST.md
+        request_file = spec_dir / "FOLLOWUP_REQUEST.md"
+        request_file.write_text(followup_task, encoding="utf-8")
+
+        # Show confirmation
+        content = [
+            success(f"{icon(Icons.SUCCESS)} FOLLOW-UP TASK SAVED"),
+            "",
+            f"Saved to: {highlight(str(request_file.name))}",
+            "",
+            muted("The planner will create new subtasks based on this task."),
+        ]
+        print()
+        print(box(content, width=70, style="heavy"))
+
+        return followup_task
+
+    # Max retries exceeded
+    print()
+    print_status("Maximum retry attempts reached. Follow-up cancelled.", "error")
+    return None
+
+
+def handle_followup_command(
+    project_dir: Path,
+    spec_dir: Path,
+    model: str,
+    verbose: bool = False,
+) -> None:
+    """
+    Handle the --followup command.
+
+    Args:
+        project_dir: Project root directory
+        spec_dir: Spec directory path
+        model: Model to use
+        verbose: Enable verbose output
+    """
+    # Lazy imports to avoid loading heavy modules
+    from agent import run_followup_planner
+
+    from .utils import print_banner, validate_environment
+
+    print_banner()
+    print(f"\nFollow-up request for: {spec_dir.name}")
+
+    # Check if implementation_plan.json exists
+    plan_file = spec_dir / "implementation_plan.json"
+    if not plan_file.exists():
+        print()
+        print(error(f"{icon(Icons.ERROR)} No implementation plan found."))
+        print()
+        content = [
+            "This spec has not been built yet.",
+            "",
+            "Follow-up tasks can only be added to specs that have been",
+            "built at least once. Run a regular build first:",
+            "",
+            highlight(f"  python auto-claude/run.py --spec {spec_dir.name}"),
+            "",
+            muted("After the build completes, you can add follow-up tasks."),
+        ]
+        print(box(content, width=70, style="light"))
+        sys.exit(1)
+
+    # Check if build is complete
+    if not is_build_complete(spec_dir):
+        completed, total = count_subtasks(spec_dir)
+        pending = total - completed
+        print()
+        print(
+            error(
+                f"{icon(Icons.ERROR)} Build not complete ({completed}/{total} subtasks)."
+            )
+        )
+        print()
+        content = [
+            f"There are still {pending} pending subtask(s) to complete.",
+            "",
+            "Follow-up tasks can only be added after all current subtasks",
+            "are finished. Complete the current build first:",
+            "",
+            highlight(f"  python auto-claude/run.py --spec {spec_dir.name}"),
+            "",
+            muted("The build will continue from where it left off."),
+        ]
+        print(box(content, width=70, style="light"))
+        sys.exit(1)
+
+    # Check for prior follow-ups (for sequential follow-up context)
+    prior_followup_count = 0
+    try:
+        with open(plan_file, encoding="utf-8") as f:
+            plan_data = json.load(f)
+        phases = plan_data.get("phases", [])
+        # Count phases that look like follow-up phases (name contains "Follow" or high phase number)
+        for phase in phases:
+            phase_name = phase.get("name", "")
+            if "follow" in phase_name.lower() or "followup" in phase_name.lower():
+                prior_followup_count += 1
+    except (json.JSONDecodeError, KeyError):
+        pass  # If plan parsing fails, just continue without prior count
+
+    # Build is complete - proceed to follow-up workflow
+    print()
+    if prior_followup_count > 0:
+        print(
+            success(
+                f"{icon(Icons.SUCCESS)} Build is complete ({prior_followup_count} prior follow-up(s)). Ready for more follow-up tasks."
+            )
+        )
+    else:
+        print(
+            success(
+                f"{icon(Icons.SUCCESS)} Build is complete. Ready for follow-up tasks."
+            )
+        )
+
+    # Collect follow-up task from user
+    followup_task = collect_followup_task(spec_dir)
+
+    if followup_task is None:
+        # User cancelled
+        print()
+        print_status("Follow-up cancelled.", "info")
+        return
+
+    # Successfully collected follow-up task
+    # The collect_followup_task() function already saved to FOLLOWUP_REQUEST.md
+    # Now run the follow-up planner to add new subtasks
+    print()
+
+    if not validate_environment(spec_dir):
+        sys.exit(1)
+
+    try:
+        success_result = asyncio.run(
+            run_followup_planner(
+                project_dir=project_dir,
+                spec_dir=spec_dir,
+                model=model,
+                verbose=verbose,
+            )
+        )
+
+        if success_result:
+            # Show next steps after successful planning
+            content = [
+                bold(f"{icon(Icons.SUCCESS)} FOLLOW-UP PLANNING COMPLETE"),
+                "",
+                "New subtasks have been added to your implementation plan.",
+                "",
+                highlight("To continue building:"),
+                f"  python auto-claude/run.py --spec {spec_dir.name}",
+            ]
+            print(box(content, width=70, style="heavy"))
+        else:
+            # Planning didn't fully succeed
+            content = [
+                bold(f"{icon(Icons.WARNING)} FOLLOW-UP PLANNING INCOMPLETE"),
+                "",
+                "Check the implementation plan manually.",
+                "",
+                muted("You may need to run the follow-up again."),
+            ]
+            print(box(content, width=70, style="light"))
+            sys.exit(1)
+
+    except KeyboardInterrupt:
+        print("\n\nFollow-up planning paused.")
+        print(f"To retry: python auto-claude/run.py --spec {spec_dir.name} --followup")
+        sys.exit(0)
+    except Exception as e:
+        print()
+        print(error(f"{icon(Icons.ERROR)} Follow-up planning error: {e}"))
+        if verbose:
+            import traceback
+
+            traceback.print_exc()
+        sys.exit(1)
@@ -0,0 +1,210 @@
+"""
+Input Handlers
+==============
+
+Reusable user input collection utilities for CLI commands.
+"""
+
+import sys
+from pathlib import Path
+
+# Ensure parent directory is in path for imports (before other imports)
+_PARENT_DIR = Path(__file__).parent.parent
+if str(_PARENT_DIR) not in sys.path:
+    sys.path.insert(0, str(_PARENT_DIR))
+
+from ui import (
+    Icons,
+    MenuOption,
+    box,
+    icon,
+    muted,
+    print_status,
+    select_menu,
+)
+
+
+def collect_user_input_interactive(
+    title: str,
+    subtitle: str,
+    prompt_text: str,
+    allow_file: bool = True,
+    allow_paste: bool = True,
+) -> str | None:
+    """
+    Collect user input through an interactive menu.
+
+    Provides multiple input methods:
+    - Type directly
+    - Paste from clipboard
+    - Read from file (optional)
+
+    Args:
+        title: Menu title
+        subtitle: Menu subtitle
+        prompt_text: Text to display in the input box
+        allow_file: Whether to allow file input (default: True)
+        allow_paste: Whether to allow paste option (default: True)
+
+    Returns:
+        The collected input string, or None if cancelled
+    """
+    # Build options list
+    options = [
+        MenuOption(
+            key="type",
+            label="Type instructions",
+            icon=Icons.EDIT,
+            description="Enter text directly",
+        ),
+    ]
+
+    if allow_paste:
+        options.append(
+            MenuOption(
+                key="paste",
+                label="Paste from clipboard",
+                icon=Icons.CLIPBOARD,
+                description="Paste text you've copied (Cmd+V / Ctrl+Shift+V)",
+            )
+        )
+
+    if allow_file:
+        options.append(
+            MenuOption(
+                key="file",
+                label="Read from file",
+                icon=Icons.DOCUMENT,
+                description="Load text from a file",
+            )
+        )
+
+    options.extend(
+        [
+            MenuOption(
+                key="skip",
+                label="Continue without input",
+                icon=Icons.SKIP,
+                description="Skip this step",
+            ),
+            MenuOption(
+                key="quit",
+                label="Quit",
+                icon=Icons.DOOR,
+                description="Exit",
+            ),
+        ]
+    )
+
+    choice = select_menu(
+        title=title,
+        options=options,
+        subtitle=subtitle,
+        allow_quit=False,  # We have explicit quit option
+    )
+
+    if choice == "quit" or choice is None:
+        return None
+
+    if choice == "skip":
+        return ""
+
+    user_input = ""
+
+    if choice == "file":
+        # Read from file
+        user_input = read_from_file()
+        if user_input is None:
+            return None
+
+    elif choice in ["type", "paste"]:
+        user_input = read_multiline_input(prompt_text)
+        if user_input is None:
+            return None
+
+    return user_input
+
+
+def read_from_file() -> str | None:
+    """
+    Read text content from a file path provided by the user.
+
+    Returns:
+        File contents as string, or None if cancelled/error
+    """
+    print()
+    print(f"{icon(Icons.DOCUMENT)} Enter the path to your file:")
+    try:
+        file_path_input = input(f"  {icon(Icons.POINTER)} ").strip()
+    except (KeyboardInterrupt, EOFError):
+        print()
+        print_status("Cancelled.", "warning")
+        return None
+
+    if not file_path_input:
+        print_status("No file path provided.", "warning")
+        return None
+
+    try:
+        # Expand ~ and resolve path
+        file_path = Path(file_path_input).expanduser().resolve()
+        if file_path.exists():
+            content = file_path.read_text(encoding="utf-8").strip()
+            if content:
+                print_status(
+                    f"Loaded {len(content)} characters from file",
+                    "success",
+                )
+                return content
+            else:
+                print_status("File is empty.", "error")
+                return None
+        else:
+            print_status(f"File not found: {file_path}", "error")
+            return None
+    except PermissionError:
+        print_status(f"Permission denied: cannot read {file_path_input}", "error")
+        return None
+    except Exception as e:
+        print_status(f"Error reading file: {e}", "error")
+        return None
+
+
+def read_multiline_input(prompt_text: str) -> str | None:
+    """
+    Read multi-line input from the user.
+
+    Args:
+        prompt_text: Text to display in the prompt box
+
+    Returns:
+        User input as string, or None if cancelled
+    """
+    print()
+    content = [
+        prompt_text,
+        muted("Press Enter on an empty line when done."),
+    ]
+    print(box(content, width=60, style="light"))
+    print()
+
+    lines = []
+    empty_count = 0
+    while True:
+        try:
+            line = input()
+            if line == "":
+                empty_count += 1
+                if empty_count >= 1:  # Stop on first empty line
+                    break
+            else:
+                empty_count = 0
+                lines.append(line)
+        except KeyboardInterrupt:
+            print()
+            print_status("Cancelled.", "warning")
+            return None
+        except EOFError:
+            break
+
+    return "\n".join(lines).strip()
@@ -0,0 +1,499 @@
+"""
+Auto Claude CLI - Main Entry Point
+===================================
+
+Command-line interface for the Auto Claude autonomous coding framework.
+"""
+
+import argparse
+import os
+import sys
+from pathlib import Path
+
+# Ensure parent directory is in path for imports (before other imports)
+_PARENT_DIR = Path(__file__).parent.parent
+if str(_PARENT_DIR) not in sys.path:
+    sys.path.insert(0, str(_PARENT_DIR))
+
+
+from .batch_commands import (
+    handle_batch_cleanup_command,
+    handle_batch_create_command,
+    handle_batch_status_command,
+)
+from .build_commands import handle_build_command
+from .followup_commands import handle_followup_command
+from .qa_commands import (
+    handle_qa_command,
+    handle_qa_status_command,
+    handle_review_status_command,
+)
+from .spec_commands import print_specs_list
+from .utils import (
+    DEFAULT_MODEL,
+    find_spec,
+    get_project_dir,
+    print_banner,
+    setup_environment,
+)
+from .workspace_commands import (
+    handle_cleanup_worktrees_command,
+    handle_create_pr_command,
+    handle_discard_command,
+    handle_list_worktrees_command,
+    handle_merge_command,
+    handle_review_command,
+)
+
+
+def parse_args() -> argparse.Namespace:
+    """Parse command line arguments."""
+    parser = argparse.ArgumentParser(
+        description="Auto Claude Framework - Autonomous multi-session coding agent",
+        formatter_class=argparse.RawDescriptionHelpFormatter,
+        epilog="""
+Examples:
+  # List all specs
+  python auto-claude/run.py --list
+
+  # Run a specific spec (by number or full name)
+  python auto-claude/run.py --spec 001
+  python auto-claude/run.py --spec 001-initial-app
+
+  # Workspace management (after build completes)
+  python auto-claude/run.py --spec 001 --merge     # Add build to your project
+  python auto-claude/run.py --spec 001 --review    # See what was built
+  python auto-claude/run.py --spec 001 --discard   # Delete build (with confirmation)
+
+  # Advanced options
+  python auto-claude/run.py --spec 001 --direct       # Skip workspace isolation
+  python auto-claude/run.py --spec 001 --isolated     # Force workspace isolation
+
+  # Status checks
+  python auto-claude/run.py --spec 001 --review-status  # Check human review status
+  python auto-claude/run.py --spec 001 --qa-status      # Check QA validation status
+
+Prerequisites:
+  1. Authenticate: Run 'claude' and type '/login'
+  2. Create a spec first: claude /spec
+
+Environment Variables:
+  CLAUDE_CODE_OAUTH_TOKEN  Your Claude Code OAuth token (auto-detected from Keychain)
+                           Or authenticate via: claude → /login
+  AUTO_BUILD_MODEL         Override default model (optional)
+        """,
+    )
+
+    parser.add_argument(
+        "--list",
+        action="store_true",
+        help="List all available specs and their status",
+    )
+
+    parser.add_argument(
+        "--spec",
+        type=str,
+        default=None,
+        help="Spec to run (e.g., '001' or '001-feature-name')",
+    )
+
+    parser.add_argument(
+        "--project-dir",
+        type=Path,
+        default=None,
+        help="Project directory (default: current working directory)",
+    )
+
+    parser.add_argument(
+        "--max-iterations",
+        type=int,
+        default=None,
+        help="Maximum number of agent sessions (default: unlimited)",
+    )
+
+    parser.add_argument(
+        "--model",
+        type=str,
+        default=None,
+        help=f"Claude model to use (default: {DEFAULT_MODEL})",
+    )
+
+    parser.add_argument(
+        "--verbose",
+        action="store_true",
+        help="Enable verbose output",
+    )
+
+    # Workspace options
+    workspace_group = parser.add_mutually_exclusive_group()
+    workspace_group.add_argument(
+        "--isolated",
+        action="store_true",
+        help="Force building in isolated workspace (safer)",
+    )
+    workspace_group.add_argument(
+        "--direct",
+        action="store_true",
+        help="Build directly in your project (no isolation)",
+    )
+
+    # Build management commands
+    build_group = parser.add_mutually_exclusive_group()
+    build_group.add_argument(
+        "--merge",
+        action="store_true",
+        help="Merge an existing build into your project",
+    )
+    build_group.add_argument(
+        "--review",
+        action="store_true",
+        help="Review what an existing build contains",
+    )
+    build_group.add_argument(
+        "--discard",
+        action="store_true",
+        help="Discard an existing build (requires confirmation)",
+    )
+    build_group.add_argument(
+        "--create-pr",
+        action="store_true",
+        help="Push branch and create a GitHub Pull Request",
+    )
+
+    # PR options
+    parser.add_argument(
+        "--pr-target",
+        type=str,
+        metavar="BRANCH",
+        help="With --create-pr: target branch for PR (default: auto-detect)",
+    )
+    parser.add_argument(
+        "--pr-title",
+        type=str,
+        metavar="TITLE",
+        help="With --create-pr: custom PR title (default: generated from spec name)",
+    )
+    parser.add_argument(
+        "--pr-draft",
+        action="store_true",
+        help="With --create-pr: create as draft PR",
+    )
+
+    # Merge options
+    parser.add_argument(
+        "--no-commit",
+        action="store_true",
+        help="With --merge: stage changes but don't commit (review in IDE first)",
+    )
+    parser.add_argument(
+        "--merge-preview",
+        action="store_true",
+        help="Preview merge conflicts without actually merging (returns JSON)",
+    )
+
+    # QA options
+    parser.add_argument(
+        "--qa",
+        action="store_true",
+        help="Run QA validation loop on a completed build",
+    )
+    parser.add_argument(
+        "--qa-status",
+        action="store_true",
+        help="Show QA validation status for a spec",
+    )
+    parser.add_argument(
+        "--skip-qa",
+        action="store_true",
+        help="Skip automatic QA validation after build completes",
+    )
+
+    # Follow-up options
+    parser.add_argument(
+        "--followup",
+        action="store_true",
+        help="Add follow-up tasks to a completed spec (extends existing implementation plan)",
+    )
+
+    # Review options
+    parser.add_argument(
+        "--review-status",
+        action="store_true",
+        help="Show human review/approval status for a spec",
+    )
+
+    # Non-interactive mode (for UI/automation)
+    parser.add_argument(
+        "--auto-continue",
+        action="store_true",
+        help="Non-interactive mode: auto-continue existing builds, skip prompts (for UI integration)",
+    )
+
+    # Worktree management
+    parser.add_argument(
+        "--list-worktrees",
+        action="store_true",
+        help="List all spec worktrees and their status",
+    )
+    parser.add_argument(
+        "--cleanup-worktrees",
+        action="store_true",
+        help="Remove all spec worktrees and their branches (with confirmation)",
+    )
+
+    # Force bypass
+    parser.add_argument(
+        "--force",
+        action="store_true",
+        help="Skip approval check and start build anyway (for debugging)",
+    )
+
+    # Base branch for worktree creation
+    parser.add_argument(
+        "--base-branch",
+        type=str,
+        default=None,
+        help="Base branch for creating worktrees (default: auto-detect or current branch)",
+    )
+
+    # Batch task management
+    parser.add_argument(
+        "--batch-create",
+        type=str,
+        default=None,
+        metavar="FILE",
+        help="Create multiple tasks from a batch JSON file",
+    )
+    parser.add_argument(
+        "--batch-status",
+        action="store_true",
+        help="Show status of all specs in the project",
+    )
+    parser.add_argument(
+        "--batch-cleanup",
+        action="store_true",
+        help="Clean up completed specs (dry-run by default)",
+    )
+    parser.add_argument(
+        "--no-dry-run",
+        action="store_true",
+        help="Actually delete files in cleanup (not just preview)",
+    )
+
+    # Issue workflow
+    parser.add_argument(
+        "--issue-workflow",
+        action="store_true",
+        help="Run build from a GitHub issue investigation (requires --issue-number)",
+    )
+    parser.add_argument(
+        "--issue-number",
+        type=int,
+        default=None,
+        help="GitHub issue number for --issue-workflow",
+    )
+
+    return parser.parse_args()
+
+
+def main() -> None:
+    """Main CLI entry point."""
+    # Set up environment first
+    setup_environment()
+
+    # Initialize Sentry early to capture any startup errors
+    from core.sentry import capture_exception, init_sentry
+
+    init_sentry(component="cli")
+
+    try:
+        _run_cli()
+    except KeyboardInterrupt:
+        # Clean exit on Ctrl+C
+        sys.exit(130)
+    except Exception as e:
+        # Capture unexpected errors to Sentry
+        capture_exception(e)
+        print(f"\nUnexpected error: {e}")
+        sys.exit(1)
+
+
+def _run_cli() -> None:
+    """Run the CLI logic (extracted for error handling)."""
+    # Import here to avoid import errors during startup
+    from core.sentry import set_context
+
+    # Parse arguments
+    args = parse_args()
+
+    # Import debug functions after environment setup
+    from debug import debug, debug_error, debug_section, debug_success
+
+    debug_section("run.py", "Starting Auto-Build Framework")
+    debug("run.py", "Arguments parsed", args=vars(args))
+
+    # Determine project directory
+    project_dir = get_project_dir(args.project_dir)
+    debug("run.py", f"Using project directory: {project_dir}")
+
+    # Get model from CLI arg or env var (None if not explicitly set)
+    # This allows get_phase_model() to fall back to task_metadata.json
+    model = args.model or os.environ.get("AUTO_BUILD_MODEL")
+
+    # Handle --list command
+    if args.list:
+        print_banner()
+        print_specs_list(project_dir)
+        return
+
+    # Handle --list-worktrees command
+    if args.list_worktrees:
+        handle_list_worktrees_command(project_dir)
+        return
+
+    # Handle --cleanup-worktrees command
+    if args.cleanup_worktrees:
+        handle_cleanup_worktrees_command(project_dir)
+        return
+
+    # Handle batch commands
+    if args.batch_create:
+        handle_batch_create_command(args.batch_create, str(project_dir))
+        return
+
+    if args.batch_status:
+        handle_batch_status_command(str(project_dir))
+        return
+
+    if args.batch_cleanup:
+        handle_batch_cleanup_command(str(project_dir), dry_run=not args.no_dry_run)
+        return
+
+    # Require --spec if not listing
+    if not args.spec:
+        print_banner()
+        print("\nError: --spec is required")
+        print("\nUsage:")
+        print("  python auto-claude/run.py --list           # See all specs")
+        print("  python auto-claude/run.py --spec 001       # Run a spec")
+        print("\nCreate a new spec with:")
+        print("  claude /spec")
+        sys.exit(1)
+
+    # Find the spec
+    debug("run.py", "Finding spec", spec_identifier=args.spec)
+    spec_dir = find_spec(project_dir, args.spec)
+    if not spec_dir:
+        debug_error("run.py", "Spec not found", spec=args.spec)
+        print_banner()
+        print(f"\nError: Spec '{args.spec}' not found")
+        print("\nAvailable specs:")
+        print_specs_list(project_dir)
+        sys.exit(1)
+
+    debug_success("run.py", "Spec found", spec_dir=str(spec_dir))
+
+    # Set Sentry context for error tracking
+    set_context(
+        "spec",
+        {
+            "name": spec_dir.name,
+            "project": str(project_dir),
+        },
+    )
+
+    # Handle build management commands
+    if args.merge_preview:
+        from cli.workspace_commands import handle_merge_preview_command
+
+        result = handle_merge_preview_command(
+            project_dir, spec_dir.name, base_branch=args.base_branch
+        )
+        # Output as JSON for the UI to parse
+        import json
+
+        print(json.dumps(result))
+        return
+
+    if args.merge:
+        success = handle_merge_command(
+            project_dir,
+            spec_dir.name,
+            no_commit=args.no_commit,
+            base_branch=args.base_branch,
+        )
+        if not success:
+            sys.exit(1)
+        return
+
+    if args.review:
+        handle_review_command(project_dir, spec_dir.name)
+        return
+
+    if args.discard:
+        handle_discard_command(project_dir, spec_dir.name)
+        return
+
+    if args.create_pr:
+        # Pass args.pr_target directly - WorktreeManager._detect_base_branch
+        # handles base branch detection internally when target_branch is None
+        result = handle_create_pr_command(
+            project_dir=project_dir,
+            spec_name=spec_dir.name,
+            target_branch=args.pr_target,
+            title=args.pr_title,
+            draft=args.pr_draft,
+        )
+        # JSON output is already printed by handle_create_pr_command
+        if not result.get("success"):
+            sys.exit(1)
+        return
+
+    # Handle QA commands
+    if args.qa_status:
+        handle_qa_status_command(spec_dir)
+        return
+
+    if args.review_status:
+        handle_review_status_command(spec_dir)
+        return
+
+    if args.qa:
+        handle_qa_command(
+            project_dir=project_dir,
+            spec_dir=spec_dir,
+            model=model,
+            verbose=args.verbose,
+        )
+        return
+
+    # Handle --followup command
+    if args.followup:
+        handle_followup_command(
+            project_dir=project_dir,
+            spec_dir=spec_dir,
+            model=model,
+            verbose=args.verbose,
+        )
+        return
+
+    # Normal build flow
+    handle_build_command(
+        project_dir=project_dir,
+        spec_dir=spec_dir,
+        model=model,
+        max_iterations=args.max_iterations,
+        verbose=args.verbose,
+        force_isolated=args.isolated,
+        force_direct=args.direct,
+        auto_continue=args.auto_continue,
+        skip_qa=args.skip_qa,
+        force_bypass_approval=args.force,
+        base_branch=args.base_branch,
+        issue_workflow=args.issue_workflow,
+        issue_number=args.issue_number,
+    )
+
+
+if __name__ == "__main__":
+    main()
@@ -0,0 +1,127 @@
+"""
+QA Commands
+===========
+
+CLI commands for QA validation (run QA, check status)
+"""
+
+import asyncio
+import sys
+from pathlib import Path
+
+# Ensure parent directory is in path for imports (before other imports)
+_PARENT_DIR = Path(__file__).parent.parent
+if str(_PARENT_DIR) not in sys.path:
+    sys.path.insert(0, str(_PARENT_DIR))
+
+from progress import count_subtasks
+from qa_loop import (
+    is_qa_approved,
+    print_qa_status,
+    run_qa_validation_loop,
+    should_run_qa,
+)
+from review import ReviewState, display_review_status
+from ui import (
+    Icons,
+    icon,
+    info,
+    success,
+    warning,
+)
+
+from .utils import print_banner, validate_environment
+
+
+def handle_qa_status_command(spec_dir: Path) -> None:
+    """
+    Handle the --qa-status command.
+
+    Args:
+        spec_dir: Spec directory path
+    """
+    print_banner()
+    print(f"\nSpec: {spec_dir.name}\n")
+    print_qa_status(spec_dir)
+
+
+def handle_review_status_command(spec_dir: Path) -> None:
+    """
+    Handle the --review-status command.
+
+    Args:
+        spec_dir: Spec directory path
+    """
+    print_banner()
+    print(f"\nSpec: {spec_dir.name}\n")
+    display_review_status(spec_dir)
+    # Also show if approval is valid for build
+    review_state = ReviewState.load(spec_dir)
+    print()
+    if review_state.is_approval_valid(spec_dir):
+        print(success(f"{icon(Icons.SUCCESS)} Ready to build - approval is valid."))
+    elif review_state.approved:
+        print(
+            warning(
+                f"{icon(Icons.WARNING)} Spec changed since approval - re-review required."
+            )
+        )
+    else:
+        print(info(f"{icon(Icons.INFO)} Review required before building."))
+    print()
+
+
+def handle_qa_command(
+    project_dir: Path,
+    spec_dir: Path,
+    model: str,
+    verbose: bool = False,
+) -> None:
+    """
+    Handle the --qa command (run QA validation loop).
+
+    Args:
+        project_dir: Project root directory
+        spec_dir: Spec directory path
+        model: Model to use for QA
+        verbose: Enable verbose output
+    """
+    print_banner()
+    print(f"\nRunning QA validation for: {spec_dir.name}")
+    if not validate_environment(spec_dir):
+        sys.exit(1)
+
+    # Check if there's pending human feedback that needs to be processed
+    # Human feedback takes priority over "already approved" status
+    fix_request_file = spec_dir / "QA_FIX_REQUEST.md"
+    has_human_feedback = fix_request_file.exists()
+
+    if not should_run_qa(spec_dir) and not has_human_feedback:
+        if is_qa_approved(spec_dir):
+            print("\n✅ Build already approved by QA.")
+        else:
+            completed, total = count_subtasks(spec_dir)
+            print(f"\n❌ Build not complete ({completed}/{total} subtasks).")
+            print("Complete all subtasks before running QA validation.")
+        return
+
+    if has_human_feedback:
+        print("\n📝 Human feedback detected - processing fix request...")
+
+    try:
+        approved = asyncio.run(
+            run_qa_validation_loop(
+                project_dir=project_dir,
+                spec_dir=spec_dir,
+                model=model,
+                verbose=verbose,
+            )
+        )
+        if approved:
+            print("\n✅ QA validation passed. Ready for merge.")
+        else:
+            print("\n❌ QA validation incomplete. See reports for details.")
+            sys.exit(1)
+    except KeyboardInterrupt:
+        print("\n\nQA validation paused.")
+        print(f"Resume with: python auto-claude/run.py --spec {spec_dir.name} --qa")
@@ -0,0 +1,217 @@
+#!/usr/bin/env python3
+"""
+JSON Recovery Utility
+=====================
+
+Detects and repairs corrupted JSON files in specs directories.
+
+Usage:
+    python -m cli.recovery --project-dir /path/to/project --detect
+    python -m cli.recovery --project-dir /path/to/project --spec-id 004-feature --delete
+    python -m cli.recovery --project-dir /path/to/project --all --delete
+"""
+
+import argparse
+import json
+import sys
+import uuid
+from pathlib import Path
+
+from cli.utils import find_specs_dir
+
+
+def check_json_file(filepath: Path) -> tuple[bool, str | None]:
+    """
+    Check if a JSON file is valid.
+
+    Returns:
+        (is_valid, error_message)
+    """
+    try:
+        with open(filepath, encoding="utf-8") as f:
+            json.load(f)
+        return True, None
+    except json.JSONDecodeError as e:
+        return False, str(e)
+    except Exception as e:
+        return False, str(e)
+
+
+def detect_corrupted_files(specs_dir: Path) -> list[tuple[Path, str]]:
+    """
+    Scan specs directory recursively for corrupted JSON files.
+
+    Returns:
+        List of (filepath, error_message) tuples
+    """
+    corrupted = []
+
+    if not specs_dir.exists():
+        return corrupted
+
+    # Recursively scan for JSON files (includes nested files like memory/*.json)
+    for json_file in specs_dir.rglob("*.json"):
+        is_valid, error = check_json_file(json_file)
+        if not is_valid:
+            # Type narrowing: error is str when is_valid is False
+            assert error is not None
+            corrupted.append((json_file, error))
+
+    return corrupted
+
+
+def backup_corrupted_file(filepath: Path) -> bool:
+    """
+    Backup a corrupted file by renaming it with a .corrupted suffix.
+
+    Args:
+        filepath: Path to the corrupted file
+
+    Returns:
+        True if backed up successfully, False otherwise
+    """
+    try:
+        # Create backup before deleting
+        base_backup_path = filepath.with_suffix(f"{filepath.suffix}.corrupted")
+        backup_path = base_backup_path
+
+        # Handle existing backup files by generating unique name with UUID
+        if backup_path.exists():
+            # Use UUID for unique naming to avoid races
+            unique_suffix = uuid.uuid4().hex[:8]
+            backup_path = filepath.with_suffix(
+                f"{filepath.suffix}.corrupted.{unique_suffix}"
+            )
+
+        filepath.rename(backup_path)
+        print(f"  [BACKUP] Moved corrupted file to: {backup_path}")
+        return True
+    except Exception as e:
+        print(f"  [ERROR] Failed to backup file: {e}")
+        return False
+
+
+def main() -> None:
+    parser = argparse.ArgumentParser(
+        description="Detect and repair corrupted JSON files in specs directories"
+    )
+    parser.add_argument(
+        "--project-dir",
+        type=Path,
+        default=Path.cwd(),
+        help="Project directory (default: current directory)",
+    )
+    parser.add_argument(
+        "--specs-dir",
+        type=Path,
+        help="Specs directory path (overrides auto-detection)",
+    )
+    parser.add_argument(
+        "--detect",
+        action="store_true",
+        help="Detect corrupted JSON files",
+    )
+    parser.add_argument(
+        "--spec-id",
+        type=str,
+        help="Specific spec ID to fix (e.g., 004-feature)",
+    )
+    parser.add_argument(
+        "--delete",
+        action="store_true",
+        help="Delete corrupted files (creates .corrupted backup)",
+    )
+    parser.add_argument(
+        "--all",
+        action="store_true",
+        help="Fix all corrupted files (requires --delete)",
+    )
+
+    args = parser.parse_args()
+
+    # Validate --all requires --delete
+    if args.all and not args.delete:
+        parser.error("--all requires --delete")
+
+    # Find specs directory
+    if args.specs_dir:
+        specs_dir = args.specs_dir
+    else:
+        specs_dir = find_specs_dir(args.project_dir)
+
+    print(f"[INFO] Scanning specs directory: {specs_dir}")
+
+    # Default to detect mode if no flags provided
+    if not args.detect and not args.delete:
+        args.detect = True
+
+    # Detect corrupted files (dry-run when detect-only, otherwise for deletion)
+    corrupted = detect_corrupted_files(specs_dir)
+
+    # Detect-only mode: show results and exit
+    if args.detect and not args.delete:
+        if not corrupted:
+            print("[OK] No corrupted JSON files found")
+            sys.exit(0)
+
+        print(f"\n[FOUND] {len(corrupted)} corrupted file(s):\n")
+        for filepath, error in corrupted:
+            print(f"  - {filepath.relative_to(specs_dir.parent)}")
+            print(f"    Error: {error}")
+        print()
+        # Exit with error code when corrupted files are found
+        sys.exit(1)
+
+    # Delete corrupted files
+    if args.delete:
+        if args.spec_id:
+            # Delete specific spec
+            spec_dir = (specs_dir / args.spec_id).resolve()
+            specs_dir_resolved = specs_dir.resolve()
+            # Validate path doesn't escape specs directory
+            if not spec_dir.is_relative_to(specs_dir_resolved):
+                print("[ERROR] Invalid spec ID: path traversal detected")
+                sys.exit(1)
+
+            if not spec_dir.exists():
+                print(f"[ERROR] Spec directory not found: {spec_dir}")
+                sys.exit(1)
+
+            print(f"[INFO] Processing spec: {args.spec_id}")
+            has_failures = False
+            for json_file in spec_dir.rglob("*.json"):
+                is_valid, error = check_json_file(json_file)
+                if not is_valid:
+                    print(f"  [CORRUPTED] {json_file.name}")
+                    if not backup_corrupted_file(json_file):
+                        has_failures = True
+
+            if has_failures:
+                sys.exit(1)
+
+        elif args.all:
+            # Delete all corrupted files
+            # Use the already-detected corrupted list, or re-scan if needed
+            if not corrupted:
+                corrupted = detect_corrupted_files(specs_dir)
+            if not corrupted:
+                print("[OK] No corrupted files to delete")
+                sys.exit(0)
+
+            print(f"\n[INFO] Backing up {len(corrupted)} corrupted file(s):\n")
+            has_failures = False
+            for filepath, _ in corrupted:
+                # backup_corrupted_file prints its own [BACKUP] message
+                if not backup_corrupted_file(filepath):
+                    has_failures = True
+
+            if has_failures:
+                sys.exit(1)
+
+        else:
+            print("[ERROR] Must specify --spec-id or --all with --delete")
+            sys.exit(1)
+
+
+if __name__ == "__main__":
+    main()
@@ -0,0 +1,191 @@
+"""
+Spec Commands
+=============
+
+CLI commands for managing specs (listing, finding, etc.)
+"""
+
+import sys
+from pathlib import Path
+
+# Ensure parent directory is in path for imports (before other imports)
+_PARENT_DIR = Path(__file__).parent.parent
+if str(_PARENT_DIR) not in sys.path:
+    sys.path.insert(0, str(_PARENT_DIR))
+
+from progress import count_subtasks
+from workspace import get_existing_build_worktree
+
+from .utils import get_specs_dir
+
+
+def list_specs(project_dir: Path) -> list[dict]:
+    """
+    List all specs in the project.
+
+    Args:
+        project_dir: Project root directory
+
+    Returns:
+        List of spec info dicts with keys: number, name, path, status, progress
+    """
+    specs_dir = get_specs_dir(project_dir)
+    specs = []
+
+    if not specs_dir.exists():
+        return specs
+
+    for spec_folder in sorted(specs_dir.iterdir()):
+        if not spec_folder.is_dir():
+            continue
+
+        # Parse folder name (e.g., "001-initial-app")
+        folder_name = spec_folder.name
+        parts = folder_name.split("-", 1)
+        if len(parts) != 2 or not parts[0].isdigit():
+            continue
+
+        number = parts[0]
+        name = parts[1]
+
+        # Check for spec.md
+        spec_file = spec_folder / "spec.md"
+        if not spec_file.exists():
+            continue
+
+        # Check for existing build in worktree
+        has_build = get_existing_build_worktree(project_dir, folder_name) is not None
+
+        # Check progress via implementation_plan.json
+        plan_file = spec_folder / "implementation_plan.json"
+        if plan_file.exists():
+            completed, total = count_subtasks(spec_folder)
+            if total > 0:
+                if completed == total:
+                    status = "complete"
+                else:
+                    status = "in_progress"
+                progress = f"{completed}/{total}"
+            else:
+                status = "initialized"
+                progress = "0/0"
+        else:
+            status = "pending"
+            progress = "-"
+
+        # Add build indicator
+        if has_build:
+            status = f"{status} (has build)"
+
+        specs.append(
+            {
+                "number": number,
+                "name": name,
+                "folder": folder_name,
+                "path": spec_folder,
+                "status": status,
+                "progress": progress,
+                "has_build": has_build,
+            }
+        )
+
+    return specs
+
+
+def print_specs_list(project_dir: Path, auto_create: bool = True) -> None:
+    """Print a formatted list of all specs.
+
+    Args:
+        project_dir: Project root directory
+        auto_create: If True and no specs exist, automatically launch spec creation
+    """
+    import subprocess
+
+    specs = list_specs(project_dir)
+
+    if not specs:
+        print("\nNo specs found.")
+
+        if auto_create:
+            # Get the backend directory and find spec_runner.py
+            backend_dir = Path(__file__).parent.parent
+            spec_runner = backend_dir / "runners" / "spec_runner.py"
+
+            # Find Python executable - use current interpreter
+            python_path = sys.executable
+
+            if spec_runner.exists() and python_path:
+                # Quick prompt for task description
+                print("\n" + "=" * 60)
+                print("  QUICK START")
+                print("=" * 60)
+                print("\nWhat do you want to build?")
+                print(
+                    "(Enter a brief description, or press Enter for interactive mode)\n"
+                )
+
+                try:
+                    task = input("> ").strip()
+                except (EOFError, KeyboardInterrupt):
+                    print("\nCancelled.")
+                    return
+
+                if task:
+                    # Direct mode: create spec and start building
+                    print(f"\nStarting build for: {task}\n")
+                    subprocess.run(
+                        [
+                            python_path,
+                            str(spec_runner),
+                            "--task",
+                            task,
+                            "--complexity",
+                            "simple",
+                            "--auto-approve",
+                        ],
+                        cwd=project_dir,
+                    )
+                else:
+                    # Interactive mode
+                    print("\nLaunching interactive mode...\n")
+                    subprocess.run(
+                        [python_path, str(spec_runner), "--interactive"],
+                        cwd=project_dir,
+                    )
+                return
+            else:
+                print("\nCreate your first spec:")
+                print("  python runners/spec_runner.py --interactive")
+        else:
+            print("\nCreate your first spec:")
+            print("  python runners/spec_runner.py --interactive")
+        return
+
+    print("\n" + "=" * 70)
+    print("  AVAILABLE SPECS")
+    print("=" * 70)
+    print()
+
+    # Status symbols
+    status_symbols = {
+        "complete": "[OK]",
+        "in_progress": "[..]",
+        "initialized": "[--]",
+        "pending": "[  ]",
+    }
+
+    for spec in specs:
+        # Get base status for symbol
+        base_status = spec["status"].split(" ")[0]
+        symbol = status_symbols.get(base_status, "[??]")
+
+        print(f"  {symbol} {spec['folder']}")
+        status_line = f"       Status: {spec['status']} | Subtasks: {spec['progress']}"
+        print(status_line)
+        print()
+
+    print("-" * 70)
+    print("\nTo run a spec:")
+    print("  python auto-claude/run.py --spec 001")
+    print("  python auto-claude/run.py --spec 001-feature-name")
+    print()
@@ -0,0 +1,278 @@
+"""
+CLI Utilities
+==============
+
+Shared utility functions for the Auto Claude CLI.
+"""
+
+import os
+import sys
+from pathlib import Path
+
+# Ensure parent directory is in path for imports (before other imports)
+_PARENT_DIR = Path(__file__).parent.parent
+if str(_PARENT_DIR) not in sys.path:
+    sys.path.insert(0, str(_PARENT_DIR))
+
+from core.auth import get_auth_token, get_auth_token_source
+from core.dependency_validator import validate_platform_dependencies
+
+
+def import_dotenv():
+    """
+    Import and return load_dotenv with helpful error message if not installed.
+
+    This centralized function ensures consistent error messaging across all
+    runner scripts when python-dotenv is not available.
+
+    Returns:
+        The load_dotenv function
+
+    Raises:
+        SystemExit: If dotenv cannot be imported, with helpful installation instructions.
+    """
+    try:
+        from dotenv import load_dotenv as _load_dotenv
+
+        return _load_dotenv
+    except ImportError:
+        sys.exit(
+            "Error: Required Python package 'python-dotenv' is not installed.\n"
+            "\n"
+            "This usually means you're not using the virtual environment.\n"
+            "\n"
+            "To fix this:\n"
+            "1. From the 'apps/backend/' directory, activate the venv:\n"
+            "   source .venv/bin/activate  # Linux/macOS\n"
+            "   .venv\\Scripts\\activate   # Windows\n"
+            "\n"
+            "2. Or install dependencies directly:\n"
+            "   pip install python-dotenv\n"
+            "   pip install -r requirements.txt\n"
+            "\n"
+            f"Current Python: {sys.executable}\n"
+        )
+
+
+# Load .env with helpful error if dependencies not installed
+load_dotenv = import_dotenv()
+# NOTE: graphiti_config is imported lazily in validate_environment() to avoid
+# triggering graphiti_core -> real_ladybug -> pywintypes import chain before
+# platform dependency validation can run. See ACS-253.
+from linear_integration import LinearManager
+from linear_updater import is_linear_enabled
+from spec.pipeline import get_specs_dir
+from ui import (
+    Icons,
+    bold,
+    box,
+    icon,
+    muted,
+)
+
+# Configuration - uses shorthand that resolves via API Profile if configured
+DEFAULT_MODEL = "sonnet"  # Changed from "opus" (fix #433)
+
+
+def setup_environment() -> Path:
+    """
+    Set up the environment and return the script directory.
+
+    Returns:
+        Path to the auto-claude directory
+    """
+    # Add auto-claude directory to path for imports
+    script_dir = Path(__file__).parent.parent.resolve()
+    sys.path.insert(0, str(script_dir))
+
+    # Load .env file - check both auto-claude/ and dev/auto-claude/ locations
+    env_file = script_dir / ".env"
+    dev_env_file = script_dir.parent / "dev" / "auto-claude" / ".env"
+    if env_file.exists():
+        load_dotenv(env_file)
+    elif dev_env_file.exists():
+        load_dotenv(dev_env_file)
+
+    return script_dir
+
+
+def find_spec(project_dir: Path, spec_identifier: str) -> Path | None:
+    """
+    Find a spec by number or full name.
+
+    Args:
+        project_dir: Project root directory
+        spec_identifier: Either "001" or "001-feature-name"
+
+    Returns:
+        Path to spec folder, or None if not found
+    """
+    specs_dir = get_specs_dir(project_dir)
+
+    if specs_dir.exists():
+        # Try exact match first
+        exact_path = specs_dir / spec_identifier
+        if exact_path.exists() and (exact_path / "spec.md").exists():
+            return exact_path
+
+        # Try matching by number prefix
+        for spec_folder in specs_dir.iterdir():
+            if spec_folder.is_dir() and spec_folder.name.startswith(
+                spec_identifier + "-"
+            ):
+                if (spec_folder / "spec.md").exists():
+                    return spec_folder
+
+    # Check worktree specs (for merge-preview, merge, review, discard operations)
+    worktree_base = project_dir / ".auto-claude" / "worktrees" / "tasks"
+    if worktree_base.exists():
+        # Try exact match in worktree
+        worktree_spec = (
+            worktree_base / spec_identifier / ".auto-claude" / "specs" / spec_identifier
+        )
+        if worktree_spec.exists() and (worktree_spec / "spec.md").exists():
+            return worktree_spec
+
+        # Try matching by prefix in worktrees
+        for worktree_dir in worktree_base.iterdir():
+            if worktree_dir.is_dir() and worktree_dir.name.startswith(
+                spec_identifier + "-"
+            ):
+                spec_in_worktree = (
+                    worktree_dir / ".auto-claude" / "specs" / worktree_dir.name
+                )
+                if (
+                    spec_in_worktree.exists()
+                    and (spec_in_worktree / "spec.md").exists()
+                ):
+                    return spec_in_worktree
+
+    return None
+
+
+def validate_environment(spec_dir: Path) -> bool:
+    """
+    Validate that the environment is set up correctly.
+
+    Returns:
+        True if valid, False otherwise (with error messages printed)
+    """
+    # Validate platform-specific dependencies first (exits if missing)
+    validate_platform_dependencies()
+
+    valid = True
+
+    # Check for OAuth token (API keys are not supported)
+    if not get_auth_token():
+        print("Error: No OAuth token found")
+        print("\nAuto Claude requires Claude Code OAuth authentication.")
+        print("Direct API keys (ANTHROPIC_API_KEY) are not supported.")
+        print("\nTo authenticate, run:")
+        print("  claude setup-token")
+        valid = False
+    else:
+        # Show which auth source is being used
+        source = get_auth_token_source()
+        if source:
+            print(f"Auth: {source}")
+
+        # Show custom base URL if set
+        base_url = os.environ.get("ANTHROPIC_BASE_URL")
+        if base_url:
+            print(f"API Endpoint: {base_url}")
+
+    # Check for spec.md in spec directory
+    spec_file = spec_dir / "spec.md"
+    if not spec_file.exists():
+        print(f"\nError: spec.md not found in {spec_dir}")
+        valid = False
+
+    # Check Linear integration (optional but show status)
+    if is_linear_enabled():
+        print("Linear integration: ENABLED")
+        # Show Linear project status if initialized
+        project_dir = (
+            spec_dir.parent.parent
+        )  # auto-claude/specs/001-name -> project root
+        linear_manager = LinearManager(spec_dir, project_dir)
+        if linear_manager.is_initialized:
+            summary = linear_manager.get_progress_summary()
+            print(f"  Project: {summary.get('project_name', 'Unknown')}")
+            print(
+                f"  Issues: {summary.get('mapped_subtasks', 0)}/{summary.get('total_subtasks', 0)} mapped"
+            )
+        else:
+            print("  Status: Will be initialized during planner session")
+    else:
+        print("Linear integration: DISABLED (set LINEAR_API_KEY to enable)")
+
+    # Check Graphiti integration (optional but show status)
+    # Lazy import to avoid triggering pywintypes import before validation (ACS-253)
+    from graphiti_config import get_graphiti_status
+
+    graphiti_status = get_graphiti_status()
+    if graphiti_status["available"]:
+        print("Graphiti memory: ENABLED")
+        print(f"  Database: {graphiti_status['database']}")
+        if graphiti_status.get("db_path"):
+            print(f"  Path: {graphiti_status['db_path']}")
+    elif graphiti_status["enabled"]:
+        print(
+            f"Graphiti memory: CONFIGURED but unavailable ({graphiti_status['reason']})"
+        )
+    else:
+        print("Graphiti memory: DISABLED (set GRAPHITI_ENABLED=true to enable)")
+
+    print()
+    return valid
+
+
+def print_banner() -> None:
+    """Print the Auto-Build banner."""
+    content = [
+        bold(f"{icon(Icons.LIGHTNING)} AUTO-BUILD FRAMEWORK"),
+        "",
+        "Autonomous Multi-Session Coding Agent",
+        muted("Subtask-Based Implementation with Phase Dependencies"),
+    ]
+    print()
+    print(box(content, width=70, style="heavy"))
+
+
+def get_project_dir(provided_dir: Path | None) -> Path:
+    """
+    Determine the project directory.
+
+    Args:
+        provided_dir: User-provided project directory (or None)
+
+    Returns:
+        Resolved project directory path
+    """
+    if provided_dir:
+        return provided_dir.resolve()
+
+    project_dir = Path.cwd()
+
+    # Auto-detect if running from within apps/backend directory (the source code)
+    if project_dir.name == "backend" and (project_dir / "run.py").exists():
+        # Running from within apps/backend/ source directory, go up 2 levels
+        project_dir = project_dir.parent.parent
+
+    return project_dir
+
+
+def find_specs_dir(project_dir: Path) -> Path:
+    """
+    Find the specs directory for a project.
+
+    Returns the '.auto-claude/specs' directory path.
+    The directory is guaranteed to exist (get_specs_dir calls init_auto_claude_dir).
+
+    Args:
+        project_dir: Project root directory
+
+    Returns:
+        Path to specs directory (always returns a valid Path)
+    """
+    return get_specs_dir(project_dir)
@@ -0,0 +1,25 @@
+"""
+Claude client module facade.
+
+Provides Claude API client utilities.
+Uses lazy imports to avoid circular dependencies.
+"""
+
+
+def __getattr__(name):
+    """Lazy import to avoid circular imports with auto_claude_tools."""
+    from core import client as _client
+
+    return getattr(_client, name)
+
+
+def create_client(*args, **kwargs):
+    """Create a Claude client instance."""
+    from core.client import create_client as _create_client
+
+    return _create_client(*args, **kwargs)
+
+
+__all__ = [
+    "create_client",
+]
@@ -0,0 +1,383 @@
+"""
+Commit Message Generator
+========================
+
+Generates high-quality commit messages using Claude Haiku.
+
+Features:
+- Conventional commits format (feat/fix/refactor/etc)
+- GitHub issue references (Fixes #123)
+- Context-aware descriptions from spec metadata
+"""
+
+from __future__ import annotations
+
+import asyncio
+import json
+import logging
+import re
+import sys
+from pathlib import Path
+from typing import TYPE_CHECKING
+
+if TYPE_CHECKING:
+    pass
+
+logger = logging.getLogger(__name__)
+
+# Map task categories to conventional commit types
+CATEGORY_TO_COMMIT_TYPE = {
+    "feature": "feat",
+    "bug_fix": "fix",
+    "bug": "fix",
+    "refactoring": "refactor",
+    "refactor": "refactor",
+    "documentation": "docs",
+    "docs": "docs",
+    "testing": "test",
+    "test": "test",
+    "performance": "perf",
+    "perf": "perf",
+    "security": "security",
+    "chore": "chore",
+    "style": "style",
+    "ci": "ci",
+    "build": "build",
+}
+
+SYSTEM_PROMPT = """You are a Git expert who writes clear, concise commit messages following conventional commits format.
+
+Rules:
+1. First line: type(scope): description (max 72 chars total)
+2. Leave blank line after first line
+3. Body: 1-3 sentences explaining WHAT changed and WHY
+4. If GitHub issue number provided, end with "Fixes #N" on its own line
+5. Be specific about the changes, not generic
+6. Use imperative mood ("Add feature" not "Added feature")
+
+Types: feat, fix, refactor, docs, test, perf, chore, style, ci, build
+
+Example output:
+feat(auth): add OAuth2 login flow
+
+Implement OAuth2 authentication with Google and GitHub providers.
+Add token refresh logic and secure storage.
+
+Fixes #42"""
+
+
+def _get_spec_context(spec_dir: Path) -> dict:
+    """
+    Extract context from spec files for commit message generation.
+
+    Returns dict with:
+    - title: Feature/task title
+    - category: Task category (feature, bug_fix, etc)
+    - description: Brief description
+    - github_issue: GitHub issue number if linked
+    """
+    context = {
+        "title": "",
+        "category": "chore",
+        "description": "",
+        "github_issue": None,
+    }
+
+    # Try to read spec.md for title
+    spec_file = spec_dir / "spec.md"
+    if spec_file.exists():
+        try:
+            content = spec_file.read_text(encoding="utf-8")
+            # Extract title from first H1 or H2
+            title_match = re.search(r"^#+ (.+)$", content, re.MULTILINE)
+            if title_match:
+                context["title"] = title_match.group(1).strip()
+
+            # Look for overview/description section
+            overview_match = re.search(
+                r"## Overview\s*\n(.+?)(?=\n##|\Z)", content, re.DOTALL
+            )
+            if overview_match:
+                context["description"] = overview_match.group(1).strip()[:200]
+        except Exception as e:
+            logger.debug(f"Could not read spec.md: {e}")
+
+    # Try to read requirements.json for metadata
+    req_file = spec_dir / "requirements.json"
+    if req_file.exists():
+        try:
+            req_data = json.loads(req_file.read_text(encoding="utf-8"))
+            if not context["title"] and req_data.get("feature"):
+                context["title"] = req_data["feature"]
+            if req_data.get("workflow_type"):
+                context["category"] = req_data["workflow_type"]
+            if req_data.get("task_description") and not context["description"]:
+                context["description"] = req_data["task_description"][:200]
+        except Exception as e:
+            logger.debug(f"Could not read requirements.json: {e}")
+
+    # Try to read implementation_plan.json for GitHub issue
+    plan_file = spec_dir / "implementation_plan.json"
+    if plan_file.exists():
+        try:
+            plan_data = json.loads(plan_file.read_text(encoding="utf-8"))
+            # Check for GitHub metadata
+            metadata = plan_data.get("metadata", {})
+            if metadata.get("githubIssueNumber"):
+                context["github_issue"] = metadata["githubIssueNumber"]
+            # Fallback title
+            if not context["title"]:
+                context["title"] = plan_data.get("feature") or plan_data.get(
+                    "title", ""
+                )
+        except Exception as e:
+            logger.debug(f"Could not read implementation_plan.json: {e}")
+
+    return context
+
+
+def _build_prompt(
+    spec_context: dict,
+    diff_summary: str,
+    files_changed: list[str],
+) -> str:
+    """Build the prompt for Claude."""
+    commit_type = CATEGORY_TO_COMMIT_TYPE.get(
+        spec_context.get("category", "").lower(), "chore"
+    )
+
+    github_ref = ""
+    if spec_context.get("github_issue"):
+        github_ref = f"\nGitHub Issue: #{spec_context['github_issue']} (include 'Fixes #{spec_context['github_issue']}' at the end)"
+
+    # Truncate file list if too long
+    if len(files_changed) > 20:
+        files_display = (
+            "\n".join(files_changed[:20])
+            + f"\n... and {len(files_changed) - 20} more files"
+        )
+    else:
+        files_display = (
+            "\n".join(files_changed) if files_changed else "(no files listed)"
+        )
+
+    prompt = f"""Generate a commit message for this change.
+
+Task: {spec_context.get("title", "Unknown task")}
+Type: {commit_type}
+Files changed: {len(files_changed)}
+{github_ref}
+
+Description: {spec_context.get("description", "No description available")}
+
+Changed files:
+{files_display}
+
+Diff summary:
+{diff_summary[:2000] if diff_summary else "(no diff available)"}
+
+Generate ONLY the commit message, nothing else. Follow the format exactly:
+type(scope): short description
+
+Body explaining changes.
+
+Fixes #N (if applicable)"""
+
+    return prompt
+
+
+async def _call_claude(prompt: str) -> str:
+    """Call Claude for commit message generation.
+
+    Reads model/thinking settings from environment variables:
+    - UTILITY_MODEL_ID: Full model ID (e.g., "claude-haiku-4-5-20251001")
+    - UTILITY_THINKING_BUDGET: Thinking budget tokens (e.g., "1024")
+    """
+    from core.auth import ensure_claude_code_oauth_token, get_auth_token
+    from core.model_config import get_utility_model_config
+
+    if not get_auth_token():
+        logger.warning("No authentication token found")
+        return ""
+
+    ensure_claude_code_oauth_token()
+
+    try:
+        from core.simple_client import create_simple_client
+    except ImportError:
+        logger.warning("core.simple_client not available")
+        return ""
+
+    # Get model settings from environment (passed from frontend)
+    model, thinking_budget = get_utility_model_config()
+
+    logger.info(
+        f"Commit message using model={model}, thinking_budget={thinking_budget}"
+    )
+
+    client = create_simple_client(
+        agent_type="commit_message",
+        model=model,
+        system_prompt=SYSTEM_PROMPT,
+        max_thinking_tokens=thinking_budget,
+    )
+
+    try:
+        async with client:
+            await client.query(prompt)
+
+            response_text = ""
+            async for msg in client.receive_response():
+                msg_type = type(msg).__name__
+                if msg_type == "AssistantMessage" and hasattr(msg, "content"):
+                    for block in msg.content:
+                        # Must check block type - only TextBlock has .text attribute
+                        block_type = type(block).__name__
+                        if block_type == "TextBlock" and hasattr(block, "text"):
+                            response_text += block.text
+
+            logger.info(f"Generated commit message: {len(response_text)} chars")
+            return response_text.strip()
+
+    except Exception as e:
+        logger.error(f"Claude SDK call failed: {e}")
+        print(f"    [WARN] Commit message generation failed: {e}", file=sys.stderr)
+        return ""
+
+
+def generate_commit_message_sync(
+    project_dir: Path,
+    spec_name: str,
+    diff_summary: str = "",
+    files_changed: list[str] | None = None,
+    github_issue: int | None = None,
+) -> str:
+    """
+    Generate a commit message synchronously.
+
+    Args:
+        project_dir: Project root directory
+        spec_name: Spec identifier (e.g., "001-add-feature")
+        diff_summary: Git diff stat or summary
+        files_changed: List of changed file paths
+        github_issue: GitHub issue number if linked (overrides spec metadata)
+
+    Returns:
+        Generated commit message or fallback message
+    """
+    # Find spec directory
+    spec_dir = project_dir / ".auto-claude" / "specs" / spec_name
+    if not spec_dir.exists():
+        # Try alternative location
+        spec_dir = project_dir / "auto-claude" / "specs" / spec_name
+
+    # Get context from spec files
+    spec_context = _get_spec_context(spec_dir) if spec_dir.exists() else {}
+
+    # Override with provided github_issue
+    if github_issue:
+        spec_context["github_issue"] = github_issue
+
+    # Build prompt
+    prompt = _build_prompt(
+        spec_context,
+        diff_summary,
+        files_changed or [],
+    )
+
+    # Call Claude
+    try:
+        # Check if we're already in an async context
+        try:
+            loop = asyncio.get_running_loop()
+        except RuntimeError:
+            loop = None
+
+        if loop and loop.is_running():
+            # Already in an async context - run in a new thread
+            # Use lambda to ensure coroutine is created inside the worker thread
+            import concurrent.futures
+
+            with concurrent.futures.ThreadPoolExecutor() as pool:
+                result = pool.submit(lambda: asyncio.run(_call_claude(prompt))).result()
+        else:
+            result = asyncio.run(_call_claude(prompt))
+
+        if result:
+            return result
+    except Exception as e:
+        logger.error(f"Failed to generate commit message: {e}")
+
+    # Fallback message
+    commit_type = CATEGORY_TO_COMMIT_TYPE.get(
+        spec_context.get("category", "").lower(), "chore"
+    )
+    title = spec_context.get("title", spec_name)
+    fallback = f"{commit_type}: {title}"
+
+    if github_issue or spec_context.get("github_issue"):
+        issue_num = github_issue or spec_context.get("github_issue")
+        fallback += f"\n\nFixes #{issue_num}"
+
+    return fallback
+
+
+async def generate_commit_message(
+    project_dir: Path,
+    spec_name: str,
+    diff_summary: str = "",
+    files_changed: list[str] | None = None,
+    github_issue: int | None = None,
+) -> str:
+    """
+    Generate a commit message asynchronously.
+
+    Args:
+        project_dir: Project root directory
+        spec_name: Spec identifier (e.g., "001-add-feature")
+        diff_summary: Git diff stat or summary
+        files_changed: List of changed file paths
+        github_issue: GitHub issue number if linked (overrides spec metadata)
+
+    Returns:
+        Generated commit message or fallback message
+    """
+    # Find spec directory
+    spec_dir = project_dir / ".auto-claude" / "specs" / spec_name
+    if not spec_dir.exists():
+        spec_dir = project_dir / "auto-claude" / "specs" / spec_name
+
+    # Get context from spec files
+    spec_context = _get_spec_context(spec_dir) if spec_dir.exists() else {}
+
+    # Override with provided github_issue
+    if github_issue:
+        spec_context["github_issue"] = github_issue
+
+    # Build prompt
+    prompt = _build_prompt(
+        spec_context,
+        diff_summary,
+        files_changed or [],
+    )
+
+    # Call Claude
+    try:
+        result = await _call_claude(prompt)
+        if result:
+            return result
+    except Exception as e:
+        logger.error(f"Failed to generate commit message: {e}")
+
+    # Fallback message
+    commit_type = CATEGORY_TO_COMMIT_TYPE.get(
+        spec_context.get("category", "").lower(), "chore"
+    )
+    title = spec_context.get("title", spec_name)
+    fallback = f"{commit_type}: {title}"
+
+    if github_issue or spec_context.get("github_issue"):
+        issue_num = github_issue or spec_context.get("github_issue")
+        fallback += f"\n\nFixes #{issue_num}"
+
+    return fallback
@@ -0,0 +1,37 @@
+"""
+Context Package
+===============
+
+Task context building for autonomous coding.
+"""
+
+from .builder import ContextBuilder
+from .categorizer import FileCategorizer
+from .graphiti_integration import fetch_graph_hints, is_graphiti_enabled
+from .keyword_extractor import KeywordExtractor
+from .models import FileMatch, TaskContext
+from .pattern_discovery import PatternDiscoverer
+from .search import CodeSearcher
+from .serialization import load_context, save_context, serialize_context
+from .service_matcher import ServiceMatcher
+
+__all__ = [
+    # Main builder
+    "ContextBuilder",
+    # Models
+    "FileMatch",
+    "TaskContext",
+    # Components
+    "CodeSearcher",
+    "ServiceMatcher",
+    "KeywordExtractor",
+    "FileCategorizer",
+    "PatternDiscoverer",
+    # Graphiti integration
+    "fetch_graph_hints",
+    "is_graphiti_enabled",
+    # Serialization
+    "serialize_context",
+    "save_context",
+    "load_context",
+]
@@ -0,0 +1,250 @@
+"""
+Context Builder
+===============
+
+Main builder class that orchestrates context building for tasks.
+"""
+
+import asyncio
+import json
+from dataclasses import asdict
+from pathlib import Path
+
+from .categorizer import FileCategorizer
+from .graphiti_integration import fetch_graph_hints, is_graphiti_enabled
+from .keyword_extractor import KeywordExtractor
+from .models import FileMatch, TaskContext
+from .pattern_discovery import PatternDiscoverer
+from .search import CodeSearcher
+from .service_matcher import ServiceMatcher
+
+
+class ContextBuilder:
+    """Builds task-specific context by searching the codebase."""
+
+    def __init__(self, project_dir: Path, project_index: dict | None = None):
+        self.project_dir = project_dir.resolve()
+        self.project_index = project_index or self._load_project_index()
+
+        # Initialize components
+        self.searcher = CodeSearcher(self.project_dir)
+        self.service_matcher = ServiceMatcher(self.project_index)
+        self.keyword_extractor = KeywordExtractor()
+        self.categorizer = FileCategorizer()
+        self.pattern_discoverer = PatternDiscoverer(self.project_dir)
+
+    def _load_project_index(self) -> dict:
+        """Load project index from file or create new one (.auto-claude is the installed instance)."""
+        index_file = self.project_dir / ".auto-claude" / "project_index.json"
+        if index_file.exists():
+            try:
+                with open(index_file, encoding="utf-8") as f:
+                    return json.load(f)
+            except (OSError, json.JSONDecodeError, UnicodeDecodeError):
+                # Corrupted or legacy-encoded file, regenerate
+                pass
+
+        # Try to create one
+        from analyzer import analyze_project
+
+        return analyze_project(self.project_dir)
+
+    def build_context(
+        self,
+        task: str,
+        services: list[str] | None = None,
+        keywords: list[str] | None = None,
+        include_graph_hints: bool = True,
+    ) -> TaskContext:
+        """
+        Build context for a specific task.
+
+        Args:
+            task: Description of the task
+            services: List of service names to search (None = auto-detect)
+            keywords: Additional keywords to search for
+            include_graph_hints: Whether to include historical hints from Graphiti
+
+        Returns:
+            TaskContext with relevant files and patterns
+        """
+        # Auto-detect services if not specified
+        if not services:
+            services = self.service_matcher.suggest_services(task)
+
+        # Extract keywords from task if not provided
+        if not keywords:
+            keywords = self.keyword_extractor.extract_keywords(task)
+
+        # Search each service
+        all_matches: list[FileMatch] = []
+        service_contexts = {}
+
+        for service_name in services:
+            service_info = self.project_index.get("services", {}).get(service_name)
+            if not service_info:
+                continue
+
+            service_path = Path(service_info.get("path", service_name))
+            if not service_path.is_absolute():
+                service_path = self.project_dir / service_path
+
+            # Search this service
+            matches = self.searcher.search_service(service_path, service_name, keywords)
+            all_matches.extend(matches)
+
+            # Load or generate service context
+            service_contexts[service_name] = self._get_service_context(
+                service_path, service_name, service_info
+            )
+
+        # Categorize matches
+        files_to_modify, files_to_reference = self.categorizer.categorize_matches(
+            all_matches, task
+        )
+
+        # Discover patterns from reference files
+        patterns = self.pattern_discoverer.discover_patterns(
+            files_to_reference, keywords
+        )
+
+        # Get graph hints (synchronously wrap async call)
+        graph_hints = []
+        if include_graph_hints and is_graphiti_enabled():
+            try:
+                # Run the async function in a new event loop if necessary
+                try:
+                    loop = asyncio.get_running_loop()
+                    # We're already in an async context - this shouldn't happen in CLI
+                    # but handle it gracefully
+                    graph_hints = []
+                except RuntimeError:
+                    # No event loop running - create one
+                    graph_hints = asyncio.run(
+                        fetch_graph_hints(task, str(self.project_dir))
+                    )
+            except Exception:
+                # Graphiti is optional - fail gracefully
+                graph_hints = []
+
+        return TaskContext(
+            task_description=task,
+            scoped_services=services,
+            files_to_modify=[
+                asdict(f) if isinstance(f, FileMatch) else f for f in files_to_modify
+            ],
+            files_to_reference=[
+                asdict(f) if isinstance(f, FileMatch) else f for f in files_to_reference
+            ],
+            patterns_discovered=patterns,
+            service_contexts=service_contexts,
+            graph_hints=graph_hints,
+        )
+
+    async def build_context_async(
+        self,
+        task: str,
+        services: list[str] | None = None,
+        keywords: list[str] | None = None,
+        include_graph_hints: bool = True,
+    ) -> TaskContext:
+        """
+        Build context for a specific task (async version).
+
+        This version is preferred when called from async code as it can
+        properly await the graph hints retrieval.
+
+        Args:
+            task: Description of the task
+            services: List of service names to search (None = auto-detect)
+            keywords: Additional keywords to search for
+            include_graph_hints: Whether to include historical hints from Graphiti
+
+        Returns:
+            TaskContext with relevant files and patterns
+        """
+        # Auto-detect services if not specified
+        if not services:
+            services = self.service_matcher.suggest_services(task)
+
+        # Extract keywords from task if not provided
+        if not keywords:
+            keywords = self.keyword_extractor.extract_keywords(task)
+
+        # Search each service
+        all_matches: list[FileMatch] = []
+        service_contexts = {}
+
+        for service_name in services:
+            service_info = self.project_index.get("services", {}).get(service_name)
+            if not service_info:
+                continue
+
+            service_path = Path(service_info.get("path", service_name))
+            if not service_path.is_absolute():
+                service_path = self.project_dir / service_path
+
+            # Search this service
+            matches = self.searcher.search_service(service_path, service_name, keywords)
+            all_matches.extend(matches)
+
+            # Load or generate service context
+            service_contexts[service_name] = self._get_service_context(
+                service_path, service_name, service_info
+            )
+
+        # Categorize matches
+        files_to_modify, files_to_reference = self.categorizer.categorize_matches(
+            all_matches, task
+        )
+
+        # Discover patterns from reference files
+        patterns = self.pattern_discoverer.discover_patterns(
+            files_to_reference, keywords
+        )
+
+        # Get graph hints asynchronously
+        graph_hints = []
+        if include_graph_hints:
+            graph_hints = await fetch_graph_hints(task, str(self.project_dir))
+
+        return TaskContext(
+            task_description=task,
+            scoped_services=services,
+            files_to_modify=[
+                asdict(f) if isinstance(f, FileMatch) else f for f in files_to_modify
+            ],
+            files_to_reference=[
+                asdict(f) if isinstance(f, FileMatch) else f for f in files_to_reference
+            ],
+            patterns_discovered=patterns,
+            service_contexts=service_contexts,
+            graph_hints=graph_hints,
+        )
+
+    def _get_service_context(
+        self,
+        service_path: Path,
+        service_name: str,
+        service_info: dict,
+    ) -> dict:
+        """Get or generate context for a service."""
+        # Check for SERVICE_CONTEXT.md
+        context_file = service_path / "SERVICE_CONTEXT.md"
+        if context_file.exists():
+            return {
+                "source": "SERVICE_CONTEXT.md",
+                "content": context_file.read_text(encoding="utf-8")[
+                    :2000
+                ],  # First 2000 chars
+            }
+
+        # Generate basic context from service info
+        return {
+            "source": "generated",
+            "language": service_info.get("language"),
+            "framework": service_info.get("framework"),
+            "type": service_info.get("type"),
+            "entry_point": service_info.get("entry_point"),
+            "key_directories": service_info.get("key_directories", {}),
+        }
@@ -0,0 +1,73 @@
+"""
+File Categorization
+===================
+
+Categorizes files into those to modify vs those to reference.
+"""
+
+from .models import FileMatch
+
+
+class FileCategorizer:
+    """Categorizes matched files based on task context."""
+
+    # Keywords that suggest modification
+    MODIFY_KEYWORDS = [
+        "add",
+        "create",
+        "implement",
+        "fix",
+        "update",
+        "change",
+        "modify",
+        "new",
+    ]
+
+    def categorize_matches(
+        self,
+        matches: list[FileMatch],
+        task: str,
+        max_modify: int = 10,
+        max_reference: int = 15,
+    ) -> tuple[list[FileMatch], list[FileMatch]]:
+        """
+        Categorize matches into files to modify vs reference.
+
+        Args:
+            matches: List of FileMatch objects to categorize
+            task: Task description string
+            max_modify: Maximum files to modify
+            max_reference: Maximum reference files
+
+        Returns:
+            Tuple of (files_to_modify, files_to_reference)
+        """
+        to_modify = []
+        to_reference = []
+
+        task_lower = task.lower()
+        is_modification = any(kw in task_lower for kw in self.MODIFY_KEYWORDS)
+
+        for match in matches:
+            # High relevance files in the "right" location are likely to be modified
+            path_lower = match.path.lower()
+
+            is_test = "test" in path_lower or "spec" in path_lower
+            is_example = "example" in path_lower or "sample" in path_lower
+            is_config = "config" in path_lower and match.relevance_score < 5
+
+            if is_test or is_example or is_config:
+                # Tests/examples are references
+                match.reason = f"Reference pattern: {match.reason}"
+                to_reference.append(match)
+            elif match.relevance_score >= 5 and is_modification:
+                # High relevance + modification task = likely to modify
+                match.reason = f"Likely to modify: {match.reason}"
+                to_modify.append(match)
+            else:
+                # Everything else is a reference
+                match.reason = f"Related: {match.reason}"
+                to_reference.append(match)
+
+        # Limit results
+        return to_modify[:max_modify], to_reference[:max_reference]
@@ -0,0 +1,44 @@
+"""
+Constants for Context Building
+================================
+
+Configuration constants for directory skipping and file filtering.
+"""
+
+# Directories to skip during code search
+SKIP_DIRS = {
+    "node_modules",
+    ".git",
+    "__pycache__",
+    ".venv",
+    "venv",
+    "dist",
+    "build",
+    ".next",
+    ".nuxt",
+    "target",
+    "vendor",
+    ".idea",
+    ".vscode",
+    "auto-claude",
+    ".pytest_cache",
+    ".mypy_cache",
+    "coverage",
+    ".turbo",
+    ".cache",
+}
+
+# File extensions to search for code files
+CODE_EXTENSIONS = {
+    ".py",
+    ".js",
+    ".jsx",
+    ".ts",
+    ".tsx",
+    ".vue",
+    ".svelte",
+    ".go",
+    ".rs",
+    ".rb",
+    ".php",
+}
@@ -0,0 +1,53 @@
+"""
+Graphiti Knowledge Graph Integration
+======================================
+
+Integration with Graphiti for historical hints and cross-session context.
+"""
+
+# Import graphiti providers for optional historical hints
+try:
+    from graphiti_providers import get_graph_hints, is_graphiti_enabled
+
+    GRAPHITI_AVAILABLE = True
+except ImportError:
+    GRAPHITI_AVAILABLE = False
+
+    def is_graphiti_enabled() -> bool:
+        return False
+
+    async def get_graph_hints(
+        query: str, project_id: str, max_results: int = 10
+    ) -> list:
+        return []
+
+
+async def fetch_graph_hints(
+    query: str, project_id: str, max_results: int = 5
+) -> list[dict]:
+    """
+    Get historical hints from Graphiti knowledge graph.
+
+    This provides context from past sessions and similar tasks.
+
+    Args:
+        query: The task description or query to search for
+        project_id: The project identifier (typically project path)
+        max_results: Maximum number of hints to return
+
+    Returns:
+        List of graph hints as dictionaries
+    """
+    if not is_graphiti_enabled():
+        return []
+
+    try:
+        hints = await get_graph_hints(
+            query=query,
+            project_id=project_id,
+            max_results=max_results,
+        )
+        return hints
+    except Exception:
+        # Graphiti is optional - fail gracefully
+        return []
@@ -0,0 +1,101 @@
+"""
+Keyword Extraction
+==================
+
+Extracts meaningful keywords from task descriptions for search.
+"""
+
+import re
+
+
+class KeywordExtractor:
+    """Extracts and filters keywords from task descriptions."""
+
+    # Common words to filter out
+    STOPWORDS = {
+        "a",
+        "an",
+        "the",
+        "to",
+        "for",
+        "of",
+        "in",
+        "on",
+        "at",
+        "by",
+        "with",
+        "and",
+        "or",
+        "but",
+        "is",
+        "are",
+        "was",
+        "were",
+        "be",
+        "been",
+        "being",
+        "have",
+        "has",
+        "had",
+        "do",
+        "does",
+        "did",
+        "will",
+        "would",
+        "could",
+        "should",
+        "may",
+        "might",
+        "must",
+        "can",
+        "this",
+        "that",
+        "these",
+        "those",
+        "i",
+        "you",
+        "we",
+        "they",
+        "it",
+        "add",
+        "create",
+        "make",
+        "implement",
+        "build",
+        "fix",
+        "update",
+        "change",
+        "modify",
+        "when",
+        "if",
+        "then",
+        "else",
+        "new",
+        "existing",
+    }
+
+    @classmethod
+    def extract_keywords(cls, task: str, max_keywords: int = 10) -> list[str]:
+        """
+        Extract search keywords from task description.
+
+        Args:
+            task: Task description string
+            max_keywords: Maximum number of keywords to return
+
+        Returns:
+            List of extracted keywords
+        """
+        # Tokenize and filter
+        words = re.findall(r"\b[a-zA-Z_][a-zA-Z0-9_]*\b", task.lower())
+        keywords = [w for w in words if w not in cls.STOPWORDS and len(w) > 2]
+
+        # Deduplicate while preserving order
+        seen = set()
+        unique_keywords = []
+        for kw in keywords:
+            if kw not in seen:
+                seen.add(kw)
+                unique_keywords.append(kw)
+
+        return unique_keywords[:max_keywords]
@@ -0,0 +1,144 @@
+#!/usr/bin/env python3
+"""
+Task Context Builder
+====================
+
+Builds focused context for a specific task by searching relevant services.
+This is the "RAG-like" component that finds what files matter for THIS task.
+
+Usage:
+    # Find context for a task across specific services
+    python auto-claude/context.py \
+        --services backend,scraper \
+        --keywords "retry,error,proxy" \
+        --task "Add retry logic when proxies fail" \
+        --output auto-claude/specs/001-retry/context.json
+
+    # Use project index to auto-suggest services
+    python auto-claude/context.py \
+        --task "Add retry logic when proxies fail" \
+        --output context.json
+
+The context builder will:
+1. Load project index (from analyzer)
+2. Search specified services for relevant files
+3. Find similar implementations to reference
+4. Output focused context for AI agents
+"""
+
+import json
+from pathlib import Path
+
+from context import (
+    ContextBuilder,
+    FileMatch,
+    TaskContext,
+)
+from context.serialization import serialize_context
+
+# Backward compatibility exports
+__all__ = [
+    "ContextBuilder",
+    "FileMatch",
+    "TaskContext",
+    "build_task_context",
+]
+
+
+def build_task_context(
+    project_dir: Path,
+    task: str,
+    services: list[str] | None = None,
+    keywords: list[str] | None = None,
+    output_file: Path | None = None,
+) -> dict:
+    """
+    Build context for a task and optionally save to file.
+
+    Args:
+        project_dir: Path to project root
+        task: Task description
+        services: Services to search (None = auto-detect)
+        keywords: Keywords to search for (None = extract from task)
+        output_file: Optional path to save JSON output
+
+    Returns:
+        Context as a dictionary
+    """
+    builder = ContextBuilder(project_dir)
+    context = builder.build_context(task, services, keywords)
+
+    result = serialize_context(context)
+
+    if output_file:
+        output_file.parent.mkdir(parents=True, exist_ok=True)
+        with open(output_file, "w", encoding="utf-8") as f:
+            json.dump(result, f, indent=2)
+        print(f"Task context saved to: {output_file}")
+
+    return result
+
+
+def main():
+    """CLI entry point."""
+    import argparse
+
+    parser = argparse.ArgumentParser(
+        description="Build task-specific context by searching the codebase"
+    )
+    parser.add_argument(
+        "--project-dir",
+        type=Path,
+        default=Path.cwd(),
+        help="Project directory (default: current directory)",
+    )
+    parser.add_argument(
+        "--task",
+        type=str,
+        required=True,
+        help="Description of the task",
+    )
+    parser.add_argument(
+        "--services",
+        type=str,
+        default=None,
+        help="Comma-separated list of services to search",
+    )
+    parser.add_argument(
+        "--keywords",
+        type=str,
+        default=None,
+        help="Comma-separated list of keywords to search for",
+    )
+    parser.add_argument(
+        "--output",
+        type=Path,
+        default=None,
+        help="Output file for JSON results",
+    )
+    parser.add_argument(
+        "--quiet",
+        action="store_true",
+        help="Only output JSON, no status messages",
+    )
+
+    args = parser.parse_args()
+
+    # Parse comma-separated args
+    services = args.services.split(",") if args.services else None
+    keywords = args.keywords.split(",") if args.keywords else None
+
+    result = build_task_context(
+        args.project_dir,
+        args.task,
+        services,
+        keywords,
+        args.output,
+    )
+
+    if not args.quiet or not args.output:
+        print(json.dumps(result, indent=2))
+
+
+if __name__ == "__main__":
+    main()
@@ -0,0 +1,34 @@
+"""
+Data Models for Task Context
+=============================
+
+Core data structures for representing file matches and task context.
+"""
+
+from dataclasses import dataclass, field
+
+
+@dataclass
+class FileMatch:
+    """A file that matched the search criteria."""
+
+    path: str
+    service: str
+    reason: str
+    relevance_score: float = 0.0
+    matching_lines: list[tuple[int, str]] = field(default_factory=list)
+
+
+@dataclass
+class TaskContext:
+    """Complete context for a task."""
+
+    task_description: str
+    scoped_services: list[str]
+    files_to_modify: list[dict]
+    files_to_reference: list[dict]
+    patterns_discovered: dict[str, str]
+    service_contexts: dict[str, dict]
+    graph_hints: list[dict] = field(
+        default_factory=list
+    )  # Historical hints from Graphiti
@@ -0,0 +1,65 @@
+"""
+Pattern Discovery
+=================
+
+Discovers code patterns from reference files to guide implementation.
+"""
+
+from pathlib import Path
+
+from .models import FileMatch
+
+
+class PatternDiscoverer:
+    """Discovers code patterns from reference files."""
+
+    def __init__(self, project_dir: Path):
+        self.project_dir = project_dir.resolve()
+
+    def discover_patterns(
+        self,
+        reference_files: list[FileMatch],
+        keywords: list[str],
+        max_files: int = 5,
+    ) -> dict[str, str]:
+        """
+        Discover code patterns from reference files.
+
+        Args:
+            reference_files: List of FileMatch objects to analyze
+            keywords: Keywords to look for in the code
+            max_files: Maximum number of files to analyze
+
+        Returns:
+            Dictionary mapping pattern keys to code snippets
+        """
+        patterns = {}
+
+        for match in reference_files[:max_files]:
+            try:
+                file_path = self.project_dir / match.path
+                content = file_path.read_text(encoding="utf-8", errors="ignore")
+
+                # Look for common patterns
+                for keyword in keywords:
+                    if keyword in content.lower():
+                        # Extract a snippet around the keyword
+                        lines = content.split("\n")
+                        for i, line in enumerate(lines):
+                            if keyword in line.lower():
+                                # Get context (3 lines before and after)
+                                start = max(0, i - 3)
+                                end = min(len(lines), i + 4)
+                                snippet = "\n".join(lines[start:end])
+
+                                pattern_key = f"{keyword}_pattern"
+                                if pattern_key not in patterns:
+                                    patterns[pattern_key] = (
+                                        f"From {match.path}:\n{snippet[:300]}"
+                                    )
+                                break
+
+            except (OSError, UnicodeDecodeError):
+                continue
+
+        return patterns
@@ -0,0 +1,101 @@
+"""
+Code Search Functionality
+==========================
+
+Search codebase for relevant files based on keywords.
+"""
+
+from pathlib import Path
+
+from .constants import CODE_EXTENSIONS, SKIP_DIRS
+from .models import FileMatch
+
+
+class CodeSearcher:
+    """Searches code files for relevant matches."""
+
+    def __init__(self, project_dir: Path):
+        self.project_dir = project_dir.resolve()
+
+    def search_service(
+        self,
+        service_path: Path,
+        service_name: str,
+        keywords: list[str],
+    ) -> list[FileMatch]:
+        """
+        Search a service for files matching keywords.
+
+        Args:
+            service_path: Path to the service directory
+            service_name: Name of the service
+            keywords: List of keywords to search for
+
+        Returns:
+            List of FileMatch objects sorted by relevance
+        """
+        matches = []
+
+        if not service_path.exists():
+            return matches
+
+        for file_path in self._iter_code_files(service_path):
+            try:
+                content = file_path.read_text(encoding="utf-8", errors="ignore")
+                content_lower = content.lower()
+
+                # Score this file
+                score = 0
+                matching_keywords = []
+                matching_lines = []
+
+                for keyword in keywords:
+                    if keyword in content_lower:
+                        # Count occurrences
+                        count = content_lower.count(keyword)
+                        score += min(count, 10)  # Cap at 10 per keyword
+                        matching_keywords.append(keyword)
+
+                        # Find matching lines (first 3 per keyword)
+                        lines = content.split("\n")
+                        found = 0
+                        for i, line in enumerate(lines, 1):
+                            if keyword in line.lower() and found < 3:
+                                matching_lines.append((i, line.strip()[:100]))
+                                found += 1
+
+                if score > 0:
+                    rel_path = str(file_path.relative_to(self.project_dir))
+                    matches.append(
+                        FileMatch(
+                            path=rel_path,
+                            service=service_name,
+                            reason=f"Contains: {', '.join(matching_keywords)}",
+                            relevance_score=score,
+                            matching_lines=matching_lines[:5],  # Top 5 lines
+                        )
+                    )
+
+            except (OSError, UnicodeDecodeError):
+                continue
+
+        # Sort by relevance
+        matches.sort(key=lambda m: m.relevance_score, reverse=True)
+        return matches[:20]  # Top 20 per service
+
+    def _iter_code_files(self, directory: Path):
+        """
+        Iterate over code files in a directory.
+
+        Args:
+            directory: Root directory to search
+
+        Yields:
+            Path objects for code files
+        """
+        for item in directory.rglob("*"):
+            if item.is_file() and item.suffix in CODE_EXTENSIONS:
+                # Check if in skip directory
+                parts = item.relative_to(directory).parts
+                if not any(part in SKIP_DIRS for part in parts):
+                    yield item
--- a/Show More
+++ b/Show More