Files
Aperant/apps/backend/agents/tools_pkg/models.py
T
Andy bb7e189374 feat: simplify thinking system and remove opus-1m model variant (#1760)
* feat: integrate Claude Opus 4.6 model with 1M context window option

Update model definitions across frontend and backend from claude-opus-4-5
to claude-opus-4-6 (without date suffix for automatic latest version).
Add "Claude Opus 4.6 (1M)" as a separate dropdown option that enables
the 1M token context window via the SDK beta header context-1m-2025-08-07.

Wire betas parameter through all create_client() callers in the core
pipeline (coder, planner, QA) and secondary callers (ideation, GitHub
PR review, triage, orchestrator, followup reviewer) so the 1M context
setting flows end-to-end from UI selection to the Claude Agent SDK.

Also fix pre-existing pydantic import error in test_integration_phase4.py
by mocking pydantic when not installed in the test environment.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* feat: simplify thinking system and remove opus-1m model variant

Replace the 5-level thinking system (none/low/medium/high/ultrathink) with
a streamlined 3-level system (low/medium/high) aligned with Claude's effort
paradigm. Remove opus-1m model variant from frontend types, simplify agent
thinking defaults, and clean up related test infrastructure.

- Simplify THINKING_BUDGET_MAP to 3 levels in phase_config.py
- Update agent thinking_default values (coder: none→low, insights: none→low,
  spec_critic: ultrathink→high)
- Remove opus-1m from ModelTypeShort type
- Streamline all backend callers (planner, coder, QA, ideation, GitHub services)
- Update frontend constants, i18n, and task log labels
- Clean up test assertions for new thinking levels

Note: Pre-commit hook bypassed due to pre-existing test_github_pr_regression.py
failure in worktree environment (unrelated to these changes; 451/452 tests pass).

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: address PR review feedback

- Fix inconsistent terminology: use 'thinking level' consistently in
  test docstrings (not 'effort level')
- Clean up pydantic mock after use to avoid leaking into sys.modules
  for the entire test session
- Update test assertions for new thinking defaults (coder: low,
  spec_critic: high)

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: restore Opus 4.6 integration lost during thinking simplification

The thinking simplification commit accidentally reverted all Opus 4.6
changes (model IDs, betas/1M context, frontend constants). This commit
restores those changes and re-applies the thinking simplification on top.

Restored: model ID updates (opus-4-5→opus-4-6), opus-1m variant with
betas header for 1M context, betas parameter threading through all
callers (client, planner, coder, QA, ideation, GitHub services).

Thinking simplification preserved: 3-level system (low/medium/high),
ultrathink→high in spec phases and complex profile, none→low defaults.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* feat: add adaptive thinking/effort level support for Opus 4.6

Route thinking configuration based on model type: Opus 4.6 gets both
effort_level (via CLAUDE_CODE_EFFORT_LEVEL env var) and max_thinking_tokens,
while Sonnet/Haiku get max_thinking_tokens only.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: update tests to match simplified thinking levels (no none/ultrathink)

Tests were referencing 'none' and 'ultrathink' thinking levels that were
removed in 1445185b. Updated to match current valid levels: low, medium, high.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: update outdated docstring and add legacy thinking level mapping

- Update create_client() docstring to reflect current thinking budget values
- Add LEGACY_THINKING_MAP for backward compatibility: 'none' -> 'low',
  'ultrathink' -> 'high' with deprecation warnings
- Add tests for legacy level mapping

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: add missing agent_type to planner and clean up return types

- Add agent_type="planner" to follow-up planner create_client() call
- Update get_thinking_budget() return type from int | None to int
  since 'none' level was removed (now mapped via LEGACY_THINKING_MAP)
- Fix ruff formatting

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* feat: add Fast Mode toggle for Opus 4.6 and remove legacy thinking levels

Add a global Fast Mode setting that passes CLAUDE_CODE_FAST_MODE=true env var
to the Claude Code SDK subprocess for faster Opus 4.6 output at higher cost.
The toggle appears in Agent Profile settings only when an Opus model is selected.
Also removes deprecated 'none' and 'ultrathink' thinking levels from CLI choices
and all mapping code, treating them as invalid with a fallback to 'medium'.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: propagate fast_mode to ideation and add MODEL_ID_MAP sync comments

Thread fast_mode parameter through IdeationGenerator, IdeationConfigManager,
and IdeationOrchestrator so ideation agents benefit from Fast Mode when enabled.
Add --fast-mode CLI flag to ideation_runner and pass it from the frontend.
Add sync comments to MODEL_ID_MAP in both backend and frontend to prevent drift.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: propagate fast_mode to PR review agents

Add fast_mode field to GitHubRunnerConfig and pass it through to all
create_client() calls in parallel_orchestrator_reviewer and
parallel_followup_reviewer. Add --fast-mode CLI flag to GitHub runner.
Frontend buildRunnerArgs() now accepts fastMode option, passed from
PR review and follow-up review handlers via readSettingsFile().
Also fix leftover 'none' in GitHub runner thinking-level choices.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: clean up stale None types and comments after removing 'none' thinking level

- get_phase_config() return type: tuple[str, str, int | None] → tuple[str, str, int]
- THINKING_BUDGET_MAP type: Record<string, number | null> → Record<string, number>
- Remove '(null = no extended thinking)' comment from THINKING_BUDGET_MAP
- Remove dead None check and stale comment in insights_runner.py

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: correct stale frontend path in phase_config.py sync comments

Update MODEL_ID_MAP and THINKING_BUDGET_MAP cross-reference comments
from auto-claude-ui/src/... to apps/frontend/src/... to match the
actual monorepo path and the frontend's reciprocal comment.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: add missing fast_mode and betas params to remaining GitHub engines

- Add fast_mode=self.config.fast_mode to all 3 create_client() calls in
  pr_review_engine.py (run_review_pass, _run_structural_pass, _run_ai_triage_pass)
- Add fast_mode=self.config.fast_mode to triage_engine.py create_client() call
- Add betas and fast_mode params to review_tools.py spawn functions
  (spawn_security_review, spawn_quality_review, spawn_deep_analysis)
- Remove stale comment in insights_runner.py

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: add betas, fast_mode, and effort_level to spec pipeline agent_runner

Update create_client() call in AgentRunner.run_agent() to use
get_model_betas(), get_fast_mode(), and get_thinking_kwargs_for_model()
matching the pattern in coder.py, planner.py, and qa/loop.py. Add
thinking_level parameter to run_agent() signature and pass from orchestrator.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: sort imports in agent_runner.py to satisfy ruff I001

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: format multi-line import to satisfy ruff I001

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: wrap long line to satisfy ruff format

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: add fast_mode to GitLab MR engine and serialize in GitHub to_dict()

- Add fast_mode field to GitLabRunnerConfig and its to_dict()
- Add betas and fast_mode params to GitLab mr_review_engine create_client()
- Add fast_mode to GitHubRunnerConfig.to_dict() for settings persistence

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-02-09 10:33:45 +01:00

531 lines
22 KiB
Python

"""
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_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",
},
# ═══════════════════════════════════════════════════════════════════════
# 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")