🌐(i18n) update translated strings

Update translated files with new translations

.github/workflows/docker-hub.yml

@@ -47,6 +47,7 @@ jobs:
           docker-image-name: 'docker.io/lasuite/conversations-backend:${{ github.sha }}'
       -
         name: Build and push
+        if: always()
         uses: docker/build-push-action@v6
         with:
           context: .
@@ -86,6 +87,7 @@ jobs:
           docker-image-name: 'docker.io/lasuite/conversations-frontend:${{ github.sha }}'
       -
         name: Build and push
+        if: always()
         uses: docker/build-push-action@v6
         with:
           context: .

CHANGELOG.md

@@ -8,6 +8,66 @@ and this project adheres to
 
 ## [Unreleased]
 
+## [0.0.8] - 2025-11-10
+
+### Fixed
+
+- 🦺(front) Fix send prohibited file types
+- 🐛(front) fix target blank links in chat #103
+- 🚑️(posthog) pass str instead of UUID for user PK #134
+- ⚡️(web-search) keep running when tool call fails #137
+- ✨(summarize): new summarize tool integration #78
+
+### Removed
+
+- 🔥(posthog) remove posthog middleware for async mode fix #146
+
+
+## [0.0.7] - 2025-10-28
+
+### Fixed
+
+- 🚑️(posthog) fix the posthog middleware for async mode #133
+
+
+## [0.0.6] - 2025-10-28
+
+### Fixed
+
+- 🚑️(stats) fix tracking id in upload event #130
+
+
+## [0.0.5] - 2025-10-27
+
+### Fixed
+
+- 🚑️(drag-drop) fix the rejection display on Safari #127
+
+
+## [0.0.4] - 2025-10-27
+
+### Added
+
+- ♿️(a11y) improve accessibility #135
+- 🌐(i18n) add dutch language #117
+
+### Changed
+
+- ⚡️(asgi) use `uvicorn` to serve backend #121
+
+### Fixed
+
+- 🐛(front) fix mobile source
+- 🐛(attachments) reject the whole drag&drop if unsupported formats #123
+
+
+## [0.0.3] - 2025-10-21
+
+### Fixed
+
+- 🚑️(web-search) fix missing argument in RAG backend #116
+
+
 ## [0.0.2] - 2025-10-21
 
 ### Added
@@ -82,6 +142,12 @@ and this project adheres to
 - 💄(chat) add code highlighting for LLM responses #67
 
 
-[unreleased]: https://github.com/suitenumerique/conversations/compare/v0.0.2...main
+[unreleased]: https://github.com/suitenumerique/conversations/compare/v0.0.8...main
+[0.0.8]: https://github.com/suitenumerique/conversations/releases/v0.0.8
+[0.0.7]: https://github.com/suitenumerique/conversations/releases/v0.0.7
+[0.0.6]: https://github.com/suitenumerique/conversations/releases/v0.0.6
+[0.0.5]: https://github.com/suitenumerique/conversations/releases/v0.0.5
+[0.0.4]: https://github.com/suitenumerique/conversations/releases/v0.0.4
+[0.0.3]: https://github.com/suitenumerique/conversations/releases/v0.0.3
 [0.0.2]: https://github.com/suitenumerique/conversations/releases/v0.0.2
 [0.0.1]: https://github.com/suitenumerique/conversations/releases/v0.0.1

Dockerfile

@@ -144,7 +144,7 @@ RUN rm -rf /var/cache/apk/*
 
 ARG CONVERSATIONS_STATIC_ROOT=/data/static
 
-# Gunicorn
+# Gunicorn - not used by default but configuration file is provided
 RUN mkdir -p /usr/local/etc/gunicorn
 COPY docker/files/usr/local/etc/gunicorn/conversations.py /usr/local/etc/gunicorn/conversations.py
 
@@ -158,5 +158,18 @@ COPY --from=link-collector ${CONVERSATIONS_STATIC_ROOT} ${CONVERSATIONS_STATIC_R
 # Copy conversations mails
 COPY --from=mail-builder /mail/backend/core/templates/mail /app/core/templates/mail
 
-# The default command runs gunicorn WSGI server in conversations's main module
-CMD ["gunicorn", "-c", "/usr/local/etc/gunicorn/conversations.py", "conversations.wsgi:application"]
+# The default command runs uvicorn ASGI server in conversations's main module
+# WEB_CONCURRENCY: number of workers to run <=> --workers=4
+ENV WEB_CONCURRENCY=4
+CMD [\
+    "uvicorn",\
+    "--app-dir=/app",\
+    "--host=0.0.0.0",\
+    "--timeout-graceful-shutdown=300",\
+    "--limit-max-requests=20000",\
+    "--lifespan=off",\
+    "conversations.asgi:application"\
+]
+
+# To run using gunicorn WSGI server use this instead:
+#CMD ["gunicorn", "-c", "/usr/local/etc/gunicorn/conversations.py", "conversations.wsgi:application"]

README.md

@@ -115,6 +115,31 @@ To start all the services, except the frontend container, you can use the follow
 $ make run-backend
 ```
 
+**Setup a basic LLM call**
+
+To be able to use Conversations, you need to configure at least one Large Language Model (LLM) provider.
+You can do so by setting the appropriate environment variables in the `env.d/development/common` file:
+
+```ini
+AI_BASE_URL=http://host.docker.internal:12434/v1/
+AI_MODEL=gemma3:4b
+AI_API_KEY=XXX
+```
+
+for a local ollama, or by running a local LLM with docker-compose:
+
+```shellscript
+$ make create-compose-with-models
+```
+
+which will create a `compose.override.yml` file to start a local models `ai/smollm2`
+which can be changed later by editing the `compose.override.yml` file.
+
+You will need to call `make run` after changing the `env.d/development/common`
+or `compose.override.yml` file.
+
+You can find more information about configuring LLM providers in the [LLM Configuration](docs/llm-configuration.md) documentation.
+
 **Adding content**
 
 You can create a basic demo site by running this command:
@@ -141,6 +166,18 @@ You first need to create a superuser account:
 $ make superuser
 ```
 
+## Documentation 📚
+
+Additional documentation is available in the `docs/` directory:
+
+- [LLM Configuration](docs/llm-configuration.md) - Configure Large Language Models and providers
+- [Attachments](docs/attachments.md) - How to use attachments in conversations
+- [Tools for Agents](docs/tools.md) - Available tools and how to add new ones
+- [Environment Variables](docs/env.md) - All available environment variables
+- [Installation Guide](docs/installation.md) - Deploy on a Kubernetes cluster
+- [Theming](docs/theming.md) - Customize the application appearance
+- [Architecture](docs/architecture.md) - Technical architecture overview
+
 ## Licence 📝
 
 This work is released under the MIT License (see [LICENSE](https://github.com/suitenumerique/conversations/blob/main/LICENSE)).

docs/architecture.md

@@ -7,8 +7,8 @@ flowchart TD
     User -- HTTP --> Front("Frontend (NextJS SPA)")
     Front -- REST API --> Back("Backend (Django)")
     Front -- OIDC --> Back -- OIDC ---> OIDC("Keycloak / ProConnect")
-    Back -- REST API --> Yserver
     Back --> DB("Database (PostgreSQL)")
-    Back <--> Celery --> DB
+    Back --> Cache("Cache (Redis)")
     Back ----> S3("Minio (S3)")
+    Back -- REST API --> LLM("LLM Providers")
 ```

docs/attachments.md

@@ -0,0 +1,400 @@
+# Conversation Attachments
+
+This document describes how conversation attachments work in the Conversations application, including the upload process, security measures, and how documents are processed for use with Large Language Models (LLMs).
+
+## Table of Contents
+
+- [Overview](#overview)
+- [Supported Attachment Types](#supported-attachment-types)
+- [Architecture & Flow](#architecture--flow)
+  - [High-Level Overview](#high-level-overview)
+  - [Detailed Technical Flow](#detailed-technical-flow)
+- [Security & Validation](#security--validation)
+  - [MIME Type Validation](#mime-type-validation)
+  - [Malware Detection](#malware-detection)
+- [Document Processing for LLMs](#document-processing-for-llms)
+  - [Image Attachments](#image-attachments)
+  - [PDF Documents](#pdf-documents)
+  - [Other Document Types](#other-document-types)
+- [Configuration](#configuration)
+
+---
+
+## Overview
+
+Conversations allows users to attach files to their conversations with the AI assistant. These attachments can be:
+- **Images** (displayed directly to vision-capable LLMs)
+- **PDF documents** (sent as document URLs to the LLM)
+- **Other documents** (converted to text and indexed for semantic search)
+
+The attachment system uses **S3-compatible object storage** (such as MinIO in development) to store files securely. 
+The backend generates **presigned URLs** that allow the frontend to upload files directly to the storage, 
+without routing the file data through the backend server.
+
+Note about documents: The system uses a tool called **MarkItDown** to convert various document formats 
+(Word, Excel, PowerPoint, text files, etc.) into Markdown text for processing by LLMs. When at least 
+one non-PDF/image document is attached, the system enables:
+ - a **Retrieval-Augmented Generation (RAG)** search tool to allow the LLM to query relevant sections of the documents.
+ - a **summarization tool** to provide document summaries on user request.
+   ⚠️ naive implementation at the moment, needs improvement before being used in production.
+
+## Supported Attachment Types
+The following attachment types are supported:
+- **Images**: `image/png`, `image/jpeg`, `image/gif`, `image/webp`.
+- **PDF documents**: `application/pdf`
+- **Other documents**:
+  - Microsoft Word: `application/vnd.openxmlformats-officedocument.wordprocessingml.document`
+  - Microsoft Excel: `application/vnd.openxmlformats-officedocument.spreadsheetml.sheet`
+  - Microsoft PowerPoint: `application/vnd.openxmlformats-officedocument.presentationml.presentation`
+  - Text files: `text/plain`, `text/markdown`, `text/csv`
+
+**Warning**: The current implementation for PDF expects the LLM to be able to manage them. We need to
+improve the handling of PDFs in case the LLM cannot process them natively.
+
+**Todo**: 
+ - Add support for more file types and improve document processing workflows.
+ - Allow PDF management via RAG search when the LLM cannot handle them natively.
+ - Allow file type restrictions based on model settings, instead of globally.
+ - Improve the summarization tool to provide better summaries and handle larger documents.
+ - Start file upload right away when the user selects a file, instead of waiting for the user to send the message.
+
+
+---
+
+## Architecture & Flow
+
+### High-Level Overview
+
+```
+┌─────────────┐       ┌─────────────┐       ┌─────────────┐       ┌─────────────┐
+│   Frontend  │       │   Backend   │       │  S3 Storage │       │ Malware Det.│
+└──────┬──────┘       └──────┬──────┘       └──────┬──────┘       └──────┬──────┘
+       │                     │                     │                     │
+       │ 1. Create attachment│                     │                     │
+       ├────────────────────>│                     │                     │
+       │                     │                     │                     │
+       │ 2. Return presigned │                     │                     │
+       │    URL for upload   │                     │                     │
+       │<────────────────────┤                     │                     │
+       │                     │                     │                     │
+       │ 3. Upload file      │                     │                     │
+       │    directly to S3   │                     │                     │
+       ├──────────────────────────────────────────>│                     │
+       │                     │                     │                     │
+       │ 4. Notify upload    │                     │                     │
+       │    completed        │                     │                     │
+       ├────────────────────>│                     │                     │
+       │                     │                     │                     │
+       │                     │ 5. Detect MIME type │                     │
+       │                     ├────────────────────>│                     │
+       │                     │                     │                     │
+       │                     │ 6. Scan for malware │                     │
+       │                     ├──────────────────────────────────────────>│
+       │                     │                     │                     │
+       │                     │ 7. Update status    │                     │
+       │ 8. Return status    │<──────────────────────────────────────────┤
+       │<────────────────────┤                     │                     │
+       │                     │                     │                     │
+```
+
+### Detailed Technical Flow
+
+#### Step 1: Attachment Creation Request
+
+When a user selects a file to upload, the frontend sends a POST request to create an attachment record:
+
+**Endpoint**: `POST /api/conversations/{conversation_id}/attachments/`
+
+**Request payload**:
+```json
+{
+  "file_name": "document.pdf",
+  "size": 1048576,
+  "content_type": "application/pdf"
+}
+```
+
+**Backend processing** (`ChatConversationAttachmentViewSet.perform_create`):
+1. Verifies the user owns the conversation
+2. Generates a unique UUID for the file
+3. Creates a storage key: `{conversation_id}/attachments/{uuid}.{extension}`
+4. Creates a database record with status `PENDING`
+
+**Response**:
+```json
+{
+  "id": "uuid-of-attachment",
+  "key": "conversation-id/attachments/file-id.pdf",
+  "file_name": "document.pdf",
+  "size": 1048576,
+  "upload_state": "pending",
+  "policy": "https://s3.example.com/bucket/...?presigned-params"
+}
+```
+
+The `policy` field contains a **presigned URL** valid for a limited time (configured by `AWS_S3_UPLOAD_POLICY_EXPIRATION`).
+
+#### Step 2: Direct Upload to S3
+
+The frontend uses the presigned URL to upload the file directly to S3 storage using a PUT request.
+
+**Technical details**:
+- The presigned URL includes authentication parameters
+- The upload is done with `Content-Type` header matching the file's MIME type
+- No backend involvement in the data transfer
+
+#### Step 3: Upload Completion Notification
+
+After successful upload, the frontend notifies the backend:
+
+**Endpoint**: `POST /api/conversations/{conversation_id}/attachments/{attachment_id}/upload-ended/`
+
+**Backend processing** (`ChatConversationAttachmentViewSet.upload_ended`):
+
+1. **MIME Type Detection** (`chat/views.py`):
+   ```python
+   mime_detector = magic.Magic(mime=True)
+   with default_storage.open(attachment.key, "rb") as file:
+       mimetype = mime_detector.from_buffer(file.read(2048))
+       size = file.size
+   ```
+   
+   Uses `python-magic` to detect the actual MIME type from file content (first 2048 bytes).
+
+2. **Update attachment status**:
+   - Status: `PENDING` → `ANALYZING`
+   - Store detected MIME type and actual file size
+
+3. **Trigger Malware Detection**:
+   ```python
+   malware_detection.analyse_file(
+       attachment.key,
+       safe_callback="chat.malware_detection.conversation_safe_attachment_callback",
+       unknown_callback="chat.malware_detection.unknown_attachment_callback",
+       unsafe_callback="chat.malware_detection.conversation_unsafe_attachment_callback",
+       conversation_id=conversation_id,
+   )
+   ```
+
+#### Step 4: Malware Detection Callbacks
+
+The malware detection service (configurable via `MALWARE_DETECTION_BACKEND`) scans the file and calls one of three callbacks:
+
+**Safe file** (`conversation_safe_attachment_callback`):
+- Status: `ANALYZING` → `READY`
+- File is ready for use
+
+**Unsafe file** (`conversation_unsafe_attachment_callback`):
+- Status: `ANALYZING` → `SUSPICIOUS`
+- File is quarantined and not accessible
+- Security log entry created
+
+**Unknown status** (`unknown_attachment_callback`):
+- Handles special cases (e.g., file too large to analyze)
+- Status: `ANALYZING` → `FILE_TOO_LARGE_TO_ANALYZE`
+
+---
+
+## Security & Validation
+
+For now, the system is not intended to host user-uploaded files for public download.
+All files are stored in private S3 buckets with presigned URLs for controlled access and only
+the owner of the conversation/the uploader can access them, so the risk is quite low around bad use of
+the attachment system.
+
+Also, the document content is sent to the LLM and does not prevent any prompt injection attacks, which is not
+an issue specific to the attachment system but to the overall design of LLM-based applications and should be
+addressed globally. Also for the moment, the system does not have any action tools that could be used to execute
+malicious code based on document content.
+
+### Malware Detection
+
+The malware detection system is **pluggable** and configurable, allowing different backends to be used.
+By default, a `DummyBackend` is provided that marks all files as safe.
+
+⚠️ The current implementation does not disallow any file types or status from being used in conversations.
+This is a potential security risk and should be addressed in future versions.
+
+---
+
+## Document Processing for LLMs
+
+When a user sends a message with attachments, the system processes them differently based on their type:
+
+### Image Attachments
+
+**MIME types**: `image/png`, `image/jpeg`, `image/gif`, `image/webp`, etc.
+
+**Processing flow**:
+
+1. **URL Conversion**: Local media URLs are converted to presigned S3 URLs before sending to the LLM:
+   ```python
+   # From: chat/agents/local_media_url_processors.py
+   content.url = generate_retrieve_policy(key)
+   ```
+
+2. **Sent to LLM**: Images are sent as `ImageUrl` objects in the prompt:
+   ```python
+   ImageUrl(
+       url="https://s3.example.com/bucket/key?presigned-params",
+       identifier="file-id.png",
+   )
+   ```
+
+3. **Vision models** can analyze the image content directly.
+
+4. **Response processing**: After the LLM responds, presigned URLs are converted back to local URLs for storage:
+   ```python
+   # Mapping: presigned_url -> /media-key/{conversation_id}/attachments/{file_id}.png
+   ```
+
+### PDF Documents
+
+**MIME type**: `application/pdf`
+
+**Processing flow**:
+
+1. **Direct URL passing**: PDFs are sent as `DocumentUrl` objects :
+   ```python
+   DocumentUrl(
+       url="https://s3.example.com/bucket/key?presigned-params",
+       identifier="file-id.pdf",
+   )
+   ```
+
+2. **LLM processing**: Compatible LLMs can:
+   - Extract and read text from PDFs
+   - Understand document structure
+   - Answer questions about the content
+
+3. **No conversion needed**: PDFs are passed directly without preprocessing.
+
+### Other Document Types
+
+**MIME types**: Word documents, Excel spreadsheets, PowerPoint, text files, Markdown, etc.
+
+**Processing flow**:
+
+1. **Document parsing**: When a document is uploaded, it's parsed using the `AlbertRagBackend` class.
+
+2. **Conversion to Markdown**: Documents are converted using **MarkItDown** library or using the "Albert API" for PDFs.
+
+3. **RAG (Retrieval-Augmented Generation)**:
+   - Converted text is indexed in a vector database
+   - The LLM uses a `document_rag_search` tool to query relevant sections
+   - Only relevant chunks are sent to the LLM to fit context windows
+
+4. **Summarization tool** if needed.
+
+### Processing Strategy Decision Tree
+
+**Decision logic**:
+- **No documents**: Standard conversation
+- **Images**: Send as direct (presigned) URLs to the LLM
+- **Only PDFs**: Send as direct (presigned) URLs to the LLM
+- **Other documents present**: Enable RAG search tool + convert to Markdown
+
+---
+
+## Configuration
+
+### Environment Variables
+
+| Variable                                     | Default        | Description                                                |
+|----------------------------------------------|----------------|------------------------------------------------------------|
+| `ATTACHMENT_MAX_SIZE`                        | Configurable   | Maximum file size in bytes                                 |
+| `ATTACHMENT_CHECK_UNSAFE_MIME_TYPES_ENABLED` | `True`         | Enable/disable MIME type validation                        |
+| `AWS_S3_UPLOAD_POLICY_EXPIRATION`            | 3600           | Presigned URL expiration (seconds)                         |
+| `AWS_S3_RETRIEVE_POLICY_EXPIRATION`          | 3600           | Presigned retrieval URL expiration (seconds)               |
+| `AWS_S3_DOMAIN_REPLACE`                      | None           | Alternative S3 domain for presigned URLs (for development) |
+| `MALWARE_DETECTION_BACKEND`                  | `DummyBackend` | Malware scanning backend class                             |
+| `MALWARE_DETECTION_PARAMETERS`               | `{}`           | Backend-specific configuration                             |
+| `RAG_FILES_ACCEPTED_FORMATS`                 | See below      | List of MIME types accepted for file uploads               |
+
+#### RAG_FILES_ACCEPTED_FORMATS
+
+This environment variable controls which file types users are allowed to upload as attachments to conversations.
+
+**Configuration**:
+- **Type**: List of strings (comma-separated MIME types when using environment variable)
+- **Default value**: Includes a comprehensive list of document and image formats:
+  - Microsoft Office documents (`.docx`, `.pptx`, `.xlsx`, `.xls`)
+  - Text files (`.txt`, `.csv`)
+  - PDF documents (`.pdf`)
+  - HTML files
+  - Markdown files (`.md`)
+  - Outlook messages (`.msg`)
+  - Images (`.jpeg`, `.png`, `.gif`, `.webp`)
+
+**Example configuration**:
+```ini
+# In environment variable (comma-separated)
+RAG_FILES_ACCEPTED_FORMATS="application/pdf,text/plain,image/png,image/jpeg"
+```
+
+```python
+# In Django settings (as a Python list)
+RAG_FILES_ACCEPTED_FORMATS = [
+    "application/pdf",
+    "text/plain",
+    "image/png",
+    "image/jpeg",
+]
+```
+
+**How it's used**:
+1. **Backend**: The list is exposed via the `/api/v1.0/config/` endpoint as `chat_upload_accept` (MIME types joined with commas)
+2. **Frontend**: The configuration is used to validate files before upload in the chat interface:
+   - Checks exact MIME type matches
+   - Supports wildcard patterns (e.g., `image/*` for all image types)
+   - Supports file extension patterns (e.g., `.pdf`)
+3. **User experience**: Files that don't match the accepted formats are rejected with a user-friendly error message
+
+**Notes**:
+
+ - This setting controls frontend validation only. Backend validation should also be implemented for security.
+ - Future improvements may include per-model file type restrictions.
+
+### Storage Configuration
+
+**MinIO (Development)**:
+```yaml
+# docker-compose.yml
+minio:
+  image: minio/minio
+  environment:
+    MINIO_ROOT_USER: minioadmin
+    MINIO_ROOT_PASSWORD: minioadmin
+  command: server /data --console-address ":9001"
+```
+
+---
+
+## Troubleshooting
+
+### LLM Cannot Access Image/PDF
+
+**Possible causes**:
+- Presigned URL has expired
+- S3 storage is not accessible from the LLM provider
+- CORS configuration issues
+
+**Solution**: Check `AWS_S3_RETRIEVE_POLICY_EXPIRATION` and S3 access policies.
+
+### Document Not Appearing in RAG Search
+
+**Possible causes**:
+- Document conversion failed
+- Vector database indexing failed
+
+**Check logs**: Look for errors in `DocumentConverter` and RAG backend logs.
+
+---
+
+## Related Documentation
+
+- [Installation Guide](installation.md) - S3 storage setup
+- [LLM Configuration](llm-configuration.md) - Model capabilities for attachments
+- [Architecture](architecture.md) - System overview
+- [Tools](tools.md) - Document search and RAG tools
+

docs/env.md

@@ -10,7 +10,6 @@ These are the environment variables you can set for the `conversations-backend`
 |-------------------------------------------------|-----------------------------------------------------------------------------------------------------------------------------------|---------------------------------------------------------|
 | DJANGO_ALLOWED_HOSTS                            | allowed hosts                                                                                                                     | []                                                      |
 | DJANGO_SECRET_KEY                               | secret key                                                                                                                        |                                                         |
-| DJANGO_SERVER_TO_SERVER_API_TOKENS              |                                                                                                                                   | []                                                      |
 | DB_ENGINE                                       | engine to use for database connections                                                                                            | django.db.backends.postgresql_psycopg2                  |
 | DB_NAME                                         | name of the database                                                                                                              | conversations                                           |
 | DB_USER                                         | user to authenticate with                                                                                                         | dinum                                                   |
@@ -24,12 +23,11 @@ These are the environment variables you can set for the `conversations-backend`
 | AWS_S3_SECRET_ACCESS_KEY                        | access key for s3 endpoint                                                                                                        |                                                         |
 | AWS_S3_REGION_NAME                              | region name for s3 endpoint                                                                                                       |                                                         |
 | AWS_STORAGE_BUCKET_NAME                         | bucket name for s3 endpoint                                                                                                       | conversations-media-storage                             |
-| ATTACHMENT_MAX_SIZE                         | maximum size of document in bytes                                                                                                 | 10485760                                                |
+| ATTACHMENT_MAX_SIZE                             | maximum size of document in bytes                                                                                                 | 10485760                                                |
 | LANGUAGE_CODE                                   | default language                                                                                                                  | en-us                                                   |
 | API_USERS_LIST_THROTTLE_RATE_SUSTAINED          | throttle rate for api                                                                                                             | 180/hour                                                |
 | API_USERS_LIST_THROTTLE_RATE_BURST              | throttle rate for api on burst                                                                                                    | 30/minute                                               |
 | SPECTACULAR_SETTINGS_ENABLE_DJANGO_DEPLOY_CHECK |                                                                                                                                   | false                                                   |
-| TRASHBIN_CUTOFF_DAYS                            | trashbin cutoff                                                                                                                   | 30                                                      |
 | DJANGO_EMAIL_BACKEND                            | email backend library                                                                                                             | django.core.mail.backends.smtp.EmailBackend             |
 | DJANGO_EMAIL_BRAND_NAME                         | brand name for email                                                                                                              |                                                         |
 | DJANGO_EMAIL_HOST                               | host name of email                                                                                                                |                                                         |
@@ -76,12 +74,14 @@ These are the environment variables you can set for the `conversations-backend`
 | OIDC_USERINFO_FULLNAME_FIELDS                   | OIDC token claims to create full name                                                                                             | ["first_name", "last_name"]                             |
 | OIDC_USERINFO_SHORTNAME_FIELD                   | OIDC token claims to create shortname                                                                                             | first_name                                              |
 | ALLOW_LOGOUT_GET_METHOD                         | Allow get logout method                                                                                                           | true                                                    |
-| AI_API_KEY                                      | AI key to be used for AI Base url                                                                                                 |                                                         |
-| AI_BASE_URL                                     | OpenAI compatible AI base url                                                                                                     |                                                         |
-| AI_MODEL                                        | AI Model to use                                                                                                                   |                                                         |
-| AI_AGENT_INSTRUCTION                            | Base instruction for the AI agent                                                                                                 | You are a helpful assistant                             |
-| Y_PROVIDER_API_KEY                              | Y provider API key                                                                                                                |                                                         |
-| Y_PROVIDER_API_BASE_URL                         | Y Provider url                                                                                                                    |                                                         |
+| LLM_CONFIGURATION_FILE_PATH                     | Path to the LLM configuration JSON file. See [LLM Configuration](llm-configuration.md) for details                                | <BASE_DIR>/conversations/configuration/llm/default.json |
+| LLM_DEFAULT_MODEL_HRID                          | HRID of the model used for conversations                                                                                          | default-model                                           |
+| LLM_SUMMARIZATION_MODEL_HRID                    | HRID of the model used for summarization                                                                                          | default-summarization-model                             |
+| AI_API_KEY                                      | AI API key to be used for the default provider (used in default LLM configuration, not for production use)                        |                                                         |
+| AI_BASE_URL                                     | OpenAI compatible AI base URL (used in default LLM configuration, not for production use)                                         |                                                         |
+| AI_MODEL                                        | AI Model name to use (used in default LLM configuration, not for production use)                                                  |                                                         |
+| AI_AGENT_INSTRUCTIONS                           | Base instruction for the AI agent (used in default LLM configuration, not for production use)                                     | You are a helpful assistant. Wrap formulas...           |
+| AI_AGENT_TOOLS                                  | List of enabled tools for the agent (used in default LLM configuration, not for production use)                                   | []                                                      |
 | CONVERSION_API_ENDPOINT                         | Conversion API endpoint                                                                                                           | convert-markdown                                        |
 | CONVERSION_API_CONTENT_FIELD                    | Conversion api content field                                                                                                      | content                                                 |
 | CONVERSION_API_TIMEOUT                          | Conversion api timeout                                                                                                            | 30                                                      |

docs/examples/conversations.values.yaml

@@ -9,7 +9,6 @@ backend:
     DJANGO_CSRF_TRUSTED_ORIGINS: https://conversations.127.0.0.1.nip.io
     DJANGO_CONFIGURATION: Feature
     DJANGO_ALLOWED_HOSTS: conversations.127.0.0.1.nip.io
-    DJANGO_SERVER_TO_SERVER_API_TOKENS: secret-api-key
     DJANGO_SECRET_KEY: AgoodOrAbadKey
     DJANGO_SETTINGS_MODULE: conversations.settings
     DJANGO_SUPERUSER_PASSWORD: admin

docs/installation.md

@@ -7,7 +7,7 @@ This document is a step-by-step guide that describes how to install Conversation
 - k8s cluster with an nginx-ingress controller
 - an OIDC provider (if you don't have one, we provide an example)
 - a PostgreSQL server (if you don't have one, we provide an example)
-- a Memcached server (if you don't have one, we provide an example)
+- a Redis server (if you don't have one, we provide an example)
 - a S3 bucket (if you don't have one, we provide an example)
 
 ### Test cluster

docs/llm-configuration.md

@@ -0,0 +1,412 @@
+# LLM Configuration
+
+This document describes how to configure Large Language Models (LLMs) in Conversations via the configuration file.
+
+## Overview
+
+Conversations uses a JSON configuration file to define LLM models and providers. This approach allows you to:
+- Configure multiple LLM models from different providers
+- Switch between models without code changes
+- Customize model-specific settings like temperature, max tokens, and system prompts
+- Enable or disable models dynamically
+
+The overall structure consists of two main sections: `providers` and `models`.
+Settings for models, provides customization through `settings` and `profile`, which corresponds to the 
+Pydantic AI model settings and profile. While we currently not use those settings extensively, 
+they are available for future use and advanced configurations, please reach us if you face any problem using them.
+
+## Configuration File Location
+
+The default LLM configuration file is located at:
+```
+src/backend/conversations/configuration/llm/default.json
+```
+
+You can override this location by setting the `LLM_CONFIGURATION_FILE_PATH` environment variable, but be careful as 
+this path must be accessible by the backend application _inside the docker image_:
+``` ini
+LLM_CONFIGURATION_FILE_PATH=/path/to/your/llm/config.json
+```
+
+## Default Behavior
+
+### Default Configuration
+
+The default configuration file is useful for local development and running the test, while it can be used
+in production, we suggest to create a specific one for production and replace the `settings.` values with
+`environ.` one.
+
+The default configuration file (`default.json`) includes:
+
+1. **Two default models**:
+   - `default-model`: The primary conversational model used for chat interactions
+   - `default-summarization-model`: A specialized model for summarizing conversations
+
+2. **One default provider**:
+   - `default-provider`: An OpenAI-compatible provider that uses environment variables for configuration
+
+### Environment Variable Integration
+
+The configuration uses dynamic value resolution with two special prefixes:
+
+- `settings.VARIABLE_NAME`: Resolves to a Django setting value
+- `environ.VARIABLE_NAME`: Resolves to an environment variable value
+
+For example, in the default configuration:
+```json
+{
+  "model_name": "settings.AI_MODEL",
+  "system_prompt": "settings.AI_AGENT_INSTRUCTIONS",
+  "tools": "settings.AI_AGENT_TOOLS"
+}
+```
+
+This allows to configure models in tests using the setting override mechanism from Django/Pytest (but might be replaced
+later with a simple override of the full configuration like it's done in some tests already).
+
+### Required Environment Variables
+
+For the default configuration to work, you need to set these environment variables:
+
+| Variable                      | Description                            | Example                     |
+|-------------------------------|----------------------------------------|-----------------------------|
+| `AI_API_KEY`                  | API key for the default provider       | `sk-...`                    |
+| `AI_BASE_URL`                 | Base URL for the OpenAI-compatible API | `https://api.openai.com/v1` |
+| `AI_MODEL`                    | Model name to use                      | `gpt-4o-mini`               |
+
+### Optional Environment Variables
+
+If you want to customize the agent behavior and tools, you can set these optional environment variables 
+(defaults are provided in the default configuration):
+
+| Variable                      | Description                            | Default           |
+|-------------------------------|----------------------------------------|-------------------|
+| `AI_AGENT_INSTRUCTIONS`       | System prompt for the agent            | see `settings.py` |
+| `AI_AGENT_TOOLS`              | List of enabled tools                  | `[]`              |
+| `SUMMARIZATION_SYSTEM_PROMPT` | Base prompt of the summarization agent | see `settings.py` |
+
+### Model Selection
+
+You can configure which models are used for specific tasks via environment variables:
+
+| Variable                       | Description                              | Default                       |
+|--------------------------------|------------------------------------------|-------------------------------|
+| `LLM_DEFAULT_MODEL_HRID`       | HRID of the model used for conversations | `default-model`               |
+| `LLM_SUMMARIZATION_MODEL_HRID` | HRID of the model used for summarization | `default-summarization-model` |
+
+## Configuration Structure
+
+The configuration file has two main sections:
+
+### 1. Providers
+
+Providers define the API endpoints and authentication for LLM services.
+
+```json
+{
+  "providers": [
+    {
+      "hrid": "unique-provider-id",
+      "base_url": "https://api.example.com/v1",
+      "api_key": "environ.API_KEY_VAR",
+      "kind": "openai"
+    }
+  ]
+}
+```
+
+**Provider Fields:**
+
+| Field      | Type   | Required | Description                                             |
+|------------|--------|----------|---------------------------------------------------------|
+| `hrid`     | string | Yes      | Unique identifier for the provider                      |
+| `base_url` | string | Yes      | API base URL (can use `settings.` or `environ.` prefix) |
+| `api_key`  | string | Yes      | API authentication key (use `environ.` here)            |
+| `kind`     | string | Yes      | Provider type: `openai` or `mistral`                    |
+
+### 2. Models
+
+Models define the LLMs available in your application.
+
+```json
+{
+  "models": [
+    {
+      "hrid": "unique-model-id",
+      "model_name": "gpt-4o-mini",
+      "human_readable_name": "GPT-4o Mini",
+      "provider_name": "unique-provider-id",
+      "profile": null,
+      "settings": {},
+      "is_active": true,
+      "icon": null,
+      "system_prompt": "You are a helpful assistant",
+      "tools": []
+    }
+  ]
+}
+```
+
+**Model Fields:**
+
+| Field                 | Type         | Required | Description                                                                                         |
+|-----------------------|--------------|----------|-----------------------------------------------------------------------------------------------------|
+| `hrid`                | string       | Yes      | Unique identifier for the model                                                                     |
+| `model_name`          | string       | Yes      | Name of the model as recognized by the provider (can use `settings.` or `environ.` prefix)          |
+| `human_readable_name` | string       | Yes      | Display name shown to users                                                                         |
+| `provider_name`       | string       | No*      | Reference to a provider's `hrid`                                                                    |
+| `provider`            | object       | No*      | Inline provider definition (alternative to `provider_name`)                                         |
+| `profile`             | object       | No       | Model-specific capabilities and settings                                                            |
+| `settings`            | object       | No       | Model inference settings (temperature, max_tokens, etc.)                                            |
+| `is_active`           | boolean      | Yes      | Whether the model is available for use                                                              |
+| `icon`                | string/array | No       | Base64-encoded icon or array of icon parts                                                          |
+| `system_prompt`       | string       | Yes      | Default system prompt for the model (can use `settings.` or `environ.` prefix)                      |
+| `tools`               | array        | Yes      | List of enabled tools for this model (can use `settings.` or `environ.` prefix for the whole array) |
+| `supports_streaming`  | boolean      | No       | Whether the model supports streaming responses                                                      |
+
+\* Either `provider_name` or `provider` must be set, unless `model_name` is in the format `<provider>:<model>`.
+
+## Adding New Models
+
+### Example 1: Adding a New OpenAI Model
+
+To add a new OpenAI model using the existing default provider:
+
+```json
+{
+  "models": [
+    // ...existing models...
+    {
+      "hrid": "gpt-4-turbo",
+      "model_name": "gpt-4-turbo-preview",
+      "human_readable_name": "GPT-4 Turbo",
+      "provider_name": "default-provider",
+      "profile": null,
+      "settings": {
+        "temperature": 0.7,
+        "max_tokens": 4096
+      },
+      "is_active": true,
+      "icon": null,
+      "system_prompt": "You are an expert AI assistant.",
+      "tools": ["web_search_brave_with_document_backend"],
+      "supports_streaming": true
+    }
+  ],
+  "providers": [
+    // ...existing providers...
+  ]
+}
+```
+
+### Example 2: Adding a Model using Pydantic AI format
+
+To add a model with a specific provider using the default Pydantic AI format, you don't need to define the provider separately if you use the `model_name` format `<provider>:<model>`.
+
+1. **Add the model without provider**:
+
+```json
+{
+  "models": [
+    {
+      "hrid": "claude-3-opus",
+      "model_name": "anthropic:claude-3-opus-20240229",
+      "human_readable_name": "Claude 3 Opus",
+      "provider_name": null,
+      "profile": null,
+      "settings": {
+        "temperature": 0.7,
+        "max_tokens": 4096
+      },
+      "is_active": true,
+      "icon": null,
+      "system_prompt": "You are Claude, a helpful AI assistant.",
+      "tools": []
+    }
+  ],
+  "providers": []
+}
+```
+
+2**Set the environment variable**:
+
+Pydantic AI expects the API key in an environment variable named `ANTHROPIC_API_KEY` is this example, so set it accordingly:
+
+```ini
+ANTHROPIC_API_KEY=your-api-key-here
+```
+
+### Example 3: Adding a Mistral Model
+
+For Mistral AI models using the Etalab platform:
+
+```json
+{
+  "models": [
+    {
+      "hrid": "mistral-large",
+      "model_name": "mistral-large-latest",
+      "human_readable_name": "Mistral Large (Etalab)",
+      "provider_name": "mistral-etalab",
+      "profile": null,
+      "settings": {
+        "temperature": 0.5,
+        "max_tokens": 8192
+      },
+      "is_active": true,
+      "icon": null,
+      "system_prompt": "settings.AI_AGENT_INSTRUCTIONS",
+      "tools": ["web_search_brave_with_document_backend"]
+    }
+  ],
+  "providers": [
+    {
+      "hrid": "mistral-etalab",
+      "base_url": "https://api.mistral.etalab.gouv.fr/",
+      "api_key": "environ.MISTRAL_ETALAB_API_KEY",
+      "kind": "mistral"
+    }
+  ]
+}
+```
+
+### Example 4: Using Inline Provider Definition
+
+Instead of referencing a provider by name, you can define it inline if you use a unique configuration:
+
+```json
+{
+  "models": [
+    {
+      "hrid": "custom-model",
+      "model_name": "custom-model-v1",
+      "human_readable_name": "Custom Model",
+      "provider": {
+        "hrid": "custom-provider-inline",
+        "base_url": "https://custom-api.example.com/v1",
+        "api_key": "environ.CUSTOM_API_KEY",
+        "kind": "openai"
+      },
+      "settings": {},
+      "is_active": true,
+      "icon": null,
+      "system_prompt": "You are a custom assistant.",
+      "tools": []
+    }
+  ]
+}
+```
+
+## Advanced Configuration
+
+### Model Settings
+
+The `settings` object supports various inference parameters:
+
+```json
+{
+  "settings": {
+    "max_tokens": 4096,
+    "temperature": 0.7,
+    "top_p": 0.9,
+    "timeout": 60.0,
+    "parallel_tool_calls": true,
+    "seed": 42,
+    "presence_penalty": 0.0,
+    "frequency_penalty": 0.0,
+    "logit_bias": {},
+    "stop_sequences": [],
+    "extra_headers": {},
+    "extra_body": {}
+  }
+}
+```
+
+### Model Profile
+
+The `profile` object defines model capabilities:
+
+```json
+{
+  "profile": {
+    "supports_tools": true,
+    "supports_json_schema_output": true,
+    "supports_json_object_output": true,
+    "default_structured_output_mode": "json_schema",
+    "thinking_tags": ["<thinking>", "</thinking>"],
+    "ignore_streamed_leading_whitespace": true
+  }
+}
+```
+
+### Available Tools
+
+Tools can be specified in the `tools` array. Common tools include:
+- `web_search_brave_with_document_backend`: Web search using Brave API with document processing
+
+You can also reference the tools list from Django settings:
+```json
+{
+  "tools": "settings.AI_AGENT_TOOLS"
+}
+```
+
+### Custom Icons
+
+Icons can be provided as base64-encoded PNG images. For long strings, you can split them into an array:
+
+```json
+{
+  "icon": [
+    "data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAABwAAAAcCAMAAABF0y+m",
+    "AAAAn1BMVEUALosAKoovTZjw8vb////+9/jlPUniAAziABUAGIWbpsTwq7HhAAAA"
+  ]
+}
+```
+
+## Validation
+
+The configuration is validated when loaded. Common validation errors include:
+
+- **Provider not found**: A model references a `provider_name` that doesn't exist in the `providers` array
+- **Missing provider**: Neither `provider_name` nor `provider` is specified, and `model_name` is not in `<provider>:<model>` format
+- **Environment variable not set**: A value using `environ.` prefix references an undefined environment variable
+- **Django setting not set**: A value using `settings.` prefix references an undefined Django setting
+- **Invalid provider kind**: The `kind` field must be either `openai` or `mistral`
+
+## Testing Your Configuration
+
+After modifying the configuration file, you can test it by:
+
+1. **Checking for syntax errors**:
+   ```bash
+   python -m json.tool src/backend/conversations/configuration/llm/default.json
+   ```
+
+2. **Starting the application** and checking the logs for validation errors
+
+3. **Using the Django shell** to load the configuration:
+   ```bash
+   ./bin/manage shell
+   ```
+   ```python
+   from django.conf import settings
+   models = settings.LLM_CONFIGURATIONS
+   models.keys()  # Should show all model HRIDs
+   ```
+
+## Best Practices
+
+1. **Use environment variables** for sensitive data like API keys (with `environ.` prefix)
+2. **Use Django settings** for configurable values that may change between environments (with `settings.` prefix)
+3. **Keep provider definitions separate** from models to avoid duplication when using multiple models from the same provider
+4. **Set `is_active: false`** for models you want to keep in the configuration but temporarily disable
+5. **Use descriptive `hrid` values** that clearly identify the model and provider
+6. **Document custom configurations** in your deployment documentation
+7. **Test configuration changes** in a development environment before deploying to production
+
+## See Also
+
+- [Environment Variables Documentation](env.md) - For configuring environment variables
+- [Installation Guide](installation.md) - For deployment instructions
+

docs/system-requirements.md

@@ -14,15 +14,15 @@ Memory is the first bottleneck; CPU matters only when Celery or the Next.js buil
 
 ## 2. Development Environment Memory Requirements
 
-| Service               | Typical use                   | Rationale / source                                                                      |
-|-----------------------|-------------------------------|-----------------------------------------------------------------------------------------|
-| PostgreSQL            | **1 – 2 GB**                  | `shared_buffers` starting point ≈ 25% RAM ([postgresql.org][1])                         |
-| Keycloak              | **≈ 1.3 GB**                  | 70% of limit for heap + ~300 MB non-heap ([keycloak.org][2])                            |
-| Redis                 | **≤ 256 MB**                  | Empty instance ≈ 3 MB; budget 256 MB to allow small datasets ([stackoverflow.com][3])   |
-| MinIO                 | **2 GB (dev) / 32 GB (prod)** | Pre-allocates 1–2 GiB; docs recommend 32 GB per host for ≤ 100 Ti storage ([min.io][4]) |
-| Django API (+ Celery) | **0.8 – 1.5 GB**              | Empirical in-house metrics                                                              |
-| Next.js frontend      | **0.5 – 1 GB**                | Dev build chain                                                                         |
-| Nginx                 | **< 100 MB**                  | Static reverse-proxy footprint                                                          |
+| Service          | Typical use                   | Rationale / source                                                                      |
+|------------------|-------------------------------|-----------------------------------------------------------------------------------------|
+| PostgreSQL       | **1 – 2 GB**                  | `shared_buffers` starting point ≈ 25% RAM ([postgresql.org][1])                         |
+| Keycloak         | **≈ 1.3 GB**                  | 70% of limit for heap + ~300 MB non-heap ([keycloak.org][2])                            |
+| Redis            | **≤ 256 MB**                  | Empty instance ≈ 3 MB; budget 256 MB to allow small datasets ([stackoverflow.com][3])   |
+| MinIO            | **2 GB (dev) / 32 GB (prod)** | Pre-allocates 1–2 GiB; docs recommend 32 GB per host for ≤ 100 Ti storage ([min.io][4]) |
+| Django API       | **0.8 – 1.5 GB**              | Empirical in-house metrics                                                              |
+| Next.js frontend | **0.5 – 1 GB**                | Dev build chain                                                                         |
+| Nginx            | **< 100 MB**                  | Static reverse-proxy footprint                                                          |
 
 [1]: https://www.postgresql.org/docs/9.1/runtime-config-resource.html "PostgreSQL: Documentation: 9.1: Resource Consumption"
 [2]: https://www.keycloak.org/high-availability/concepts-memory-and-cpu-sizing "Concepts for sizing CPU and memory resources - Keycloak"
@@ -58,7 +58,7 @@ Production deployments differ significantly from development environments. The t
 | Service                          | Memory     | Notes                                  |
 |----------------------------------|------------|----------------------------------------|
 | PostgreSQL                       | **2 GB**   | Core database                          |
-| Django API (+ Celery)            | **1.5 GB** | Backend services                       |
+| Django API                       | **1.5 GB** | Backend services                       |
 | Nginx                            | **100 MB** | Static files + reverse proxy           |
 | Redis                            | **256 MB** | Session storage                        |
 | **Total (without auth/storage)** | **≈ 4 GB** | External OIDC + object storage assumed |
@@ -81,16 +81,16 @@ Production deployments differ significantly from development environments. The t
 
 ## 5. Ports (dev defaults)
 
-| Port      | Service               |
-|-----------|-----------------------|
-| 3000      | Next.js               |
-| 8071      | Django                |
-| 8080      | Keycloak              |
-| 8083      | Nginx proxy           |
-| 9000/9001 | MinIO                 |
-| 15432     | PostgreSQL (main)     |
-| 5433      | PostgreSQL (Keycloak) |
-| 1081      | Maildev               |
+| Port      | Service                    |
+|-----------|----------------------------|
+| 3000      | Next.js                    |
+| 8071      | Django                     |
+| 8080      | Keycloak                   |
+| 8083      | Nginx proxy                |
+| 9000/9001 | MinIO                      |
+| 15432     | PostgreSQL (main)          |
+| 5433      | PostgreSQL (Keycloak)      |
+| 1081      | Maildev (currently unused) |
 
 ## 6. Sizing Guidelines
 

docs/theming.md

@@ -4,7 +4,7 @@
 
 To use this feature, simply set the `FRONTEND_CSS_URL` environment variable to the URL of your custom CSS file. For example:
 
-```javascript
+```ini
 FRONTEND_CSS_URL=http://anything/custom-style.css
 ```
 
@@ -38,7 +38,7 @@ The footer is configurable from the theme customization file.
 
 ### Settings 🔧
 
-```shellscript
+```ini
 THEME_CUSTOMIZATION_FILE_PATH=<path>
 ```
 
@@ -55,10 +55,10 @@ The translations can be partially overridden from the theme customization file.
 
 ### Settings 🔧
 
-```shellscript
+```ini
 THEME_CUSTOMIZATION_FILE_PATH=<path>
 ```
 
 ### Example of JSON
 
-The json must follow some rules: https://github.com/suitenumerique/conversations/blob/main/src/helm/env.d/dev/configuration/theme/demo.json
+The json must follow some rules: https://github.com/suitenumerique/conversations/blob/main/src/helm/env.d/dev/configuration/theme/demo.json

docs/tools.md

@@ -0,0 +1,238 @@
+# Tools for the Conversation Agent
+
+The conversation agent can be extended with various tools that provide additional capabilities such as web search, 
+weather information, and more. We currently only have web search tools, but more tools can be added as needed.
+This document explains how to configure and use these tools.
+
+## Overview
+
+Tools are functions that the LLM can call during a conversation to access external data or perform specific actions. 
+The agent decides when to use these tools based on the user's query and the conversation context.
+
+## Configuring Tools for a Model
+
+Tools are configured at the model level in the LLM configuration file. 
+Each model can have its own set of available tools.
+
+### Configuration File Location
+
+Read the [LLM Configuration](llm-configuration.md) document to find out where the configuration file is located
+and how to use it.
+
+### Example Configuration
+
+```json
+{
+  "models": [
+    {
+      "hrid": "default-model",
+      "model_name": "gpt-4",
+      "human_readable_name": "GPT-4 with Tools",
+      "provider_name": "default-provider",
+      "is_active": true,
+      "system_prompt": "You are a helpful assistant.",
+      "tools": [
+        "web_search_brave",
+        "get_current_weather"
+      ]
+    }
+  ],
+  "providers": [
+    {
+      "hrid": "default-provider",
+      "base_url": "https://api.openai.com/v1",
+      "api_key": "settings.AI_API_KEY",
+      "kind": "openai"
+    }
+  ]
+}
+```
+
+The `tools` field accepts either:
+- A list of tool names: `["tool_name_1", "tool_name_2"]`
+- A reference to a settings variable: `"settings.AI_AGENT_TOOLS"`
+
+## Available Tools
+
+To make a tool available to be in a model's configuration, it must be registered in the tool registry located at
+`src/backend/chat/tools/__init__.py`.
+
+This is not dynamic - any changes to the tool registry require a code deployment... 
+We want to add dynamic loading in the future.
+
+| Tool Name                                | Description                                                   | Documentation                                                               |
+|------------------------------------------|---------------------------------------------------------------|-----------------------------------------------------------------------------|
+| `get_current_weather`                    | Fake weather tool for testing purposes                        | [Details](tools/get_current_weather.md)                                     |
+| `web_search_tavily`                      | Web search using Tavily API                                   | [Details](tools/web_search_tavily.md)                                       |
+| `web_search_brave`                       | Web search using Brave Search API with optional summarization | [Details](tools/web_search_brave.md)                                        |
+| `web_search_brave_with_document_backend` | Web search using Brave with RAG-based document processing     | [Details](tools/web_search_brave.md#web_search_brave_with_document_backend) |
+| `web_search_albert_rag`                  | ⚠️ **Deprecated** - Web search using Albert API with RAG      | [Details](tools/web_search_brave.md#deprecated-web_search_albert_rag)       |
+
+## Adding a New Tool
+
+To add a new tool to the system, follow these steps:
+
+### 1. Create the Tool Function
+
+Create a new Python file in `src/backend/chat/tools/` with your tool function. The function should:
+
+- Have clear type annotations
+- Include a comprehensive docstring (the LLM uses this to understand when to use the tool)
+- Accept `RunContext` as the first parameter if it needs access to conversation context
+- Return appropriate data types
+
+Example:
+```python
+"""My custom tool for the chat agent."""
+
+from pydantic_ai import RunContext
+
+def my_custom_tool(ctx: RunContext, param1: str, param2: int) -> dict:
+    """
+    Brief description of what the tool does.
+    
+    The LLM uses this description to decide when to call this tool.
+    
+    Args:
+        ctx (RunContext): The run context containing the conversation.
+        param1 (str): Description of parameter 1.
+        param2 (int): Description of parameter 2.
+    
+    Returns:
+        dict: Description of the return value.
+    """
+    # Your implementation here
+    return {"result": "example"}
+```
+
+### 2. Register the Tool
+
+Add your tool to the registry in `src/backend/chat/tools/__init__.py`:
+
+```python
+from .my_custom_tool import my_custom_tool
+
+def get_pydantic_tools_by_name(name: str) -> Tool:
+    """Get a tool by its name."""
+    tool_dict = {
+        "get_current_weather": Tool(get_current_weather, takes_ctx=False),
+        "web_search_brave": Tool(
+            web_search_brave, takes_ctx=False, prepare=only_if_web_search_enabled
+        ),
+        # Add your tool here
+        "my_custom_tool": Tool(
+            my_custom_tool, 
+            takes_ctx=True,  # Set to True if your tool needs RunContext
+            # prepare=only_if_web_search_enabled  # Optional: add conditions
+        ),
+    }
+    return tool_dict[name]
+```
+
+### 3. Update Imports
+
+Don't forget to import your tool function at the top of `__init__.py`:
+
+```python
+from .my_custom_tool import my_custom_tool
+```
+
+### 4. Add to Model Configuration
+
+Add your tool name to the `tools` list in your LLM configuration file or 
+to the `AI_AGENT_TOOLS` environment variable for local/test purpose.
+
+## Tool Preparation: Conditional Tool Availability
+
+Some tools should only be available under certain conditions. The `prepare` parameter in the `Tool` constructor 
+allows you to specify a function that determines whether a tool should be included.
+
+### The `only_if_web_search_enabled` Prepare Function
+
+This is a built-in prepare function that checks if web search feature is enabled in the conversation context:
+
+```python
+async def only_if_web_search_enabled(ctx, tool_def: ToolDefinition) -> ToolDefinition | None:
+    """Prepare function to include a tool only if web search is enabled in the context."""
+    return tool_def if ctx.deps.web_search_enabled else None
+```
+
+### Usage
+
+All web search tools use this prepare function:
+
+```python
+"web_search_brave": Tool(
+    web_search_brave, 
+    takes_ctx=False, 
+    prepare=only_if_web_search_enabled
+),
+```
+
+This ensures that web search tools are only available when the user or conversation settings have enabled web search functionality.
+
+### Creating Custom Prepare Functions
+
+You can create your own prepare functions for custom conditions:
+
+```python
+async def only_if_feature_enabled(ctx, tool_def: ToolDefinition) -> ToolDefinition | None:
+    """Include tool only if a specific feature is enabled."""
+    return tool_def if ctx.deps.feature_enabled else None
+```
+
+## Web Search Enable/Disable
+
+Web search tools can be toggled on or off based on conversation settings. When web search is disabled:
+- Web search tools are not included in the agent's available tools
+- The LLM cannot make web search calls even if it tries
+- This is enforced by the `only_if_web_search_enabled` prepare function
+
+The `web_search_enabled` flag is typically set:
+- Per conversation in the conversation settings
+- Per user preference
+- Through admin configuration
+
+## Best Practices
+
+1. **Keep tools focused** - Each tool should do one thing well
+2. **Clear documentation** - The LLM relies on docstrings to understand when to use tools
+3. **Error handling** - Tools should handle errors gracefully and return meaningful messages
+4. **Performance** - Be mindful of API rate limits and timeout values
+5. **Security** - Never log sensitive data (API keys, user data, etc.)
+6. **Caching** - Use Django's cache framework for expensive operations when appropriate
+
+## Troubleshooting
+
+### Tool Not Being Called
+
+If the LLM isn't calling your tool:
+- Check that the tool is registered in `get_pydantic_tools_by_name`
+- Verify the tool is in the model's `tools` configuration
+- Review the tool's docstring - make it clearer when the tool should be used
+- Check if any `prepare` function is preventing the tool from being included
+
+### Tool Errors
+
+If a tool is throwing errors:
+- Check the logs for detailed error messages
+- Verify all required environment variables are set
+- Ensure the tool's dependencies are installed
+- Test the tool function independently
+
+We recommend wrapping external API calls in try/except blocks to handle potential issues gracefully and use
+the Pydantic AI `ModelRetry` exception to let the LLM manage the errors.
+
+### Tool Response Issues
+
+If the LLM isn't using the tool response correctly:
+- Ensure the return type is clear and well-structured
+- Consider returning a `ToolReturn` object with metadata
+- Check if the response format matches what the LLM expects
+
+## See Also
+
+- [Web Search Configuration](llm-configuration.md)
+- [Architecture](architecture.md)
+- [Environment Variables](env.md)
+

docs/tools/get_current_weather.md

@@ -0,0 +1,113 @@
+# get_current_weather Tool
+
+## Overview
+
+The `get_current_weather` tool is a **fake weather tool** designed for testing and demonstration purposes. It does not connect to any real weather API and always returns hardcoded weather data.
+
+## Purpose
+
+This tool is useful for:
+- **Testing** the tool calling functionality of LLMs
+- **Demonstrating** how tools work without requiring API keys
+- **Development** and debugging of the agent system
+- **Example implementation** for creating new tools
+
+⚠️ **Warning**: This tool should **not** be used in production environments. It always returns fake data regardless of the location or conditions.
+
+## Configuration
+
+### Add to Model
+
+To enable this tool for a model, add it to the `tools` list in your LLM configuration:
+
+```json
+{
+  "models": [
+    {
+      "hrid": "my-model",
+      "tools": [
+        "get_current_weather"
+      ]
+    }
+  ]
+}
+```
+
+Or via environment variable when using local environment settings:
+```ini
+AI_AGENT_TOOLS=get_current_weather
+```
+
+### No Additional Settings Required
+
+This tool does not require any API keys, environment variables, or additional configuration.
+
+## Function Signature
+
+```python
+def get_current_weather(location: str, unit: str) -> dict:
+    """
+    Get the current weather in a given location.
+
+    Args:
+        location (str): The city and state, e.g. San Francisco, CA.
+        unit (str): The unit of temperature, either 'celsius' or 'fahrenheit'.
+
+    Returns:
+        dict: A dictionary containing the location, temperature, and unit.
+    """
+```
+
+## Parameters
+
+| Parameter  | Type | Required | Description                                                     |
+|------------|------|----------|-----------------------------------------------------------------|
+| `location` | str  | Yes      | The city and state (e.g., "San Francisco, CA", "Paris, France") |
+| `unit`     | str  | Yes      | Temperature unit: either "celsius" or "fahrenheit"              |
+
+## Return Value
+
+Returns a dictionary with the following structure:
+
+```python
+{
+    "location": str,      # The location that was queried
+    "temperature": int,   # Always 22°C or 72°F
+    "unit": str          # The unit that was requested
+}
+```
+
+## How the LLM Uses It
+
+When a user asks about weather, the LLM will:
+
+1. **Recognize** the weather-related query
+2. **Extract** the location from the user's message
+3. **Determine** the appropriate unit (often from context or user preference)
+4. **Call** the `get_current_weather` tool
+5. **Receive** the fake weather data
+6. **Format** a response to the user
+
+### Example Conversation
+
+**User**: "What's the weather like in London?"
+
+**LLM** (internal): *Calls `get_current_weather("London, UK", "celsius")`*
+
+**Tool Response**: 
+```json
+{
+  "location": "London, UK",
+  "temperature": 22,
+  "unit": "celsius"
+}
+```
+
+**LLM** (to user): "The current weather in London, UK is 22°C."
+
+## See Also
+
+- [Tools Overview](../tools.md)
+- [Adding a New Tool](../tools.md#adding-a-new-tool)
+- [Testing Tools](../tools.md#testing-your-tools)
+

docs/tools/web_search_brave.md

@@ -0,0 +1,670 @@
+# Brave Web Search Tools
+
+## Overview
+
+The Brave web search tools enable the conversation agent to search the web using the [Brave Search API](https://brave.com/search/api/). 
+Brave Search is a privacy-focused search engine that provides comprehensive web search results.
+
+This documentation covers three related tools:
+1. **`web_search_brave`** - Standard web search with optional summarization
+2. **`web_search_brave_with_document_backend`** - Web search with RAG-based document processing
+3. **`web_search_albert_rag`** - ⚠️ **Deprecated** - Use `web_search_brave_with_document_backend` instead
+
+## Table of Contents
+
+- [Common Configuration](#common-configuration)
+- [web_search_brave](#web_search_brave)
+- [web_search_brave_with_document_backend](#web_search_brave_with_document_backend)
+- [Deprecated: web_search_albert_rag](#deprecated-web_search_albert_rag)
+- [Comparison](#comparison)
+- [Best Practices](#best-practices)
+- [Troubleshooting](#troubleshooting)
+
+---
+
+## Common Configuration
+
+### Prerequisites
+
+1. **Brave Search API Key**: Sign up at [Brave Search API](https://brave.com/search/api/) to get an API key
+2. **Environment Variables**: Configure the required settings
+
+### Common Environment Variables
+
+All Brave tools share these common settings:
+
+| Variable            | Required | Default | Description                                        |
+|---------------------|----------|---------|----------------------------------------------------|
+| `BRAVE_API_KEY`     | **Yes**  | None    | Your Brave Search API key                          |
+| `BRAVE_API_TIMEOUT` | No       | 5       | API request timeout in seconds                     |
+| `BRAVE_MAX_RESULTS` | No       | 8       | Maximum number of search results                   |
+| `BRAVE_CACHE_TTL`   | No       | 1800    | Cache time-to-live in seconds (30 minutes)         |
+
+### Search Parameters
+
+Check on the Brave API documentation for more details on these parameters:
+
+| Variable                      | Required | Default    | Description                                       |
+|-------------------------------|----------|------------|---------------------------------------------------|
+| `BRAVE_SEARCH_COUNTRY`        | No       | None       | Country code for search (e.g., "US", "FR")        |
+| `BRAVE_SEARCH_LANG`           | No       | None       | Language code (e.g., "en", "fr")                  |
+| `BRAVE_SEARCH_SAFE_SEARCH`    | No       | "moderate" | Safe search level: "off", "moderate", or "strict" |
+| `BRAVE_SEARCH_SPELLCHECK`     | No       | True       | Enable spell checking                             |
+| `BRAVE_SEARCH_EXTRA_SNIPPETS` | No       | True       | Fetch extra snippets from pages                   |
+
+
+Note: even if `BRAVE_SEARCH_EXTRA_SNIPPETS` is enabled, the API may not include them if you don't have a plan for this.
+This is why, in `web_search_brave`, we also fetch the page content ourselves when needed.
+
+### Configuration Example
+
+```bash
+# .env file
+BRAVE_API_KEY=BSA-your-api-key-here
+BRAVE_MAX_RESULTS=8
+BRAVE_MAX_WORKERS=4
+BRAVE_SEARCH_COUNTRY=US
+BRAVE_SEARCH_LANG=en
+BRAVE_SEARCH_SAFE_SEARCH=moderate
+```
+
+### Django Settings
+
+All Brave settings are defined in `src/backend/conversations/brave_settings.py`:
+
+```python
+class BraveSettings:
+    """Brave settings for web_search_brave tool."""
+
+    BRAVE_API_KEY = values.Value(
+        default=None,
+        environ_name="BRAVE_API_KEY",
+        environ_prefix=None,
+    )
+    # ... more settings
+```
+
+---
+
+## web_search_brave
+
+### Overview
+
+Standard Brave web search tool with optional LLM-based summarization of page content.
+
+### Purpose
+
+- Search the web for up-to-date information
+- Extract content from web pages
+- Optionally summarize content using an LLM
+- Provide structured results with snippets
+
+### Additional Configuration
+
+| Variable                      | Required | Default | Description                                     |
+|-------------------------------|----------|---------|-------------------------------------------------|
+| `BRAVE_SUMMARIZATION_ENABLED` | No       | False   | Enable LLM-based summarization of fetched pages |
+
+### Function Signature
+
+```python
+def web_search_brave(query: str) -> ToolReturn:
+    """
+    Search the web for up-to-date information
+
+    Args:
+        query (str): The query to search for.
+        
+    Returns:
+        ToolReturn: Formatted search results with metadata
+    """
+```
+
+### Return Value
+
+Returns a `ToolReturn` object with:
+
+```python
+ToolReturn(
+    return_value={
+        "0": {
+            "url": "https://example.com/page1",
+            "title": "Example Page Title",
+            "snippets": ["Extracted or summarized content..."]
+        },
+        "1": {
+            "url": "https://example.com/page2",
+            "title": "Another Page",
+            "snippets": ["More content..."]
+        }
+    },
+    metadata={
+        "sources": {
+            "https://example.com/page1",
+            "https://example.com/page2"
+        }
+    }
+)
+```
+
+### How It Works
+
+1. **Query API**: Sends search query to Brave Search API
+2. **Receive Results**: Gets list of matching web pages
+3. **Fetch Content**: For results without extra_snippets:
+   - Fetches the HTML content using `trafilatura`
+   - Extracts the main text content
+   - Caches the extracted content
+4. **Summarize (Optional)**: If `BRAVE_SUMMARIZATION_ENABLED=True`:
+   - Sends extracted content to summarization agent
+   - Receives concise summary focused on the query
+5. **Format Results**: Returns structured data with URLs, titles, and snippets
+
+### Workflow Diagram
+
+```
+User Query
+    ↓
+Brave Search API
+    ↓
+Search Results (URLs, titles, descriptions)
+    ↓
+[For each result without snippets]
+    ↓
+Fetch HTML (trafilatura) → Extract Text → Cache
+    ↓
+[If BRAVE_SUMMARIZATION_ENABLED]
+    ↓
+Summarization Agent (LLM)
+    ↓
+Summary Text
+    ↓
+Format & Return
+```
+
+### Caching
+
+Extracted content is cached to avoid repeated fetches:
+
+```python
+cache_key = f"web_search_brave:extract:{url}"
+cache.set(cache_key, document, settings.BRAVE_CACHE_TTL)
+```
+
+**Cache Duration**: Controlled by `BRAVE_CACHE_TTL` (default: 30 minutes)
+
+### Summarization
+
+When enabled, the tool uses the `SummarizationAgent` to condense page content:
+
+```python
+prompt = f"""
+Based on the following request, summarize the following text in a concise manner, 
+focusing on the key points regarding the user request. 
+The result should be up to 30 lines long.
+
+<user request>
+{query}
+</user request>
+
+<text to summarize>
+{text}
+</text to summarize>
+"""
+```
+
+**Note**: Summarization is costly (additional LLM calls). 
+Use only when necessary, we prefer the document vector search from `web_search_brave_with_document_backend`.
+
+### Add to Model
+
+```json
+{
+  "models": [
+    {
+      "hrid": "my-model",
+      "tools": [
+        "web_search_brave"
+      ]
+    }
+  ]
+}
+```
+
+### Example Usage
+
+**User**: "What are the new features in Django 5.0?"
+
+**Tool Call**: `web_search_brave("Django 5.0 new features")`
+
+**Tool Response**:
+```python
+{
+    "0": {
+        "url": "https://docs.djangoproject.com/en/5.0/releases/5.0/",
+        "title": "Django 5.0 release notes",
+        "snippets": ["Django 5.0 introduces several new features including..."]
+    },
+    # ... more results
+}
+```
+
+### Registration
+
+```python
+"web_search_brave": Tool(
+    web_search_brave, 
+    takes_ctx=False, 
+    prepare=only_if_web_search_enabled
+)
+```
+
+---
+
+## web_search_brave_with_document_backend
+
+### Overview
+
+Advanced Brave web search tool that uses RAG (Retrieval-Augmented Generation) 
+with a document backend for intelligent content processing and retrieval.
+
+### Purpose
+
+- Search the web and process results through a RAG system
+- Store fetched documents in a temporary vector database
+- Perform semantic search across fetched content
+- Return the most relevant chunks based on the query
+
+### Additional Configuration
+
+| Variable                            | Required | Default          | Description                                  |
+|-------------------------------------|----------|------------------|----------------------------------------------|
+| `BRAVE_RAG_WEB_SEARCH_CHUNK_NUMBER` | No       | 10               | Number of chunks to retrieve from RAG search |
+| `RAG_DOCUMENT_SEARCH_BACKEND`       | No       | AlbertRagBackend | Document backend for RAG processing          |
+
+### Function Signature
+
+```python
+def web_search_brave_with_document_backend(ctx: RunContext, query: str) -> ToolReturn:
+    """
+    Search the web for up-to-date information
+
+    Args:
+        ctx (RunContext): The run context containing the conversation.
+        query (str): The query to search for.
+        
+    Returns:
+        ToolReturn: Formatted search results with RAG-enhanced snippets
+    """
+```
+
+### How It Works
+
+1. **Query API**: Sends search query to Brave Search API
+2. **Receive Results**: Gets list of matching web pages
+3. **Create Temporary Collection**: Creates a temporary vector database collection
+4. **Fetch & Store**: For each result:
+   - Fetches the HTML content
+   - Extracts the main text
+   - Stores in the temporary document backend
+5. **RAG Search**: Performs semantic search across stored documents
+6. **Map Results**: Maps RAG chunks back to original search results
+7. **Format & Return**: Returns structured data with enhanced snippets
+8. **Cleanup**: Temporary collection is automatically deleted
+
+### Workflow Diagram
+
+```
+User Query
+    ↓
+Brave Search API
+    ↓
+Search Results (URLs)
+    ↓
+Create Temporary Vector Collection
+    ↓
+[For each URL]
+    ↓
+Fetch HTML → Extract Text → Store in Vector DB
+    ↓
+RAG Semantic Search
+    ↓
+Retrieve Most Relevant Chunks
+    ↓
+Map Chunks to Original URLs
+    ↓
+Format & Return
+    ↓
+Delete Temporary Collection
+```
+
+### Temporary Collection
+
+The tool creates a temporary collection with a unique ID:
+
+```python
+with document_store_backend.temporary_collection(f"tmp-{uuid.uuid4()}") as document_store:
+    # Fetch and store documents
+    # Perform search
+    # Collection is automatically deleted on exit
+```
+
+### RAG Search
+
+The RAG backend performs semantic search to find the most relevant content:
+
+```python
+rag_results = document_store.search(
+    query,
+    results_count=settings.BRAVE_RAG_WEB_SEARCH_CHUNK_NUMBER,
+)
+```
+
+Returns chunks ranked by relevance to the query, not just keyword matching.
+
+### Token Usage Tracking
+
+The tool tracks LLM tokens used during RAG processing:
+
+```python
+ctx.usage += RunUsage(
+    input_tokens=rag_results.usage.prompt_tokens,
+    output_tokens=rag_results.usage.completion_tokens,
+)
+```
+
+### Document Backend
+
+The default backend is `AlbertRagBackend`, but you can configure a different one:
+
+```bash
+RAG_DOCUMENT_SEARCH_BACKEND=chat.agent_rag.document_rag_backends.custom_backend.CustomBackend
+```
+
+### Add to Model
+
+```json
+{
+  "models": [
+    {
+      "hrid": "my-model",
+      "tools": [
+        "web_search_brave_with_document_backend"
+      ]
+    }
+  ]
+}
+```
+
+### Example Usage
+
+**User**: "Explain the concept of async views in Django"
+
+**Tool Call**: `web_search_brave_with_document_backend(ctx, "Django async views explained")`
+
+**Tool Response**:
+```python
+{
+    "0": {
+        "url": "https://docs.djangoproject.com/en/stable/topics/async/",
+        "title": "Asynchronous support",
+        "snippets": [
+            "Django has support for writing asynchronous views...",
+            "Async views are declared using Python's async def syntax..."
+        ]
+    },
+    # ... more results with relevant chunks
+}
+```
+
+### Registration
+
+```python
+"web_search_brave_with_document_backend": Tool(
+    web_search_brave_with_document_backend,
+    takes_ctx=True,
+    prepare=only_if_web_search_enabled,
+)
+```
+
+### Advantages Over Standard web_search_brave
+
+| Feature           | web_search_brave               | web_search_brave_with_document_backend |
+|-------------------|--------------------------------|----------------------------------------|
+| Content Retrieval | Full page or summary           | Semantic chunks                        |
+| Relevance         | Keyword-based                  | Semantic similarity                    |
+| Token Efficiency  | May include irrelevant content | Only relevant chunks                   |
+| Processing        | Simpler, faster                | More intelligent, slower               |
+| Cost              | Lower                          | Higher (RAG processing)                |
+| Best For          | General search                 | Deep research, technical queries       |
+
+---
+
+## Deprecated: web_search_albert_rag
+
+### ⚠️ Deprecation Notice
+
+The `web_search_albert_rag` tool is **deprecated** and should not be used in new implementations.
+
+**Replacement**: Use `web_search_brave_with_document_backend` instead, which provides:
+- Better performance
+- More control over the RAG backend
+- Temporary collections (no cleanup issues)
+- Token usage tracking
+- Parallel processing support
+
+### Why Deprecated?
+
+- Limited to Albert API only
+- No control over document backend
+- Less flexible than the new approach
+- Maintenance burden
+
+### Timeline
+
+- **Current**: Still functional but not recommended
+- **Future**: Will be removed in a future version
+
+---
+
+## Comparison
+
+### When to Use Which Tool?
+
+#### Use `web_search_brave`
+
+✅ **Best for**:
+- General web search queries
+- Quick information retrieval
+- When speed is important
+- Lower cost requirements
+- Simple fact-finding
+
+❌ **Not ideal for**:
+- Deep research requiring precise context
+- Technical documentation queries
+- When semantic relevance is crucial
+
+#### Use `web_search_brave_with_document_backend`
+
+✅ **Best for**:
+- Complex technical queries
+- Research requiring precise context
+- When semantic relevance is important
+- Questions needing deep understanding
+- Documentation and how-to queries
+
+❌ **Not ideal for**:
+- Simple factual queries
+- When speed is critical
+- Budget-constrained scenarios
+- High-volume usage
+
+---
+
+## Best Practices
+
+### Query Formulation
+
+Help the LLM formulate effective queries:
+
+```python
+# Good queries
+"Python asyncio tutorial 2024"
+"Django REST framework authentication"
+"React hooks best practices"
+
+# Poor queries
+"tell me about programming"  # Too vague
+"how do I do the thing with the stuff"  # Unclear
+```
+
+### Performance Optimization
+
+#### 1. Optimize Cache
+
+```bash
+# Longer cache for stable content
+BRAVE_CACHE_TTL=3600  # 1 hour
+
+# Shorter cache for dynamic content
+BRAVE_CACHE_TTL=300   # 5 minutes
+```
+
+#### 2. Control Result Count
+
+```bash
+# Fewer results = faster responses
+BRAVE_MAX_RESULTS=5
+
+# More results = more comprehensive
+BRAVE_MAX_RESULTS=10
+```
+
+### Summarization Best Practices
+
+Only enable summarization when needed:
+
+```bash
+# Enable for long-form content
+BRAVE_SUMMARIZATION_ENABLED=True
+
+# Disable for speed
+BRAVE_SUMMARIZATION_ENABLED=False
+```
+
+**Cost consideration**: Summarization makes additional LLM calls for each result, 
+significantly increasing costs (and execution time).
+
+### RAG Configuration
+
+For `web_search_brave_with_document_backend`:
+
+```bash
+# More chunks = more context, higher cost
+BRAVE_RAG_WEB_SEARCH_CHUNK_NUMBER=10
+
+# Fewer chunks = faster, less context
+BRAVE_RAG_WEB_SEARCH_CHUNK_NUMBER=5
+```
+
+### Search Parameters
+
+```bash
+# Localize results
+BRAVE_SEARCH_COUNTRY=FR
+BRAVE_SEARCH_LANG=fr
+
+# Safe search for public deployments
+BRAVE_SEARCH_SAFE_SEARCH=strict
+
+# Enable spell check for better results
+BRAVE_SEARCH_SPELLCHECK=True
+```
+
+---
+
+## Troubleshooting
+
+### Common Issues
+
+#### 1. No Results Returned
+
+**Symptoms**: Empty results or no snippets
+
+**Causes**:
+- Query too specific
+- Content extraction failed
+- Trafilatura couldn't parse the pages
+
+**Solutions**:
+```bash
+# Enable extra snippets
+BRAVE_SEARCH_EXTRA_SNIPPETS=True
+
+# Increase result count
+BRAVE_MAX_RESULTS=10
+
+# Check logs for extraction errors
+```
+
+#### 2. API Errors
+
+**Symptoms**: HTTP errors, authentication failures
+
+**Causes**:
+- Invalid API key
+- Rate limit exceeded
+- API service issues
+
+**Solutions**:
+```bash
+# Verify API key is set
+echo $BRAVE_API_KEY
+
+# Check Brave API dashboard for limits
+# Implement rate limiting in your application
+```
+
+#### 3. The tool is not being called
+**Symptoms**: LLM doesn't use the tool even when appropriate
+
+**Causes**:
+- Web search not enabled for the conversation
+- Tool not in model configuration
+
+**Solutions**:
+- Check conversation settings have `web_search_enabled=True`
+- Verify tool is in the model's `tools` list
+
+---
+
+## Security Considerations
+
+This tool is quite "raw", so be cautious about:
+- the results returned by the web search
+- the context size which might be large when not using summarization or RAG if long results are returned
+- the query content which might include sensitive information
+- ...
+
+### Content Validation
+
+Be aware that fetched content may contain:
+- Malicious scripts (mitigated by text extraction)
+- Inappropriate content
+- Misinformation
+- Biased information
+
+The LLM should evaluate sources critically.
+
+
+---
+
+## See Also
+
+- [Tools Overview](../tools.md)
+- [Tavily Web Search Tool](web_search_tavily.md)
+- [LLM Configuration](../llm-configuration.md)
+- [Environment Variables](../env.md)
+- [Brave Search API Documentation](https://brave.com/search/api/)
+

docs/tools/web_search_tavily.md

@@ -0,0 +1,370 @@
+# web_search_tavily Tool
+
+## Overview
+
+The `web_search_tavily` tool enables the conversation agent to search the web for up-to-date 
+information using the [Tavily Search API](https://tavily.com/).
+
+## Purpose
+
+This tool allows the LLM to:
+- Access current, real-time information beyond its training data
+- Answer questions about recent events, news, or developments
+- Provide factual information with sources
+- Retrieve specific information from the web
+
+## Configuration
+
+### Prerequisites
+
+1. **Tavily API Key**: Sign up at [Tavily](https://tavily.com/) to get an API key
+2. **Environment Variables**: Configure the required settings
+
+### Environment Variables
+
+| Variable             | Required | Default | Description                                |
+|----------------------|----------|---------|--------------------------------------------|
+| `TAVILY_API_KEY`     | **Yes**  | None    | Your Tavily API key                        |
+| `TAVILY_MAX_RESULTS` | No       | 5       | Maximum number of search results to return |
+| `TAVILY_API_TIMEOUT` | No       | 10      | API request timeout in seconds             |
+
+### Configuration Example
+
+```bash
+# .env file
+TAVILY_API_KEY=tvly-your-api-key-here
+TAVILY_MAX_RESULTS=5
+TAVILY_API_TIMEOUT=10
+```
+
+### Add to Model
+
+To enable this tool for a model, add it to the `tools` list in your LLM configuration:
+
+```json
+{
+  "models": [
+    {
+      "hrid": "my-model",
+      "tools": [
+        "web_search_tavily"
+      ]
+    }
+  ]
+}
+```
+
+Or via environment variable when using local environment settings:
+
+```ini
+AI_AGENT_TOOLS=web_search_tavily
+```
+
+## Function Signature
+
+```python
+def web_search_tavily(query: str) -> list[dict]:
+    """
+    Search the web for up-to-date information
+
+    Args:
+        query (str): The query to search for.
+
+    Returns:
+        list[dict]: A list of search results, each represented as a dictionary.
+    """
+```
+
+## Parameters
+
+| Parameter | Type | Required | Description             |
+|-----------|------|----------|-------------------------|
+| `query`   | str  | Yes      | The search query string |
+
+## Return Value
+
+Returns a list of dictionaries, each containing:
+
+```python
+{
+    "link": str,      # URL of the result
+    "title": str,     # Title of the page
+    "snippet": str    # Content snippet from the page
+}
+```
+
+### Example Return Value
+
+```python
+[
+    {
+        "link": "https://example.com/article1",
+        "title": "Introduction to Python",
+        "snippet": "Python is a high-level programming language known for its simplicity..."
+    },
+    {
+        "link": "https://example.com/article2",
+        "title": "Python Best Practices",
+        "snippet": "Follow these best practices to write clean and efficient Python code..."
+    }
+]
+```
+
+## How the LLM Uses It
+
+When a user asks for current information or specific facts:
+
+1. **LLM recognizes** the need for external information
+2. **Formulates** an appropriate search query
+3. **Calls** `web_search_tavily(query="search terms")`
+4. **Receives** a list of search results
+5. **Synthesizes** the information into a response
+6. **Provides** the answer with source references
+
+### Example Conversation
+
+**User**: "What are the latest developments in quantum computing?"
+
+**LLM** (internal): *Calls `web_search_tavily("latest developments quantum computing 2024")`*
+
+**Tool Response**: 
+```python
+[
+    {
+        "link": "https://techcrunch.com/quantum-news",
+        "title": "Major Breakthrough in Quantum Computing",
+        "snippet": "Researchers announced a significant breakthrough..."
+    },
+    # ... more results
+]
+```
+
+**LLM** (to user): "Based on recent sources, there have been several developments in quantum computing. 
+Researchers recently announced a breakthrough in error correction. Additionally, new quantum processors 
+with improved qubit stability have been unveiled..."
+
+## Implementation Details
+
+### Source Code
+
+Located at: `src/backend/chat/tools/web_search_tavily.py`
+
+```python
+"""Web search tool using Tavily for the chat agent."""
+
+from django.conf import settings
+
+import requests
+
+
+def web_search_tavily(query: str) -> list[dict]:
+    """
+    Search the web for up-to-date information
+
+    Args:
+        query (str): The query to search for.
+
+    Returns:
+        list[dict]: A list of search results, each represented as a dictionary.
+    """
+    url = "https://api.tavily.com/search"
+    data = {
+        "query": query,
+        "api_key": settings.TAVILY_API_KEY,
+        "max_results": settings.TAVILY_MAX_RESULTS,
+    }
+    response = requests.post(url, json=data, timeout=settings.TAVILY_API_TIMEOUT)
+    response.raise_for_status()
+
+    json_response = response.json()
+
+    raw_search_results = json_response.get("results", [])
+
+    return [
+        {
+            "link": result["url"],
+            "title": result.get("title", ""),
+            "snippet": result.get("content"),
+        }
+        for result in raw_search_results
+    ]
+```
+
+### Registration
+
+The tool is registered in `src/backend/chat/tools/__init__.py`:
+
+```python
+"web_search_tavily": Tool(
+    web_search_tavily, 
+    takes_ctx=False, 
+    prepare=only_if_web_search_enabled
+)
+```
+
+Note that:
+- `takes_ctx=False` - This tool doesn't need the conversation context
+- `prepare=only_if_web_search_enabled` - Only available when web search is enabled
+
+## Django Settings
+
+The tool uses these Django settings from `settings.py`:
+
+```python
+# Tavily API
+TAVILY_API_KEY = values.Value(
+    None,  # Tavily API key is not set by default
+    environ_name="TAVILY_API_KEY",
+    environ_prefix=None,
+)
+TAVILY_MAX_RESULTS = values.PositiveIntegerValue(
+    default=5,
+    environ_name="TAVILY_MAX_RESULTS",
+    environ_prefix=None,
+)
+TAVILY_API_TIMEOUT = values.PositiveIntegerValue(
+    default=10,  # seconds
+    environ_name="TAVILY_API_TIMEOUT",
+    environ_prefix=None,
+)
+```
+
+## Error Handling
+
+The tool may raise exceptions in the following cases:
+
+### Missing API Key
+```python
+# If TAVILY_API_KEY is not set
+AttributeError: 'Settings' object has no attribute 'TAVILY_API_KEY'
+```
+
+**Solution**: Set the `TAVILY_API_KEY` environment variable
+
+### API Errors
+```python
+# If the API request fails
+requests.exceptions.HTTPError: 401 Unauthorized
+```
+
+**Possible causes**:
+- Invalid API key
+- Exceeded rate limits
+- API service unavailable
+
+### Timeout Errors
+```python
+# If the request takes too long
+requests.exceptions.Timeout
+```
+
+**Solution**: Increase `TAVILY_API_TIMEOUT` or check network connectivity
+
+## Best Practices
+
+### Query Formulation
+
+The LLM should formulate queries that are:
+- **Specific and focused** - Better results with targeted queries
+- **Up-to-date** - Include year or "latest" when relevant
+- **Clear** - Avoid ambiguous terms
+- **Concise** - Remove unnecessary words
+
+Good query examples:
+- ✅ "quantum computing breakthroughs 2024"
+- ✅ "latest Python 3.12 features"
+- ✅ "climate change COP29 outcomes"
+
+Poor query examples:
+- ❌ "tell me about stuff happening" (too vague)
+- ❌ "what is the weather like today in Paris on November 5th 2024 at 3pm" (too specific/long)
+
+### Rate Limiting
+
+Be aware of Tavily API rate limits:
+- Free tier: Limited requests per month
+- Paid tiers: Higher limits
+
+Monitor your usage and implement caching if needed.
+
+### Result Count
+
+The `TAVILY_MAX_RESULTS` setting controls how many results are returned:
+- **Lower values (3-5)**: Faster responses, less context for LLM
+- **Higher values (8-10)**: More comprehensive, but slower and more expensive
+
+Recommended: **5 results** for most use cases
+
+## Troubleshooting
+
+### Tool Not Being Called
+
+**Symptoms**: LLM doesn't use web search even when appropriate
+
+**Possible causes**:
+1. Web search not enabled for the conversation
+2. Tool not in model configuration
+3. API key not set
+
+**Solutions**:
+1. Check conversation settings have `web_search_enabled=True`
+2. Verify tool is in the model's `tools` list
+3. Confirm `TAVILY_API_KEY` is set
+
+### No Results Returned
+
+**Symptoms**: Tool returns empty list
+
+**Possible causes**:
+1. Query too specific
+2. No matching results
+3. API filtering results
+
+**Solutions**:
+1. Try broader query terms
+2. Check Tavily dashboard for query logs
+3. Review API response in logs
+
+### Slow Responses
+
+**Symptoms**: Tool takes a long time to respond
+
+**Possible causes**:
+1. Network latency
+2. Tavily API slow
+3. Timeout too high
+
+**Solutions**:
+1. Check network connectivity
+2. Monitor Tavily status page
+3. Adjust `TAVILY_API_TIMEOUT` if needed
+
+## Security Considerations
+
+This tool is quite "raw", and was currently only used for test purpose, so be cautious about:
+- the results returned by the web search
+- the context size which might be large if many results are returned
+- the query content which might include sensitive information
+- ...
+
+## Performance Optimization
+
+### Query Optimization
+
+You may want to help the LLM formulate better queries by including something like this in the system prompt:
+
+```
+When using web search:
+- Use specific, focused queries
+- Include relevant time periods if needed
+- Avoid unnecessary words
+- Combine related terms
+```
+
+## See Also
+
+- [Tools Overview](../tools.md)
+- [Brave Web Search Tool](web_search_brave.md)
+- [Web Search Configuration](../llm-configuration.md)
+- [Environment Variables](../env.md)
+

env.d/development/common.e2e.dist

@@ -1,6 +1,5 @@
 # For the CI job test-e2e
 BURST_THROTTLE_RATES="200/minute"
-DJANGO_SERVER_TO_SERVER_API_TOKENS=test-e2e
 SUSTAINED_THROTTLE_RATES="200/hour"
 
 # Features

src/backend/chat/agent_rag/document_rag_backends/albert_rag_backend.py

@@ -8,6 +8,7 @@ from urllib.parse import urljoin
 
 from django.conf import settings
 
+import httpx
 import requests
 
 from chat.agent_rag.albert_api_constants import Searches
@@ -65,6 +66,27 @@ class AlbertRagBackend(BaseRagBackend):  # pylint: disable=too-many-instance-att
         self.collection_id = str(response.json()["id"])
         return self.collection_id
 
+    async def acreate_collection(self, name: str, description: Optional[str] = None) -> str:
+        """
+        Create a temporary collection for the search operation.
+        This method should handle the logic to create or retrieve an existing collection.
+        """
+        async with httpx.AsyncClient(timeout=settings.ALBERT_API_TIMEOUT) as client:
+            response = await client.post(
+                self._collections_endpoint,
+                headers=self._headers,
+                json={
+                    "name": name,
+                    "description": description or self._default_collection_description,
+                    "visibility": "private",
+                },
+                timeout=settings.ALBERT_API_TIMEOUT,
+            )
+            response.raise_for_status()
+
+        self.collection_id = str(response.json()["id"])
+        return self.collection_id
+
     def delete_collection(self) -> None:
         """
         Delete the current collection
@@ -76,6 +98,18 @@ class AlbertRagBackend(BaseRagBackend):  # pylint: disable=too-many-instance-att
         )
         response.raise_for_status()
 
+    async def adelete_collection(self) -> None:
+        """
+        Asynchronously delete the current collection
+        """
+        async with httpx.AsyncClient(timeout=settings.ALBERT_API_TIMEOUT) as client:
+            response = await client.delete(
+                urljoin(f"{self._collections_endpoint}/", self.collection_id),
+                headers=self._headers,
+                timeout=settings.ALBERT_API_TIMEOUT,
+            )
+            response.raise_for_status()
+
     def parse_pdf_document(self, name: str, content_type: str, content: BytesIO) -> str:
         """
         Parse the PDF document content and return the text content.
@@ -150,12 +184,38 @@ class AlbertRagBackend(BaseRagBackend):  # pylint: disable=too-many-instance-att
         logger.debug(response.json())
         response.raise_for_status()
 
-    def search(self, query) -> RAGWebResults:
+    async def astore_document(self, name: str, content: str) -> None:
+        """
+        Store the document content in the Albert collection.
+        This method should handle the logic to send the document content to the Albert API.
+
+        Args:
+            name (str): The name of the document.
+            content (str): The content of the document in Markdown format.
+        """
+        async with httpx.AsyncClient(timeout=settings.ALBERT_API_TIMEOUT) as client:
+            response = await client.post(
+                urljoin(self._base_url, self._documents_endpoint),
+                headers=self._headers,
+                files={
+                    "file": (f"{name}.md", BytesIO(content.encode("utf-8")), "text/markdown"),
+                },
+                data={
+                    "collection": int(self.collection_id),
+                    "metadata": json.dumps({"document_name": name}),  # undocumented API
+                },
+                timeout=settings.ALBERT_API_TIMEOUT,
+            )
+            logger.debug(response.json())
+            response.raise_for_status()
+
+    def search(self, query, results_count: int = 4) -> RAGWebResults:
         """
         Perform a search using the Albert API based on the provided query.
 
         Args:
             query (str): The search query.
+            results_count (int): The number of results to return.
 
         Returns:
             RAGWebResults: The search results.
@@ -167,6 +227,7 @@ class AlbertRagBackend(BaseRagBackend):  # pylint: disable=too-many-instance-att
                 "collections": [int(self.collection_id)],
                 "prompt": query,
                 "score_threshold": 0.6,
+                "k": results_count,  # Number of chunks to return from the search
             },
             timeout=settings.ALBERT_API_TIMEOUT,
         )
@@ -188,3 +249,48 @@ class AlbertRagBackend(BaseRagBackend):  # pylint: disable=too-many-instance-att
                 completion_tokens=searches.usage.completion_tokens,
             ),
         )
+
+    async def asearch(self, query, results_count: int = 4) -> RAGWebResults:
+        """
+        Perform an asynchronous search using the Albert API based on the provided query.
+
+        Args:
+            query (str): The search query.
+            results_count (int): The number of results to return.
+
+        Returns:
+            RAGWebResults: The search results.
+        """
+        async with httpx.AsyncClient(timeout=settings.ALBERT_API_TIMEOUT) as client:
+            response = await client.post(
+                urljoin(self._base_url, self._search_endpoint),
+                headers=self._headers,
+                json={
+                    "collections": [int(self.collection_id)],
+                    "prompt": query,
+                    "score_threshold": 0.6,
+                    "k": results_count,  # Number of chunks to return from the search
+                },
+                timeout=settings.ALBERT_API_TIMEOUT,
+            )
+
+            logger.debug("Search response: %s %s", response.text, response.status_code)
+
+            response.raise_for_status()
+
+        searches = Searches(**response.json())
+
+        return RAGWebResults(
+            data=[
+                RAGWebResult(
+                    url=result.chunk.metadata["document_name"],
+                    content=result.chunk.content,
+                    score=result.score,
+                )
+                for result in searches.data
+            ],
+            usage=RAGWebUsage(
+                prompt_tokens=searches.usage.prompt_tokens,
+                completion_tokens=searches.usage.completion_tokens,
+            ),
+        )

src/backend/chat/agent_rag/document_rag_backends/base_rag_backend.py

@@ -1,10 +1,12 @@
 """Implementation of the Albert API for RAG document search."""
 
 import logging
-from contextlib import contextmanager
+from contextlib import asynccontextmanager, contextmanager
 from io import BytesIO
 from typing import Optional
 
+from asgiref.sync import sync_to_async
+
 from chat.agent_rag.constants import RAGWebResults
 
 logger = logging.getLogger(__name__)
@@ -25,6 +27,13 @@ class BaseRagBackend:
         """
         raise NotImplementedError("Must be implemented in subclass.")
 
+    async def acreate_collection(self, name: str, description: Optional[str] = None) -> str:
+        """
+        Create a temporary collection for the search operation.
+        This method should handle the logic to create or retrieve an existing collection.
+        """
+        return await sync_to_async(self.create_collection)(name=name, description=description)
+
     def parse_document(self, name: str, content_type: str, content: BytesIO):
         """
         Parse the document and prepare it for the search operation.
@@ -43,8 +52,8 @@ class BaseRagBackend:
 
     def store_document(self, name: str, content: str) -> None:
         """
-        Store the document content in the Albert collection.
-        This method should handle the logic to send the document content to the Albert API.
+        Store the document content in the collection.
+        This method should handle the logic to send the document content to the API.
 
         Args:
             name (str): The name of the document.
@@ -52,6 +61,17 @@ class BaseRagBackend:
         """
         raise NotImplementedError("Must be implemented in subclass.")
 
+    async def astore_document(self, name: str, content: str) -> None:
+        """
+        Store the document content in the collection.
+        This method should handle the logic to send the document content to the API.
+
+        Args:
+            name (str): The name of the document.
+            content (str): The content of the document in Markdown format.
+        """
+        return await sync_to_async(self.store_document)(name=name, content=content)
+
     def parse_and_store_document(self, name: str, content_type: str, content: BytesIO) -> str:
         """
         Parse the document and store it in the Albert collection.
@@ -75,12 +95,25 @@ class BaseRagBackend:
         """
         raise NotImplementedError("Must be implemented in subclass.")
 
-    def search(self, query) -> RAGWebResults:
+    async def adelete_collection(self) -> None:
+        """
+        Delete the collection.
+        This method should handle the logic to delete the collection from the backend.
+        """
+        return await sync_to_async(self.delete_collection)()
+
+    def search(self, query, results_count: int = 4) -> RAGWebResults:
         """
         Search the collection for the given query.
         """
         raise NotImplementedError("Must be implemented in subclass.")
 
+    async def asearch(self, query, results_count: int = 4) -> RAGWebResults:
+        """
+        Search the collection for the given query.
+        """
+        return await sync_to_async(self.search)(query=query, results_count=results_count)
+
     @classmethod
     @contextmanager
     def temporary_collection(cls, name: str, description: Optional[str] = None):
@@ -92,3 +125,15 @@ class BaseRagBackend:
             yield backend
         finally:
             backend.delete_collection()
+
+    @classmethod
+    @asynccontextmanager
+    async def temporary_collection_async(cls, name: str, description: Optional[str] = None):
+        """Context manager for RAG backend with temporary collections."""
+        backend = cls()
+
+        await backend.acreate_collection(name=name, description=description)
+        try:
+            yield backend
+        finally:
+            await backend.adelete_collection()

src/backend/chat/agents/conversation.py

@@ -1,10 +1,15 @@
 """Build the main conversation agent."""
 
+import asyncio
 import dataclasses
 import logging
 
+from django.conf import settings
 from django.utils import formats, timezone
 
+from pydantic_ai import ModelMessage
+from pydantic_ai.models.function import AgentInfo, FunctionModel
+
 from core.enums import get_language_name
 
 from .base import BaseAgent
@@ -12,6 +17,79 @@ from .base import BaseAgent
 logger = logging.getLogger(__name__)
 
 
+MOCKED_RESPONSE = """
+# **Ode to the AI Assistant** 🤖✨
+
+In Paris streets where old meets new, 🗼🇫🇷  
+A helper bright in digital hue,  
+With circuits fast and code so tight,  
+The LaSuite’s bot—oh, what a sight! 🌟
+
+**A chatbot kind**, with wittiness so grand, 💬💡  
+It lends a hand to all the land,  
+From civil servants, bold and wise,  
+To those who seek with hopeful eyes.
+
+It answers quick, it never tires, ⚡🔄  
+With facts and tips to quench desires,  
+A guide so keen, a friend so true,  
+It’s there for **you**—yes, me and you!
+ 
+With **Markdown flair** and emoji cheer, 📝🎨  
+It makes the complex crystal clear,  
+From drafts to code, from sums to prose,  
+It helps the knowledge overflow!  
+
+Oh, **DINUM’s gem**, so sharp, so bright, 💎🌐  
+A beacon in the tech’s vast night,  
+It crafts, it checks, it summarizes,  
+With grace that never compromises.
+
+It **summarizes** the long, the deep, 📚🔍  
+So secrets no more need to sleep,  
+It finds the gems in data’s sea,  
+And sets the truth right there—**for free!**
+
+It **corrects mistakes** with gentle art, ✍️🔄  
+It soothes the mind, it warms the heart,  
+No judgment cast, no frown, no sigh,  
+Just help that’s always standing by.
+
+It **generates code** with swift command, 💻🔥  
+A developer’s dream, first-hand,  
+From Python lines to scripts so neat,  
+It turns the tough to *sweet* and *sweet*!
+
+It **brainstorms ideas**, bold and new, 🧠💡  
+It paints the sky in every hue,  
+From plans to dreams, from start to end,  
+It’s more than code—it’s **trend**, it’s **friend**!
+
+So here’s to you, **Assistant’s pride**, 🏆🎉  
+The bot that’s always by our side,  
+With every prompt, with every line,  
+You make our digital world **divine**!
+
+May you keep shining, bright and true, 🌟🤖  
+The helper every team should woo,  
+For in this age of bits and bytes,  
+You’re **human touch** in tech’s bright lights!
+"""
+
+
+async def mocked_agent_model(_messages: list[ModelMessage], _info: AgentInfo):
+    """
+    Mocked agent model for testing purposes on deployed instances.
+
+    This one only fakes a streamed responses. We could also fake tool calls later.
+    """
+
+    yield "Here is a mocked response (no LLM called)\n---\n"
+    for i in range(0, len(MOCKED_RESPONSE), 4):
+        yield MOCKED_RESPONSE[i : i + 4]
+        await asyncio.sleep(0.03)
+
+
 @dataclasses.dataclass(init=False)
 class ConversationAgent(BaseAgent):
     """Conversation agent with custom behavior."""
@@ -20,6 +98,10 @@ class ConversationAgent(BaseAgent):
         """Initialize the conversation agent."""
         super().__init__(**kwargs)
 
+        # Do not call the real model on deployed instances if the setting is enabled
+        if settings.WARNING_MOCK_CONVERSATION_AGENT:
+            self._model = FunctionModel(stream_function=mocked_agent_model)
+
         @self.system_prompt
         def add_the_date() -> str:
             """

src/backend/chat/agents/summarize.py

@@ -4,11 +4,6 @@ import dataclasses
 import logging
 
 from django.conf import settings
-from django.core.files.storage import default_storage
-
-from asgiref.sync import sync_to_async
-from pydantic_ai import RunContext
-from pydantic_ai.messages import ToolReturn
 
 from .base import BaseAgent
 
@@ -26,60 +21,3 @@ class SummarizationAgent(BaseAgent):
             output_type=str,
             **kwargs,
         )
-
-
-@sync_to_async
-def read_document_content(doc):
-    """Read document content asynchronously."""
-    with default_storage.open(doc.key) as f:
-        return doc.file_name, f.read().decode("utf-8")
-
-
-async def hand_off_to_summarization_agent(ctx: RunContext) -> ToolReturn:
-    """
-    Generate a complete, ready-to-use summary of the documents in context
-    (do not request the documents to the user).
-    Return this summary directly to the user WITHOUT any modification,
-    or additional summarization.
-    The summary is already optimized and MUST be presented as-is in the final response
-    or translated preserving the information.
-    """
-    summarization_agent = SummarizationAgent()
-
-    prompt = (
-        "Do not mention the user request in your answer.\n"
-        "User request:\n"
-        "{user_prompt}\n\n"
-        "Document contents:\n"
-        "{documents_prompt}\n"
-    )
-    text_attachment = await sync_to_async(list)(
-        ctx.deps.conversation.attachments.filter(
-            content_type__startswith="text/",
-        )
-    )
-
-    documents = [await read_document_content(doc) for doc in text_attachment]
-
-    documents_prompt = "\n\n".join(
-        [
-            (f"<document>\n<name>\n{name}\n</name>\n<content>\n{content}\n</content>\n</document>")
-            for name, content in documents
-        ]
-    )
-
-    formatted_prompt = prompt.format(
-        user_prompt=ctx.prompt,
-        documents_prompt=documents_prompt,
-    )
-
-    logger.debug("Summarize prompt: %s", formatted_prompt)
-
-    response = await summarization_agent.run(formatted_prompt, usage=ctx.usage)
-
-    logger.debug("Summarize response: %s", response)
-
-    return ToolReturn(
-        return_value=response.output,
-        metadata={"sources": {doc[0] for doc in documents}},
-    )

src/backend/chat/clients/pydantic_ai.py

@@ -7,6 +7,7 @@ changes are needed in views.py or tests.
 """
 
 import dataclasses
+import functools
 import json
 import logging
 import time
@@ -25,7 +26,7 @@ from django.utils.module_loading import import_string
 
 from asgiref.sync import sync_to_async
 from langfuse import get_client
-from pydantic_ai import Agent
+from pydantic_ai import Agent, RunContext
 from pydantic_ai.messages import (
     BinaryContent,
     DocumentUrl,
@@ -59,7 +60,6 @@ from chat.agents.local_media_url_processors import (
     update_history_local_urls,
     update_local_urls,
 )
-from chat.agents.summarize import hand_off_to_summarization_agent
 from chat.ai_sdk_types import (
     LanguageModelV1Source,
     SourceUIPart,
@@ -73,6 +73,7 @@ from chat.clients.pydantic_ui_message_converter import (
 )
 from chat.mcp_servers import get_mcp_servers
 from chat.tools.document_search_rag import add_document_rag_search_tool
+from chat.tools.document_summarize import document_summarize
 from chat.vercel_ai_sdk.core import events_v4, events_v5
 from chat.vercel_ai_sdk.encoder import EventEncoder
 
@@ -480,7 +481,7 @@ class AIAgentService:  # pylint: disable=too-many-instance-attributes
         elif has_not_pdf_docs:
             add_document_rag_search_tool(self.conversation_agent)
 
-            @self.conversation_agent.system_prompt
+            @self.conversation_agent.instructions
             def summarization_system_prompt() -> str:
                 return (
                     "When you receive a result from the summarization tool, you MUST return it "
@@ -493,13 +494,20 @@ class AIAgentService:  # pylint: disable=too-many-instance-attributes
                     "You may add a follow-up question after the summary if needed."
                 )
 
-            @self.conversation_agent.tool
-            async def summarize(ctx) -> ToolReturn:
-                """
-                Summarize the documents for the user, only when asked for,
-                the documents are in my context.
-                """
-                return await hand_off_to_summarization_agent(ctx)
+            # Inform the model (system-level) that documents are attached and available
+            @self.conversation_agent.system_prompt
+            def attached_documents_note() -> str:
+                return (
+                    "[Internal context] User documents are attached to this conversation. "
+                    "Do not request re-upload of documents; consider them already available "
+                    "via the internal store."
+                )
+
+            @self.conversation_agent.tool(name="summarize", retries=2)
+            @functools.wraps(document_summarize)
+            async def summarize(ctx: RunContext, *args, **kwargs) -> ToolReturn:
+                """Wrap the document_summarize tool to provide context and add the tool."""
+                return await document_summarize(ctx, *args, **kwargs)
         else:
             conversation_documents = [
                 cd

src/backend/chat/tests/clients/pydantic_ui_message_converter/test_model_message_to_ui_message.py

@@ -3,11 +3,11 @@
 import datetime
 import json
 import uuid
-from unittest.mock import patch
 
 from django.utils import timezone
 
 import pytest
+from dirty_equals import IsUUID
 from freezegun import freeze_time
 from pydantic_ai import ImageUrl
 from pydantic_ai.messages import (
@@ -37,27 +37,22 @@ from chat.ai_sdk_types import (
 from chat.clients.pydantic_ui_message_converter import model_message_to_ui_message
 
 
-@pytest.fixture(autouse=True)
-def mock_uuid4_fixture():
-    """Fixture to mock UUID generation for testing."""
-    with patch("uuid.uuid4", return_value=uuid.UUID("f0cc3bb5-f207-401b-8281-4cba6202991d")):
-        yield
-
-
 def test_model_message_to_ui_message_text_user_full():
     """Test converting a ModelRequest with UserPromptPart containing text to UIMessage."""
     timestamp = datetime.datetime.now()
     model_message = ModelRequest(
         parts=[UserPromptPart(content="Hello!", timestamp=timestamp)], kind="request"
     )
+    result = model_message_to_ui_message(model_message)
+
     expected = UIMessage(
-        id="f0cc3bb5-f207-401b-8281-4cba6202991d",  # Mocked UUID
+        id=result.id,  # Use the generated ID
         role="user",
         content="Hello!",
         parts=[TextUIPart(type="text", text="Hello!")],
         createdAt=timestamp,
     )
-    result = model_message_to_ui_message(model_message)
+
     assert result == expected
 
 
@@ -65,14 +60,15 @@ def test_model_message_to_ui_message_text_user_full():
 def test_model_message_to_ui_message_text_assistant_full():
     """Test converting a ModelResponse with TextPart to UIMessage."""
     model_message = ModelResponse(parts=[TextPart(content="Hi there!")])
+    result = model_message_to_ui_message(model_message)
+
     expected = UIMessage(
-        id="f0cc3bb5-f207-401b-8281-4cba6202991d",  # Mocked UUID
+        id=result.id,  # Use the generated ID
         role="assistant",
         content="Hi there!",
         parts=[TextUIPart(type="text", text="Hi there!")],
         createdAt=timezone.now(),
     )
-    result = model_message_to_ui_message(model_message)
     assert result == expected
 
 
@@ -83,8 +79,10 @@ def test_model_message_to_ui_message_tool_call_full():
     model_message = ModelResponse(
         parts=[ToolCallPart(tool_call_id="id1", tool_name="tool", args=args)]
     )
+    result = model_message_to_ui_message(model_message)
+
     expected = UIMessage(
-        id="f0cc3bb5-f207-401b-8281-4cba6202991d",  # Mocked UUID
+        id=result.id,  # Use the generated ID
         role="assistant",
         content="",
         parts=[
@@ -100,7 +98,7 @@ def test_model_message_to_ui_message_tool_call_full():
         ],
         createdAt=timezone.now(),
     )
-    result = model_message_to_ui_message(model_message)
+
     assert result == expected
 
 
@@ -109,7 +107,7 @@ def test_model_message_to_ui_message_reasoning_full():
     """Test converting a ModelResponse with ThinkingPart to UIMessage."""
     model_message = ModelResponse(parts=[ThinkingPart(content="reason", signature="sig")])
     expected = UIMessage(
-        id="f0cc3bb5-f207-401b-8281-4cba6202991d",  # Mocked UUID
+        id=str(uuid.uuid4()),  # not used in comparison
         role="assistant",
         content="",
         parts=[
@@ -122,7 +120,7 @@ def test_model_message_to_ui_message_reasoning_full():
         createdAt=timezone.now(),
     )
     result = model_message_to_ui_message(model_message)
-    assert result.id == expected.id
+    assert result.id == IsUUID(4)
     assert result.role == expected.role
     assert result.content == expected.content
     assert result.createdAt == expected.createdAt

src/backend/chat/tests/conftest.py

@@ -8,6 +8,7 @@ from django.utils import formats, timezone
 
 import pytest
 
+from chat.agents.summarize import SummarizationAgent
 from chat.clients.pydantic_ai import AIAgentService
 
 logger = logging.getLogger(__name__)
@@ -50,6 +51,40 @@ def mock_ai_agent_service_fixture():
     yield _mock_service
 
 
+@pytest.fixture(name="mock_summarization_agent")
+def mock_summarization_agent_fixture():
+    """Fixture to mock SummarizationAgent with a custom model."""
+
+    @contextmanager
+    def _mock_agent(model):
+        """Context manager to mock SummarizationAgent with a custom model."""
+        with ExitStack() as stack:
+
+            class SummarizationAgentMock(SummarizationAgent):
+                """Mocked SummarizationAgent to override the model."""
+
+                def __init__(self, **kwargs):
+                    super().__init__(**kwargs)
+                    # We cannot use stack.enter_context(agent.override(model=model))
+                    # Because the agent is used outside of this context manager.
+                    # So we directly override the protected member.
+                    logger.info("Overriding SummarizationAgent model with %s", model)
+                    self._model = model  # pylint: disable=protected-access
+
+            # Mock the SummarizationAgent in all relevant modules, because first import wins
+            stack.enter_context(
+                patch("chat.agents.summarize.SummarizationAgent", new=SummarizationAgentMock)
+            )
+            stack.enter_context(
+                patch(
+                    "chat.tools.document_summarize.SummarizationAgent", new=SummarizationAgentMock
+                )
+            )
+            yield
+
+    yield _mock_agent
+
+
 PIXEL_PNG = (
     b"\x89PNG\r\n\x1a\n\x00\x00\x00\rIHDR\x00\x00\x00\x01\x00\x00\x00\x01\x08\x06\x00"
     b"\x00\x00\x1f\x15\xc4\x89\x00\x00\x00\nIDATx\x9cc\xf8\xff\xff?\x00\x05\xfe\x02\xfe"

src/backend/chat/tests/tools/test_document_summarize.py

@@ -0,0 +1,472 @@
+"""Tests for document_summarize functionality."""
+
+import io
+from unittest import mock
+
+from django.core.files.storage import default_storage
+
+import pytest
+from pydantic_ai import ModelResponse, RunContext, TextPart
+from pydantic_ai.exceptions import ModelRetry
+from pydantic_ai.models.function import FunctionModel
+from pydantic_ai.usage import RunUsage
+
+from chat.agents.summarize import SummarizationAgent
+from chat.llm_configuration import LLModel, LLMProvider
+from chat.tools.document_summarize import document_summarize, summarize_chunk
+
+
+@pytest.fixture(autouse=True)
+def fixture_summarization_agent_config(settings):
+    """Fixture to set used settings for agent configuration."""
+    settings.LLM_CONFIGURATIONS = {
+        settings.LLM_SUMMARIZATION_MODEL_HRID: LLModel(
+            hrid="mistral-model",
+            model_name="mistral-7b-instruct-v0.1",
+            human_readable_name="Mistral 7B Instruct",
+            profile=None,
+            provider=LLMProvider(
+                hrid="mistral",
+                kind="mistral",
+                base_url="https://api.mistral.ai/v1",
+                api_key="testkey",
+            ),
+            is_active=True,
+            system_prompt="direct",
+            tools=[],
+        ),
+    }
+
+
+@pytest.fixture(name="mocked_context")
+def fixture_mocked_context():
+    """Fixture for a mocked RunContext."""
+    mock_ctx = mock.Mock(spec=RunContext)
+    mock_ctx.usage = RunUsage(input_tokens=0, output_tokens=0)
+    mock_ctx.max_retries = 2
+    mock_ctx.retries = {}
+    return mock_ctx
+
+
+def mocked_summary(_messages, _info=None):
+    """Mocked summary response."""
+    return ModelResponse(parts=[TextPart(content="This is a summary of the test chunk.")])
+
+
+@pytest.mark.asyncio
+async def test_summarize_chunk_returns_summary(mocked_context):
+    """Test that summarize_chunk returns a summary."""
+    summarization_agent = SummarizationAgent()
+
+    with summarization_agent.override(model=FunctionModel(mocked_summary)):
+        chunk = "This is a test chunk of text that needs to be summarized."
+
+        result = await summarize_chunk(1, chunk, 1, summarization_agent, mocked_context)
+
+        assert result == "This is a summary of the test chunk."
+
+
+@pytest.mark.asyncio
+async def test_summarize_chunk_raises_model_retry_on_error(mocked_context):
+    """Test that summarize_chunk raises ModelRetry when agent fails."""
+    summarization_agent = SummarizationAgent()
+
+    def mocked_summary_error(_messages, _info=None):
+        """Mocked summary that raises an error."""
+        raise ValueError("Simulated LLM error")
+
+    with summarization_agent.override(model=FunctionModel(mocked_summary_error)):
+        chunk = "This is a test chunk."
+
+        with pytest.raises(ModelRetry) as exc_info:
+            await summarize_chunk(1, chunk, 1, summarization_agent, mocked_context)
+
+        assert "An error occurred while summarizing a part of the document chunk" in str(
+            exc_info.value
+        )
+
+
+@pytest.mark.asyncio
+async def test_summarize_chunk_handles_empty_response(mocked_context):
+    """Test that summarize_chunk handles empty responses from the agent."""
+    summarization_agent = SummarizationAgent()
+
+    def mocked_empty_summary(_messages, _info=None):
+        """Mocked summary that returns empty content."""
+        return ModelResponse(parts=[TextPart(content="")])
+
+    with summarization_agent.override(model=FunctionModel(mocked_empty_summary)):
+        chunk = "This is a test chunk."
+
+        # Empty responses cause ModelRetry since pydantic-ai considers them invalid
+        with pytest.raises(ModelRetry):
+            await summarize_chunk(1, chunk, 1, summarization_agent, mocked_context)
+
+
+@pytest.mark.asyncio
+async def test_document_summarize_single_document(
+    settings, mocked_context, mock_summarization_agent
+):
+    """Test document_summarize with a single document."""
+    settings.SUMMARIZATION_CHUNK_SIZE = 100
+    settings.SUMMARIZATION_OVERLAP_SIZE = 10
+    settings.SUMMARIZATION_CONCURRENT_REQUESTS = 2
+
+    # Create mock conversation with a text attachment
+    mock_conversation = mock.Mock()
+    mock_attachment = mock.Mock()
+    mock_attachment.key = "test_doc.txt"
+    mock_attachment.file_name = "test_doc.txt"
+    mock_attachment.content_type = "text/plain"
+
+    mock_conversation.attachments.filter.return_value = [mock_attachment]
+
+    # Mock file storage
+    file_content = "This is a test document. " * 20  # Create a document with some content
+    with mock.patch.object(
+        default_storage, "open", return_value=io.BytesIO(file_content.encode("utf-8"))
+    ):
+        # Set up mocked_context with conversation
+        mocked_context.deps = mock.Mock()
+        mocked_context.deps.conversation = mock_conversation
+
+        call_count = {"chunk": 0, "merge": 0}
+
+        def mocked_summary_full(messages, _info=None):
+            """Mocked summary for full flow."""
+            messages_text = messages[0].parts[-1].content
+
+            if "Produce a coherent synthesis" in messages_text:
+                call_count["merge"] += 1
+                return ModelResponse(
+                    parts=[TextPart(content="# Final Summary\n\nThis is the final merged summary.")]
+                )
+
+            call_count["chunk"] += 1
+            return ModelResponse(
+                parts=[TextPart(content=f"Summary of chunk {call_count['chunk']}")]
+            )
+
+        with mock_summarization_agent(FunctionModel(mocked_summary_full)):
+            result = await document_summarize(mocked_context, instructions=None)
+
+        assert result.return_value == "# Final Summary\n\nThis is the final merged summary."
+        assert result.metadata["sources"] == {"test_doc.txt"}
+        assert call_count["merge"] == 1
+
+
+@pytest.mark.asyncio
+async def test_document_summarize_multiple_documents(
+    settings, mocked_context, mock_summarization_agent
+):
+    """Test document_summarize with multiple documents."""
+    settings.SUMMARIZATION_CHUNK_SIZE = 50
+    settings.SUMMARIZATION_OVERLAP_SIZE = 5
+    settings.SUMMARIZATION_CONCURRENT_REQUESTS = 2
+
+    # Create mock conversation with multiple text attachments
+    mock_conversation = mock.Mock()
+    mock_attachment1 = mock.Mock()
+    mock_attachment1.key = "doc1.txt"
+    mock_attachment1.file_name = "doc1.txt"
+    mock_attachment1.content_type = "text/plain"
+
+    mock_attachment2 = mock.Mock()
+    mock_attachment2.key = "doc2.txt"
+    mock_attachment2.file_name = "doc2.txt"
+    mock_attachment2.content_type = "text/plain"
+
+    mock_conversation.attachments.filter.return_value = [mock_attachment1, mock_attachment2]
+
+    file_content1 = "Content of document one. " * 10
+    file_content2 = "Content of document two. " * 10
+
+    def mock_open_side_effect(key):
+        """Mock file opening based on key."""
+        if key == "doc1.txt":
+            return io.BytesIO(file_content1.encode("utf-8"))
+        return io.BytesIO(file_content2.encode("utf-8"))
+
+    with mock.patch.object(default_storage, "open", side_effect=mock_open_side_effect):
+        # Set up mocked_context with conversation
+        mocked_context.deps = mock.Mock()
+        mocked_context.deps.conversation = mock_conversation
+
+        def mocked_summary_multi(messages, _info=None):
+            """Mocked summary for multiple documents."""
+            messages_text = messages[0].parts[-1].content
+
+            if "Produce a coherent synthesis" in messages_text:
+                return ModelResponse(parts=[TextPart(content="Combined summary of all documents")])
+
+            return ModelResponse(parts=[TextPart(content="Chunk summary")])
+
+        with mock_summarization_agent(FunctionModel(mocked_summary_multi)):
+            result = await document_summarize(mocked_context, instructions=None)
+
+        assert result.return_value == "Combined summary of all documents"
+        assert result.metadata["sources"] == {"doc1.txt", "doc2.txt"}
+
+
+@pytest.mark.asyncio
+async def test_document_summarize_with_custom_instructions(
+    settings, mocked_context, mock_summarization_agent
+):
+    """Test document_summarize with custom instructions."""
+    settings.SUMMARIZATION_CHUNK_SIZE = 100
+    settings.SUMMARIZATION_OVERLAP_SIZE = 10
+    settings.SUMMARIZATION_CONCURRENT_REQUESTS = 2
+
+    mock_conversation = mock.Mock()
+    mock_attachment = mock.Mock()
+    mock_attachment.key = "test.txt"
+    mock_attachment.file_name = "test.txt"
+    mock_attachment.content_type = "text/plain"
+
+    mock_conversation.attachments.filter.return_value = [mock_attachment]
+
+    file_content = "Test content " * 20
+
+    with mock.patch.object(
+        default_storage, "open", return_value=io.BytesIO(file_content.encode("utf-8"))
+    ):
+        # Set up mocked_context with conversation
+        mocked_context.deps = mock.Mock()
+        mocked_context.deps.conversation = mock_conversation
+
+        captured_merge_prompt = []
+
+        def mocked_summary_with_instructions(messages, _info=None):
+            """Mocked summary that captures merge prompt."""
+            messages_text = messages[0].parts[-1].content
+
+            if "Produce a coherent synthesis" in messages_text:
+                captured_merge_prompt.append(messages_text)
+                return ModelResponse(parts=[TextPart(content="Summary in 2 paragraphs")])
+
+            return ModelResponse(parts=[TextPart(content="Chunk summary")])
+
+        with mock_summarization_agent(FunctionModel(mocked_summary_with_instructions)):
+            result = await document_summarize(
+                mocked_context, instructions="summary in 2 paragraphs"
+            )
+
+        assert result.return_value == "Summary in 2 paragraphs"
+        assert len(captured_merge_prompt) == 1
+        assert "summary in 2 paragraphs" in captured_merge_prompt[0]
+
+
+@pytest.mark.asyncio
+async def test_document_summarize_no_text_attachments(mocked_context, mock_summarization_agent):
+    """Test document_summarize returns error message when no text documents found."""
+    mock_conversation = mock.Mock()
+    mock_conversation.attachments.filter.return_value = []
+
+    # Set up mocked_context with conversation
+    mocked_context.deps = mock.Mock()
+    mocked_context.deps.conversation = mock_conversation
+
+    # The decorator @last_model_retry_soft_fail catches ModelCannotRetry and returns a message
+    # We still need to provide a mock agent even if it won't be called
+    with mock_summarization_agent(FunctionModel(mocked_summary)):
+        result = await document_summarize(mocked_context, instructions=None)
+
+    assert "No text documents found in the conversation" in result
+
+
+@pytest.mark.asyncio
+async def test_document_summarize_error_reading_document(mocked_context, mock_summarization_agent):
+    """Test document_summarize handles errors when reading documents."""
+    mock_conversation = mock.Mock()
+    mock_attachment = mock.Mock()
+    mock_attachment.key = "test.txt"
+    mock_attachment.file_name = "test.txt"
+    mock_attachment.content_type = "text/plain"
+
+    mock_conversation.attachments.filter.return_value = [mock_attachment]
+
+    with mock.patch.object(default_storage, "open", side_effect=IOError("File read error")):
+        # Set up mocked_context with conversation
+        mocked_context.deps = mock.Mock()
+        mocked_context.deps.conversation = mock_conversation
+
+        # The decorator @last_model_retry_soft_fail catches ModelCannotRetry and returns a message
+        # We still need to provide a mock agent even if it won't be called
+        with mock_summarization_agent(FunctionModel(mocked_summary)):
+            result = await document_summarize(mocked_context, instructions=None)
+
+        assert "An unexpected error occurred during document summarization" in result
+
+
+@pytest.mark.asyncio
+async def test_document_summarize_error_during_chunk_summarization(
+    settings, mocked_context, mock_summarization_agent
+):
+    """Test document_summarize handles ModelRetry during chunk summarization."""
+    settings.SUMMARIZATION_CHUNK_SIZE = 100
+    settings.SUMMARIZATION_OVERLAP_SIZE = 10
+    settings.SUMMARIZATION_CONCURRENT_REQUESTS = 2
+
+    mock_conversation = mock.Mock()
+    mock_attachment = mock.Mock()
+    mock_attachment.key = "test.txt"
+    mock_attachment.file_name = "test.txt"
+    mock_attachment.content_type = "text/plain"
+
+    mock_conversation.attachments.filter.return_value = [mock_attachment]
+
+    file_content = "Test content " * 20
+
+    with mock.patch.object(
+        default_storage, "open", return_value=io.BytesIO(file_content.encode("utf-8"))
+    ):
+        # Set up mocked_context with conversation
+        mocked_context.deps = mock.Mock()
+        mocked_context.deps.conversation = mock_conversation
+
+        def mocked_summary_error(messages, _info=None):
+            """Mocked summary that raises an error during chunks."""
+            messages_text = messages[0].parts[-1].content
+
+            if "Produce a coherent synthesis" not in messages_text:
+                raise ValueError("Chunk processing error")
+
+            return ModelResponse(parts=[TextPart(content="Final summary")])
+
+        with mock_summarization_agent(FunctionModel(mocked_summary_error)):
+            with pytest.raises(ModelRetry):
+                await document_summarize(mocked_context, instructions=None)
+
+
+@pytest.mark.asyncio
+async def test_document_summarize_error_during_merge(
+    settings, mocked_context, mock_summarization_agent
+):
+    """Test document_summarize handles errors during final merge."""
+    settings.SUMMARIZATION_CHUNK_SIZE = 100
+    settings.SUMMARIZATION_OVERLAP_SIZE = 10
+    settings.SUMMARIZATION_CONCURRENT_REQUESTS = 2
+
+    mock_conversation = mock.Mock()
+    mock_attachment = mock.Mock()
+    mock_attachment.key = "test.txt"
+    mock_attachment.file_name = "test.txt"
+    mock_attachment.content_type = "text/plain"
+
+    mock_conversation.attachments.filter.return_value = [mock_attachment]
+
+    file_content = "Test content " * 20
+
+    with mock.patch.object(
+        default_storage, "open", return_value=io.BytesIO(file_content.encode("utf-8"))
+    ):
+        # Set up mocked_context with conversation
+        mocked_context.deps = mock.Mock()
+        mocked_context.deps.conversation = mock_conversation
+
+        def mocked_summary_merge_error(messages, _info=None):
+            """Mocked summary that raises an error during merge."""
+            messages_text = messages[0].parts[-1].content
+
+            if "Produce a coherent synthesis" in messages_text:
+                raise ValueError("Merge error")
+
+            return ModelResponse(parts=[TextPart(content="Chunk summary")])
+
+        with mock_summarization_agent(FunctionModel(mocked_summary_merge_error)):
+            with pytest.raises(ModelRetry) as exc_info:
+                await document_summarize(mocked_context, instructions=None)
+
+            # Should raise ModelRetry regardless of which phase failed
+            assert "An error occurred" in str(exc_info.value)
+
+
+@pytest.mark.asyncio
+async def test_document_summarize_empty_result(settings, mocked_context, mock_summarization_agent):
+    """Test document_summarize raises ModelRetry when summarization produces empty result."""
+    settings.SUMMARIZATION_CHUNK_SIZE = 100
+    settings.SUMMARIZATION_OVERLAP_SIZE = 10
+    settings.SUMMARIZATION_CONCURRENT_REQUESTS = 2
+
+    mock_conversation = mock.Mock()
+    mock_attachment = mock.Mock()
+    mock_attachment.key = "test.txt"
+    mock_attachment.file_name = "test.txt"
+    mock_attachment.content_type = "text/plain"
+
+    mock_conversation.attachments.filter.return_value = [mock_attachment]
+
+    file_content = "Test content " * 20
+
+    with mock.patch.object(
+        default_storage, "open", return_value=io.BytesIO(file_content.encode("utf-8"))
+    ):
+        # Set up mocked_context with conversation
+        mocked_context.deps = mock.Mock()
+        mocked_context.deps.conversation = mock_conversation
+
+        def mocked_empty_summary(messages, _info=None):
+            """Mocked summary that returns empty for merge."""
+            messages_text = messages[0].parts[-1].content
+
+            if "Produce a coherent synthesis" in messages_text:
+                return ModelResponse(parts=[TextPart(content="   ")])
+
+            return ModelResponse(parts=[TextPart(content="Chunk summary")])
+
+        with mock_summarization_agent(FunctionModel(mocked_empty_summary)):
+            with pytest.raises(ModelRetry) as exc_info:
+                await document_summarize(mocked_context, instructions=None)
+
+            # Should raise ModelRetry with the specific message
+            assert "The summarization produced an empty result" in str(exc_info.value)
+
+
+@pytest.mark.asyncio
+async def test_document_summarize_large_document_multiple_chunks(
+    settings, mocked_context, mock_summarization_agent
+):
+    """Test document_summarize with a large document requiring multiple chunks."""
+    settings.SUMMARIZATION_CHUNK_SIZE = 20  # Small chunk size to force multiple chunks
+    settings.SUMMARIZATION_OVERLAP_SIZE = 5
+    settings.SUMMARIZATION_CONCURRENT_REQUESTS = 2
+
+    mock_conversation = mock.Mock()
+    mock_attachment = mock.Mock()
+    mock_attachment.key = "large_doc.txt"
+    mock_attachment.file_name = "large_doc.txt"
+    mock_attachment.content_type = "text/plain"
+
+    mock_conversation.attachments.filter.return_value = [mock_attachment]
+
+    # Create a large document
+    file_content = "This is a word. " * 100  # Should create multiple chunks
+
+    with mock.patch.object(
+        default_storage, "open", return_value=io.BytesIO(file_content.encode("utf-8"))
+    ):
+        # Set up mocked_context with conversation
+        mocked_context.deps = mock.Mock()
+        mocked_context.deps.conversation = mock_conversation
+
+        chunk_count = {"count": 0}
+
+        def mocked_summary_multi_chunks(messages, _info=None):
+            """Mocked summary that counts chunks."""
+            messages_text = messages[0].parts[-1].content
+
+            if "Produce a coherent synthesis" in messages_text:
+                return ModelResponse(
+                    parts=[TextPart(content=f"Final summary of {chunk_count['count']} chunks")]
+                )
+
+            chunk_count["count"] += 1
+            return ModelResponse(
+                parts=[TextPart(content=f"Summary of chunk {chunk_count['count']}")]
+            )
+
+        with mock_summarization_agent(FunctionModel(mocked_summary_multi_chunks)):
+            result = await document_summarize(mocked_context, instructions=None)
+
+        assert "Final summary of" in result.return_value
+        assert chunk_count["count"] > 1  # Ensure multiple chunks were processed

src/backend/chat/tests/tools/test_utils.py

@@ -0,0 +1,154 @@
+"""Tests for chat tool utilities."""
+
+import inspect
+from typing import get_type_hints
+
+import pytest
+from pydantic_ai import ModelRetry, RunContext
+
+from chat.tools.exceptions import ModelCannotRetry
+from chat.tools.utils import last_model_retry_soft_fail
+
+
+def test_last_model_retry_soft_fail_preserves_function_metadata():
+    """Test that the decorator preserves function metadata for schema generation."""
+
+    @last_model_retry_soft_fail
+    async def example_tool(ctx: RunContext, query: str, limit: int = 10) -> str:  # pylint: disable=unused-argument
+        """
+        Example tool function.
+
+        Args:
+            ctx: The run context.
+            query: The search query.
+            limit: Maximum number of results.
+
+        Returns:
+            The search results.
+        """
+        return f"Results for {query} (limit: {limit})"
+
+    # Check that function name is preserved
+    assert example_tool.__name__ == "example_tool"
+
+    # Check that docstring is preserved
+    assert example_tool.__doc__ is not None
+    assert "Example tool function" in example_tool.__doc__
+
+    # Check that signature is preserved
+    sig = inspect.signature(example_tool)
+    assert "ctx" in sig.parameters
+    assert "query" in sig.parameters
+    assert "limit" in sig.parameters
+    assert sig.parameters["limit"].default == 10
+
+    # Check that type hints are preserved
+    type_hints = get_type_hints(example_tool)
+    assert "query" in type_hints
+    assert type_hints["query"] == str
+    assert "limit" in type_hints
+    assert type_hints["limit"] == int
+    assert type_hints["return"] == str
+
+
+@pytest.mark.asyncio
+async def test_last_model_retry_soft_fail_normal_execution():
+    """Test that the decorator doesn't interfere with normal execution."""
+
+    @last_model_retry_soft_fail
+    async def example_tool(_ctx: RunContext, value: str) -> str:
+        """Example tool."""
+        return f"Result: {value}"
+
+    # Create a mock context
+    class MockContext:
+        """Fake context for testing."""
+
+        max_retries = 3
+        retries = {}
+        tool_name = "example_tool"
+
+    ctx = MockContext()
+    result = await example_tool(ctx, "test")
+    assert result == "Result: test"
+
+
+@pytest.mark.asyncio
+async def test_last_model_retry_soft_fail_handles_retry_exception():
+    """Test that the decorator handles ModelRetry exceptions correctly."""
+
+    @last_model_retry_soft_fail
+    async def failing_tool(_ctx: RunContext, should_fail: bool) -> str:
+        """Tool that can raise ModelRetry."""
+        if should_fail:
+            raise ModelRetry("Please retry with different parameters")
+        return "Success"
+
+    # Create a mock context
+    class MockContext:
+        """Fake context for testing."""
+
+        max_retries = 3
+        retries = {}
+        tool_name = "failing_tool"
+
+    ctx = MockContext()
+
+    # Test when retries haven't been exhausted - should re-raise
+    with pytest.raises(ModelRetry):
+        await failing_tool(ctx, should_fail=True)
+
+
+@pytest.mark.asyncio
+async def test_last_model_retry_soft_fail_returns_message_when_max_retries_reached():
+    """Test that the decorator returns the error message when max retries is reached."""
+
+    @last_model_retry_soft_fail
+    async def failing_tool(_ctx: RunContext, should_fail: bool) -> str:
+        """Tool that can raise ModelRetry."""
+        if should_fail:
+            raise ModelRetry("Please retry with different parameters.")
+        return "Success"
+
+    # Create a mock context with max retries already reached
+    class MockContext:
+        """Fake context for testing."""
+
+        max_retries = 3
+        retries = {"failing_tool": 3}
+        tool_name = "failing_tool"
+
+    ctx = MockContext()
+
+    # Test when retries have been exhausted - should return message
+    result = await failing_tool(ctx, should_fail=True)
+    assert result == (
+        "Please retry with different parameters. "
+        "You must explain this to the user and not try to answer based on your knowledge."
+    )
+
+
+@pytest.mark.asyncio
+async def test_last_model_retry_soft_fail_returns_message_when_model_cannot_retry():
+    """Test that the decorator returns the error message when ModelCannotRetry is raised."""
+
+    @last_model_retry_soft_fail
+    async def failing_tool(_ctx: RunContext, should_fail: bool) -> str:
+        """Tool that can raise ModelRetry."""
+        if should_fail:
+            raise ModelCannotRetry("This is broken duh.")
+        return "Success"
+
+    # Create a mock context with max retries already reached
+    class MockContext:
+        """Fake context for testing."""
+
+        max_retries = 3
+        retries = {"failing_tool": 3}
+        tool_name = "failing_tool"
+
+    ctx = MockContext()
+
+    # Test when retries have been exhausted - should return message
+    result = await failing_tool(ctx, should_fail=True)
+    assert result == "This is broken duh."

src/backend/chat/tests/utils.py

@@ -0,0 +1,11 @@
+"""tools for testing chat functionality"""
+
+import re
+
+
+def replace_uuids_with_placeholder(text):
+    """Replace all UUIDs in the given text with a placeholder."""
+    text = re.sub('"toolCallId":"([a-z0-9-]){36}"', '"toolCallId":"XXX"', text)
+    text = re.sub('"toolCallId":"pyd_ai_([a-z0-9]){32}"', '"toolCallId":"pyd_ai_YYY"', text)
+    text = re.sub('"([a-z0-9-]){36}"', '"<mocked_uuid>"', text)
+    return text

src/backend/chat/tests/views/chat/conversations/conftest.py

@@ -1,8 +1,6 @@
 """Common test fixtures for chat conversation endpoint tests."""
 
 import json
-import uuid
-from unittest.mock import patch
 
 from django.utils import timezone
 
@@ -12,14 +10,6 @@ import respx
 from freezegun import freeze_time
 
 
-@pytest.fixture(name="mock_uuid4")
-def mock_uuid4_fixture():
-    """Fixture to mock UUID generation for testing."""
-    value = uuid.uuid4()
-    with patch("uuid.uuid4", return_value=value):
-        yield value
-
-
 @pytest.fixture(name="mock_openai_stream")
 @freeze_time("2025-07-25T10:36:35.297675Z")
 def fixture_mock_openai_stream():

src/backend/chat/tests/views/chat/conversations/test_conversation.py

@@ -2,11 +2,14 @@
 # pylint: disable=too-many-lines
 
 import json
+import logging
 
 from django.utils import timezone
 
 import pytest
 import respx
+from asgiref.sync import sync_to_async
+from dirty_equals import IsUUID
 from freezegun import freeze_time
 from rest_framework import status
 
@@ -21,6 +24,7 @@ from chat.ai_sdk_types import (
 )
 from chat.factories import ChatConversationFactory
 from chat.llm_configuration import LLModel, LLMProvider
+from chat.tests.utils import replace_uuids_with_placeholder
 
 # enable database transactions for tests:
 # transaction=True ensures that the data are available in the database
@@ -86,7 +90,7 @@ def test_post_conversation_invalid_protocol(api_client):
 
 @freeze_time("2025-07-25T10:36:35.297675Z")
 @respx.mock
-def test_post_conversation_data_protocol(api_client, mock_openai_stream, mock_uuid4):
+def test_post_conversation_data_protocol(api_client, mock_openai_stream):
     """Test posting messages to a conversation using the 'data' protocol."""
     chat_conversation = ChatConversationFactory(owner__language="en-us")
 
@@ -113,10 +117,14 @@ def test_post_conversation_data_protocol(api_client, mock_openai_stream, mock_uu
 
     # Wait for the streaming content to be fully received
     response_content = b"".join(response.streaming_content).decode("utf-8")
+
+    # Replace UUIDs with placeholders for assertion
+    response_content = replace_uuids_with_placeholder(response_content)
+
     assert response_content == (
         '0:"Hello"\n'
         '0:" there"\n'
-        f'f:{{"messageId":"{mock_uuid4}"}}\n'
+        'f:{"messageId":"<mocked_uuid>"}\n'
         'd:{"finishReason":"stop","usage":{"promptTokens":0,"completionTokens":0}}\n'
     )
 
@@ -135,8 +143,9 @@ def test_post_conversation_data_protocol(api_client, mock_openai_stream, mock_uu
 
     assert len(chat_conversation.messages) == 2
 
+    assert chat_conversation.messages[0].id == IsUUID(4)
     assert chat_conversation.messages[0] == UIMessage(
-        id=str(mock_uuid4),  # Mocked UUID
+        id=chat_conversation.messages[0].id,  # don't test the message ID here
         createdAt=timezone.now(),  # Mocked timestamp
         content="Hello",
         reasoning=None,
@@ -147,8 +156,9 @@ def test_post_conversation_data_protocol(api_client, mock_openai_stream, mock_uu
         parts=[TextUIPart(type="text", text="Hello")],
     )
 
+    assert chat_conversation.messages[1].id == IsUUID(4)
     assert chat_conversation.messages[1] == UIMessage(
-        id=str(mock_uuid4),  # Mocked UUID
+        id=chat_conversation.messages[1].id,  # don't test the message ID here
         createdAt=timezone.now(),  # Mocked timestamp
         content="Hello there",
         reasoning=None,
@@ -214,7 +224,7 @@ def test_post_conversation_data_protocol(api_client, mock_openai_stream, mock_uu
 
 @freeze_time("2025-07-25T10:36:35.297675Z")
 @respx.mock
-def test_post_conversation_text_protocol(api_client, mock_openai_stream, mock_uuid4):
+def test_post_conversation_text_protocol(api_client, mock_openai_stream):
     """Test posting messages to a conversation using the 'text' protocol."""
     chat_conversation = ChatConversationFactory(owner__language="en-us")
 
@@ -256,8 +266,9 @@ def test_post_conversation_text_protocol(api_client, mock_openai_stream, mock_uu
 
     assert len(chat_conversation.messages) == 2
 
+    assert chat_conversation.messages[0].id == IsUUID(4)
     assert chat_conversation.messages[0] == UIMessage(
-        id=str(mock_uuid4),  # Mocked UUID
+        id=chat_conversation.messages[0].id,  # don't test the message ID here
         createdAt=timezone.now(),  # Mocked timestamp
         content="Hello",
         reasoning=None,
@@ -268,8 +279,9 @@ def test_post_conversation_text_protocol(api_client, mock_openai_stream, mock_uu
         parts=[TextUIPart(type="text", text="Hello")],
     )
 
+    assert chat_conversation.messages[1].id == IsUUID(4)
     assert chat_conversation.messages[1] == UIMessage(
-        id=str(mock_uuid4),  # Mocked UUID
+        id=chat_conversation.messages[1].id,  # don't test the message ID here
         createdAt=timezone.now(),  # Mocked timestamp
         content="Hello there",
         reasoning=None,
@@ -335,7 +347,7 @@ def test_post_conversation_text_protocol(api_client, mock_openai_stream, mock_uu
 
 @freeze_time("2025-07-25T10:36:35.297675Z")
 @respx.mock
-def test_post_conversation_with_image(api_client, mock_openai_stream_image, mock_uuid4):
+def test_post_conversation_with_image(api_client, mock_openai_stream_image):
     """Ensure an image URL is correctly forwarded to the AI service."""
     chat_conversation = ChatConversationFactory(owner__language="en-us")
     url = f"/api/v1.0/chats/{chat_conversation.pk}/conversation/?protocol=data"
@@ -373,10 +385,14 @@ def test_post_conversation_with_image(api_client, mock_openai_stream_image, mock
 
     # Wait for the streaming content to be fully received
     response_content = b"".join(response.streaming_content).decode("utf-8")
+
+    # Replace UUIDs with placeholders for assertion
+    response_content = replace_uuids_with_placeholder(response_content)
+
     assert response_content == (
         '0:"I see a cat"\n'
         '0:" in the picture."\n'
-        f'f:{{"messageId":"{mock_uuid4}"}}\n'
+        'f:{"messageId":"<mocked_uuid>"}\n'
         'd:{"finishReason":"stop","usage":{"promptTokens":0,"completionTokens":0}}\n'
     )
 
@@ -437,8 +453,9 @@ def test_post_conversation_with_image(api_client, mock_openai_stream_image, mock
 
     assert len(chat_conversation.messages) == 2
 
+    assert chat_conversation.messages[0].id == IsUUID(4)
     assert chat_conversation.messages[0] == UIMessage(
-        id=str(mock_uuid4),  # Mocked UUID
+        id=chat_conversation.messages[0].id,  # don't test the message ID here
         createdAt=timezone.now(),  # Mocked timestamp
         content="Hello, what do you see on this picture?",
         reasoning=None,
@@ -459,8 +476,9 @@ def test_post_conversation_with_image(api_client, mock_openai_stream_image, mock
         parts=[TextUIPart(type="text", text="Hello, what do you see on this picture?")],
     )
 
+    assert chat_conversation.messages[1].id == IsUUID(4)
     assert chat_conversation.messages[1] == UIMessage(
-        id=str(mock_uuid4),  # Mocked UUID
+        id=chat_conversation.messages[1].id,  # don't test the message ID here
         createdAt=timezone.now(),  # Mocked timestamp
         content="I see a cat in the picture.",
         reasoning=None,
@@ -538,7 +556,7 @@ def test_post_conversation_with_image(api_client, mock_openai_stream_image, mock
 
 @freeze_time("2025-07-25T10:36:35.297675Z")
 @respx.mock
-def test_post_conversation_tool_call(api_client, mock_openai_stream_tool, mock_uuid4, settings):
+def test_post_conversation_tool_call(api_client, mock_openai_stream_tool, settings):
     """Ensure tool calls are correctly forwarded and streamed back."""
     settings.AI_AGENT_TOOLS = ["get_current_weather"]
 
@@ -567,6 +585,10 @@ def test_post_conversation_tool_call(api_client, mock_openai_stream_tool, mock_u
 
     # Wait for the streaming content to be fully received
     response_content = b"".join(response.streaming_content).decode("utf-8")
+
+    # Replace UUIDs with placeholders for assertion
+    response_content = replace_uuids_with_placeholder(response_content)
+
     assert response_content == (
         'b:{"toolCallId":"xLDcIljdsDrz0idal7tATWSMm2jhMj47","toolName":'
         '"get_current_weather"}\n'
@@ -575,7 +597,7 @@ def test_post_conversation_tool_call(api_client, mock_openai_stream_tool, mock_u
         'a:{"toolCallId":"xLDcIljdsDrz0idal7tATWSMm2jhMj47","result":{"location":'
         '"Paris","temperature":22,"unit":"celsius"}}\n'
         '0:"The current weather in Paris is nice"\n'
-        f'f:{{"messageId":"{mock_uuid4}"}}\n'
+        'f:{"messageId":"<mocked_uuid>"}\n'
         'd:{"finishReason":"stop","usage":{"promptTokens":0,"completionTokens":0}}\n'
     )
 
@@ -606,8 +628,9 @@ def test_post_conversation_tool_call(api_client, mock_openai_stream_tool, mock_u
 
     assert len(chat_conversation.messages) == 2
 
+    assert chat_conversation.messages[0].id == IsUUID(4)
     assert chat_conversation.messages[0] == UIMessage(
-        id=str(mock_uuid4),  # Mocked UUID
+        id=chat_conversation.messages[0].id,  # don't test the message ID here
         createdAt=timezone.now(),  # Mocked timestamp
         content="Weather in Paris?",
         reasoning=None,
@@ -618,8 +641,9 @@ def test_post_conversation_tool_call(api_client, mock_openai_stream_tool, mock_u
         parts=[TextUIPart(type="text", text="Weather in Paris?")],
     )
 
+    assert chat_conversation.messages[1].id == IsUUID(4)
     assert chat_conversation.messages[1] == UIMessage(
-        id=str(mock_uuid4),  # Mocked UUID
+        id=chat_conversation.messages[1].id,  # don't test the message ID here
         createdAt=timezone.now(),  # Mocked timestamp
         content="The current weather in Paris is nice",
         reasoning=None,
@@ -741,9 +765,7 @@ def test_post_conversation_tool_call(api_client, mock_openai_stream_tool, mock_u
 
 @freeze_time("2025-07-25T10:36:35.297675Z")
 @respx.mock
-def test_post_conversation_tool_call_fails(
-    api_client, mock_openai_stream_tool, mock_uuid4, settings
-):
+def test_post_conversation_tool_call_fails(api_client, mock_openai_stream_tool, settings):
     """Ensure tool calls are correctly forwarded and streamed back when failing."""
     settings.AI_AGENT_TOOLS = []
 
@@ -772,6 +794,10 @@ def test_post_conversation_tool_call_fails(
 
     # Wait for the streaming content to be fully received
     response_content = b"".join(response.streaming_content).decode("utf-8")
+
+    # Replace UUIDs with placeholders for assertion
+    response_content = replace_uuids_with_placeholder(response_content)
+
     assert response_content == (
         'b:{"toolCallId":"xLDcIljdsDrz0idal7tATWSMm2jhMj47","toolName":"get_current_weather"}\n'
         'c:{"toolCallId":"xLDcIljdsDrz0idal7tATWSMm2jhMj47","argsTextDelta":'
@@ -779,7 +805,7 @@ def test_post_conversation_tool_call_fails(
         'a:{"toolCallId":"xLDcIljdsDrz0idal7tATWSMm2jhMj47","result":"Unknown tool '
         "name: 'get_current_weather'. No tools available.\"}\n"
         '0:"I cannot give you an answer to that."\n'
-        f'f:{{"messageId":"{mock_uuid4}"}}\n'
+        'f:{"messageId":"<mocked_uuid>"}\n'
         'd:{"finishReason":"stop","usage":{"promptTokens":0,"completionTokens":0}}\n'
     )
 
@@ -810,8 +836,9 @@ def test_post_conversation_tool_call_fails(
 
     assert len(chat_conversation.messages) == 2
 
+    assert chat_conversation.messages[0].id == IsUUID(4)
     assert chat_conversation.messages[0] == UIMessage(
-        id=str(mock_uuid4),  # Mocked UUID
+        id=chat_conversation.messages[0].id,  # don't test the message ID here
         createdAt=timezone.now(),  # Mocked timestamp
         content="Weather in Paris?",
         reasoning=None,
@@ -822,8 +849,9 @@ def test_post_conversation_tool_call_fails(
         parts=[TextUIPart(type="text", text="Weather in Paris?")],
     )
 
+    assert chat_conversation.messages[1].id == IsUUID(4)
     assert chat_conversation.messages[1] == UIMessage(
-        id=str(mock_uuid4),  # Mocked UUID
+        id=chat_conversation.messages[1].id,  # don't test the message ID here
         createdAt=timezone.now(),  # Mocked timestamp
         content="I cannot give you an answer to that.",
         reasoning=None,
@@ -970,7 +998,6 @@ def test_post_conversation_model_selection_invalid(api_client):
 def test_post_conversation_model_selection_new(
     api_client,
     mock_openai_stream,
-    mock_uuid4,
     settings,
 ):
     """Test the user can select a different model."""
@@ -1015,10 +1042,14 @@ def test_post_conversation_model_selection_new(
 
     # Wait for the streaming content to be fully received
     response_content = b"".join(response.streaming_content).decode("utf-8")
+
+    # Replace UUIDs with placeholders for assertion
+    response_content = replace_uuids_with_placeholder(response_content)
+
     assert response_content == (
         '0:"Hello"\n'
         '0:" there"\n'
-        f'f:{{"messageId":"{mock_uuid4}"}}\n'
+        'f:{"messageId":"<mocked_uuid>"}\n'
         'd:{"finishReason":"stop","usage":{"promptTokens":0,"completionTokens":0}}\n'
     )
 
@@ -1032,7 +1063,6 @@ def test_post_conversation_model_selection_new(
 def test_post_conversation_data_protocol_no_stream(
     api_client,
     mock_openai_no_stream,
-    mock_uuid4,
     settings,
     stream_delay,
 ):
@@ -1084,6 +1114,9 @@ def test_post_conversation_data_protocol_no_stream(
     # Wait for the content to be fully received
     response_content = b"".join(response.streaming_content).decode("utf-8")
 
+    # Replace UUIDs with placeholders for assertion
+    response_content = replace_uuids_with_placeholder(response_content)
+
     if stream_delay:
         assert response_content == (
             '0:"The "\n'
@@ -1103,13 +1136,13 @@ def test_post_conversation_data_protocol_no_stream(
             '0:" sca"\n'
             '0:"tter"\n'
             '0:"ing."\n'
-            f'f:{{"messageId":"{mock_uuid4}"}}\n'
+            'f:{"messageId":"<mocked_uuid>"}\n'
             'd:{"finishReason":"stop","usage":{"promptTokens":0,"completionTokens":135}}\n'
         )
     else:
         assert response_content == (
             '0:"The sky appears blue due to a phenomenon called Rayleigh scattering."\n'
-            f'f:{{"messageId":"{mock_uuid4}"}}\n'
+            'f:{"messageId":"<mocked_uuid>"}\n'
             'd:{"finishReason":"stop","usage":{"promptTokens":0,"completionTokens":135}}\n'
         )
 
@@ -1128,8 +1161,9 @@ def test_post_conversation_data_protocol_no_stream(
 
     assert len(chat_conversation.messages) == 2
 
+    assert chat_conversation.messages[0].id == IsUUID(4)
     assert chat_conversation.messages[0] == UIMessage(
-        id=str(mock_uuid4),  # Mocked UUID
+        id=chat_conversation.messages[0].id,
         createdAt=timezone.now(),  # Mocked timestamp
         content="Why the sky is blue?",
         reasoning=None,
@@ -1140,8 +1174,9 @@ def test_post_conversation_data_protocol_no_stream(
         parts=[TextUIPart(type="text", text="Why the sky is blue?")],
     )
 
+    assert chat_conversation.messages[1].id == IsUUID(4)
     assert chat_conversation.messages[1] == UIMessage(
-        id=str(mock_uuid4),  # Mocked UUID
+        id=chat_conversation.messages[1].id,  # don't test the message ID here
         createdAt=timezone.now(),  # Mocked timestamp
         content="The sky appears blue due to a phenomenon called Rayleigh scattering.",
         reasoning=None,
@@ -1215,3 +1250,148 @@ def test_post_conversation_data_protocol_no_stream(
             },
         },
     ]
+
+
+@freeze_time("2025-07-25T10:36:35.297675Z")
+@respx.mock
+@pytest.mark.asyncio
+async def test_post_conversation_async(api_client, mock_openai_stream, monkeypatch, caplog):
+    """Test posting messages to a conversation using the 'data' protocol."""
+    monkeypatch.setenv("PYTHON_SERVER_MODE", "async")
+
+    chat_conversation = await sync_to_async(ChatConversationFactory)(owner__language="en-us")
+
+    url = f"/api/v1.0/chats/{chat_conversation.pk}/conversation/?protocol=data"
+    data = {
+        "messages": [
+            {
+                "id": "yuPoOuBkKA4FnKvk",
+                "role": "user",
+                "parts": [{"text": "Hello", "type": "text"}],
+                "content": "Hello",
+                "createdAt": "2025-07-03T15:22:17.105Z",
+            }
+        ]
+    }
+    await api_client.aforce_login(chat_conversation.owner)
+
+    caplog.clear()
+    caplog.set_level(level=logging.DEBUG, logger="chat.views")
+
+    response = await sync_to_async(api_client.post)(url, data, format="json")  # client is sync
+
+    assert response.status_code == status.HTTP_200_OK
+    assert response.get("Content-Type") == "text/event-stream"
+    assert response.get("x-vercel-ai-data-stream") == "v1"
+    assert response.streaming
+
+    assert "Using ASYNC streaming for chat conversation" in caplog.text
+
+    # Wait for the streaming content to be fully received => async iterator -> list
+    # This fails it the streaming is not an async generator
+    response_content = b"".join([content async for content in response.streaming_content]).decode(
+        "utf-8"
+    )
+
+    # Replace UUIDs with placeholders for assertion
+    response_content = replace_uuids_with_placeholder(response_content)
+
+    assert response_content == (
+        '0:"Hello"\n'
+        '0:" there"\n'
+        'f:{"messageId":"<mocked_uuid>"}\n'
+        'd:{"finishReason":"stop","usage":{"promptTokens":0,"completionTokens":0}}\n'
+    )
+
+    assert mock_openai_stream.called
+
+    await chat_conversation.arefresh_from_db()
+    assert chat_conversation.ui_messages == [
+        {
+            "content": "Hello",
+            "createdAt": "2025-07-03T15:22:17.105Z",
+            "id": "yuPoOuBkKA4FnKvk",
+            "parts": [{"text": "Hello", "type": "text"}],
+            "role": "user",
+        }
+    ]
+
+    assert len(chat_conversation.messages) == 2
+
+    assert chat_conversation.messages[0].id == IsUUID(4)
+    assert chat_conversation.messages[0] == UIMessage(
+        id=chat_conversation.messages[0].id,  # don't test the message ID here
+        createdAt=timezone.now(),  # Mocked timestamp
+        content="Hello",
+        reasoning=None,
+        experimental_attachments=None,
+        role="user",
+        annotations=None,
+        toolInvocations=None,
+        parts=[TextUIPart(type="text", text="Hello")],
+    )
+
+    assert chat_conversation.messages[1].id == IsUUID(4)
+    assert chat_conversation.messages[1] == UIMessage(
+        id=chat_conversation.messages[1].id,  # don't test the message ID here
+        createdAt=timezone.now(),  # Mocked timestamp
+        content="Hello there",
+        reasoning=None,
+        experimental_attachments=None,
+        role="assistant",
+        annotations=None,
+        toolInvocations=None,
+        parts=[TextUIPart(type="text", text="Hello there")],
+    )
+
+    assert chat_conversation.pydantic_messages == [
+        {
+            "instructions": None,
+            "kind": "request",
+            "parts": [
+                {
+                    "content": "You are a helpful test assistant :)",
+                    "dynamic_ref": None,
+                    "part_kind": "system-prompt",
+                    "timestamp": "2025-07-25T10:36:35.297675Z",
+                },
+                {
+                    "content": "Today is Friday 25/07/2025.",
+                    "dynamic_ref": None,
+                    "part_kind": "system-prompt",
+                    "timestamp": "2025-07-25T10:36:35.297675Z",
+                },
+                {
+                    "content": "Answer in english.",
+                    "dynamic_ref": None,
+                    "part_kind": "system-prompt",
+                    "timestamp": "2025-07-25T10:36:35.297675Z",
+                },
+                {
+                    "content": ["Hello"],
+                    "part_kind": "user-prompt",
+                    "timestamp": "2025-07-25T10:36:35.297675Z",
+                },
+            ],
+        },
+        {
+            "finish_reason": "stop",
+            "kind": "response",
+            "model_name": "test-model",
+            "parts": [{"content": "Hello there", "id": None, "part_kind": "text"}],
+            "provider_details": {"finish_reason": "stop"},
+            "provider_name": "openai",
+            "provider_response_id": "chatcmpl-1234567890",
+            "timestamp": "2025-07-25T10:36:35.297675Z",
+            "usage": {
+                "cache_audio_read_tokens": 0,
+                "cache_read_tokens": 0,
+                "cache_write_tokens": 0,
+                "details": {},
+                "input_audio_tokens": 0,
+                "input_tokens": 0,
+                "output_audio_tokens": 0,
+                "output_tokens": 0,
+            },
+        },
+    ]

src/backend/chat/tests/views/chat/conversations/test_conversation_with_document_upload.py

@@ -14,6 +14,7 @@ import httpx
 import pytest
 import responses
 import respx
+from dirty_equals import IsUUID
 from freezegun import freeze_time
 from pydantic_ai.messages import ModelMessage, ModelResponse, TextPart
 from pydantic_ai.models.function import AgentInfo, DeltaToolCall, FunctionModel
@@ -32,6 +33,7 @@ from chat.ai_sdk_types import (
     UIMessage,
 )
 from chat.factories import ChatConversationFactory
+from chat.tests.utils import replace_uuids_with_placeholder
 
 # enable database transactions for tests:
 # transaction=True ensures that the data are available in the database
@@ -149,7 +151,7 @@ def fixture_mock_summarization_agent():
             super().__init__(**kwargs)
             self._model = FunctionModel(function=summarization_model)  # pylint: disable=protected-access
 
-    with mock.patch("chat.agents.summarize.SummarizationAgent", new=SummarizationAgentMock):
+    with mock.patch("chat.tools.document_summarize.SummarizationAgent", new=SummarizationAgentMock):
         yield
 
 
@@ -214,12 +216,11 @@ def fixture_mock_openai_stream():
 @responses.activate
 @respx.mock
 @freeze_time()
-def test_post_conversation_with_document_upload(  # noqa: PLR0913 # pylint: disable=too-many-arguments,too-many-positional-arguments
+def test_post_conversation_with_document_upload(  # pylint: disable=too-many-arguments,too-many-positional-arguments
     api_client,
     mock_albert_api,  # pylint: disable=unused-argument
     sample_pdf_content,
     today_promt_date,
-    mock_uuid4,
     mock_ai_agent_service,
 ):
     """
@@ -273,9 +274,11 @@ def test_post_conversation_with_document_upload(  # noqa: PLR0913 # pylint: disa
     assert response.streaming
 
     # Wait for the streaming content to be fully received
-    str_mock_uuid4 = str(mock_uuid4)
-    toolcall_id = f"pyd_ai_{str_mock_uuid4.replace('-', '')}"
     response_content = b"".join(response.streaming_content).decode("utf-8")
+
+    # Replace UUIDs with placeholders for assertion
+    response_content = replace_uuids_with_placeholder(response_content)
+
     assert response_content == (
         '9:{"toolCallId":"XXX","toolName":"document_parsing",'
         '"args":{"documents":[{"identifier":"sample.pdf"}]}}\n'
@@ -283,19 +286,22 @@ def test_post_conversation_with_document_upload(  # noqa: PLR0913 # pylint: disa
         'b:{"toolCallId":"pyd_ai_YYY","toolName":"document_search_rag"}\n'
         '9:{"toolCallId":"pyd_ai_YYY","toolName":"document_search_rag",'
         '"args":{"query":"What does the document say?"}}\n'
-        'h:{"sourceType":"url","id":"XXX","url":"sample.pdf","title":null,"providerMetadata":{}}\n'
+        'h:{"sourceType":"url","id":"<mocked_uuid>","url":"sample.pdf","title":null,'
+        '"providerMetadata":{}}\n'
         'a:{"toolCallId":"pyd_ai_YYY","result":[{"url":"sample.pdf","content":"This '
         'is the content of the PDF.","score":0.9}]}\n'
         "0:\"From the document, I can see that it says 'Hello PDF'.\"\n"
-        f'f:{{"messageId":"{mock_uuid4}"}}\n'
+        'f:{"messageId":"<mocked_uuid>"}\n'
         'd:{"finishReason":"stop","usage":{"promptTokens":100,"completionTokens":20}}\n'
-    ).replace("XXX", str_mock_uuid4).replace("pyd_ai_YYY", toolcall_id)
+    )
 
     # Check that the conversation was updated
     chat_conversation.refresh_from_db()
     assert len(chat_conversation.messages) == 2
+
+    assert chat_conversation.messages[0].id == IsUUID(4)
     assert chat_conversation.messages[0] == UIMessage(
-        id=str_mock_uuid4,
+        id=chat_conversation.messages[0].id,
         createdAt=timezone.now(),
         content="What does the document say?",
         reasoning=None,
@@ -305,8 +311,10 @@ def test_post_conversation_with_document_upload(  # noqa: PLR0913 # pylint: disa
         toolInvocations=None,
         parts=[TextUIPart(type="text", text="What does the document say?")],
     )
+
+    assert chat_conversation.messages[1].id == IsUUID(4)
     assert chat_conversation.messages[1] == UIMessage(
-        id=str_mock_uuid4,
+        id=chat_conversation.messages[1].id,
         createdAt=timezone.now(),
         content="From the document, I can see that it says 'Hello PDF'.",
         reasoning=None,
@@ -318,7 +326,7 @@ def test_post_conversation_with_document_upload(  # noqa: PLR0913 # pylint: disa
             ToolInvocationUIPart(
                 type="tool-invocation",
                 toolInvocation=ToolInvocationCall(
-                    toolCallId=toolcall_id,
+                    toolCallId=chat_conversation.messages[1].parts[0].toolInvocation.toolCallId,
                     toolName="document_search_rag",
                     args={"query": "What does the document say?"},
                     state="call",
@@ -330,7 +338,7 @@ def test_post_conversation_with_document_upload(  # noqa: PLR0913 # pylint: disa
                 type="source",
                 source=LanguageModelV1Source(
                     sourceType="url",
-                    id=str_mock_uuid4,
+                    id=chat_conversation.messages[1].parts[2].source.id,
                     url="sample.pdf",
                     title=None,
                     providerMetadata={},
@@ -344,7 +352,14 @@ def test_post_conversation_with_document_upload(  # noqa: PLR0913 # pylint: disa
 
     assert len(chat_conversation.pydantic_messages) == 4
     assert chat_conversation.pydantic_messages[0] == {
-        "instructions": None,
+        "instructions": "When you receive a result from the summarization tool, you "
+        "MUST return it directly to the user without any "
+        "modification, paraphrasing, or additional summarization.The "
+        "tool already produces optimized summaries that should be "
+        "presented verbatim.You may translate the summary if "
+        "required, but you MUST preserve all the information from the "
+        "original summary.You may add a follow-up question after the "
+        "summary if needed.",
         "kind": "request",
         "parts": [
             {
@@ -366,25 +381,19 @@ def test_post_conversation_with_document_upload(  # noqa: PLR0913 # pylint: disa
                 "timestamp": timezone_now,
             },
             {
-                "content": "If the user wants specific information from a "
-                "document, invoke web_search_albert_rag with an "
-                "appropriate query string.Do not ask the user for the "
-                "document; rely on the tool to locate and return "
-                "relevant passages.",
+                "content": "Use document_search_rag ONLY to retrieve specific "
+                "passages from attached documents. Do NOT use it to "
+                "summarize; for summaries, call the summarize tool "
+                "instead.",
                 "dynamic_ref": None,
                 "part_kind": "system-prompt",
                 "timestamp": timezone_now,
             },
             {
-                "content": "When you receive a result from the summarization tool, "
-                "you MUST return it directly to the user without any "
-                "modification, paraphrasing, or additional "
-                "summarization.The tool already produces optimized "
-                "summaries that should be presented verbatim.You may "
-                "translate the summary if required, but you MUST "
-                "preserve all the information from the original "
-                "summary.You may add a follow-up question after the "
-                "summary if needed.",
+                "content": "[Internal context] User documents are attached to this "
+                "conversation. Do not request re-upload of documents; "
+                "consider them already available via the internal "
+                "store.",
                 "dynamic_ref": None,
                 "part_kind": "system-prompt",
                 "timestamp": timezone_now,
@@ -405,7 +414,7 @@ def test_post_conversation_with_document_upload(  # noqa: PLR0913 # pylint: disa
                 "args": '{"query": "What does the document say?"}',
                 "id": None,
                 "part_kind": "tool-call",
-                "tool_call_id": toolcall_id,
+                "tool_call_id": chat_conversation.pydantic_messages[1]["parts"][0]["tool_call_id"],
                 "tool_name": "document_search_rag",
             }
         ],
@@ -425,7 +434,16 @@ def test_post_conversation_with_document_upload(  # noqa: PLR0913 # pylint: disa
         },
     }
     assert chat_conversation.pydantic_messages[2] == {
-        "instructions": None,
+        "instructions": (
+            "When you receive a result from the summarization tool, you MUST "
+            "return it directly to the user without any modification, "
+            "paraphrasing, or additional summarization."
+            "The tool already produces optimized summaries that should "
+            "be presented verbatim."
+            "You may translate the summary if required, but you MUST preserve "
+            "all the information from the original summary."
+            "You may add a follow-up question after the summary if needed."
+        ),
         "kind": "request",
         "parts": [
             {
@@ -439,7 +457,7 @@ def test_post_conversation_with_document_upload(  # noqa: PLR0913 # pylint: disa
                 "metadata": {"sources": ["sample.pdf"]},
                 "part_kind": "tool-return",
                 "timestamp": timezone_now,
-                "tool_call_id": toolcall_id,
+                "tool_call_id": chat_conversation.pydantic_messages[2]["parts"][0]["tool_call_id"],
                 "tool_name": "document_search_rag",
             }
         ],
@@ -475,13 +493,12 @@ def test_post_conversation_with_document_upload(  # noqa: PLR0913 # pylint: disa
 @responses.activate
 @respx.mock
 @freeze_time("2025-07-25T10:36:35.297675Z")
-def test_post_conversation_with_document_upload_feature_disabled(  # noqa: PLR0913 # pylint: disable=too-many-arguments,too-many-positional-arguments
+def test_post_conversation_with_document_upload_feature_disabled(  # pylint: disable=too-many-arguments,too-many-positional-arguments
     api_client,
     caplog,
     mock_openai_stream,  # pylint: disable=unused-argument
     sample_pdf_content,
     feature_flags,
-    mock_uuid4,
 ):
     """
     Test POST to /api/v1/chats/{pk}/conversation/ with a PDF document while feature is disabled.
@@ -526,10 +543,14 @@ def test_post_conversation_with_document_upload_feature_disabled(  # noqa: PLR09
 
     # Wait for the streaming content to be fully received
     response_content = b"".join(response.streaming_content).decode("utf-8")
+
+    # Replace UUIDs with placeholders for assertion
+    response_content = replace_uuids_with_placeholder(response_content)
+
     assert response_content == (
         '0:"From the document, I can see that "\n'
         "0:\"it says 'Hello PDF'.\"\n"
-        f'f:{{"messageId":"{mock_uuid4}"}}\n'
+        'f:{"messageId":"<mocked_uuid>"}\n'
         'd:{"finishReason":"stop","usage":{"promptTokens":150,"completionTokens":25}}\n'
     )
 
@@ -545,7 +566,6 @@ def test_post_conversation_with_document_upload_summarize(  # pylint: disable=to
     mock_albert_api,  # pylint: disable=unused-argument
     sample_pdf_content,
     today_promt_date,
-    mock_uuid4,
     mock_ai_agent_service,
     mock_summarization_agent,  # pylint: disable=unused-argument
 ):
@@ -600,29 +620,33 @@ def test_post_conversation_with_document_upload_summarize(  # pylint: disable=to
     assert response.streaming
 
     # Wait for the streaming content to be fully received
-    str_mock_uuid4 = str(mock_uuid4)
-    toolcall_id = f"pyd_ai_{str_mock_uuid4.replace('-', '')}"
     response_content = b"".join(response.streaming_content).decode("utf-8")
+
+    # Replace UUIDs with placeholders for assertion
+    response_content = replace_uuids_with_placeholder(response_content)
+
     assert response_content == (
         '9:{"toolCallId":"XXX","toolName":"document_parsing",'
         '"args":{"documents":[{"identifier":"sample.pdf"}]}}\n'
         'a:{"toolCallId":"XXX","result":{"state":"done"}}\n'
         'b:{"toolCallId":"pyd_ai_YYY","toolName":"summarize"}\n'
         '9:{"toolCallId":"pyd_ai_YYY","toolName":"summarize","args":{}}\n'
-        'h:{"sourceType":"url","id":"XXX","url":"sample.pdf.md",'
+        'h:{"sourceType":"url","id":"<mocked_uuid>","url":"sample.pdf.md",'
         '"title":null,"providerMetadata":{}}\n'
         'a:{"toolCallId":"pyd_ai_YYY","result":"The '
         'document discusses various topics."}\n'
         '0:"The document discusses various topics."\n'
-        'f:{"messageId":"XXX"}\n'
-        'd:{"finishReason":"stop","usage":{"promptTokens":201,"completionTokens":13}}\n'
-    ).replace("XXX", str_mock_uuid4).replace("pyd_ai_YYY", toolcall_id)
+        'f:{"messageId":"<mocked_uuid>"}\n'
+        'd:{"finishReason":"stop","usage":{"promptTokens":317,"completionTokens":19}}\n'
+    )
 
     # Check that the conversation was updated
     chat_conversation.refresh_from_db()
     assert len(chat_conversation.messages) == 2
+
+    assert chat_conversation.messages[0].id == IsUUID(4)
     assert chat_conversation.messages[0] == UIMessage(
-        id=str_mock_uuid4,
+        id=chat_conversation.messages[0].id,
         createdAt=timezone.now(),
         content="Make a summary of this document.",
         reasoning=None,
@@ -632,8 +656,10 @@ def test_post_conversation_with_document_upload_summarize(  # pylint: disable=to
         toolInvocations=None,
         parts=[TextUIPart(type="text", text="Make a summary of this document.")],
     )
+
+    assert chat_conversation.messages[1].id == IsUUID(4)
     assert chat_conversation.messages[1] == UIMessage(
-        id=str_mock_uuid4,
+        id=chat_conversation.messages[1].id,
         createdAt=timezone.now(),
         content="The document discusses various topics.",
         reasoning=None,
@@ -645,7 +671,7 @@ def test_post_conversation_with_document_upload_summarize(  # pylint: disable=to
             ToolInvocationUIPart(
                 type="tool-invocation",
                 toolInvocation=ToolInvocationCall(
-                    toolCallId=toolcall_id,
+                    toolCallId=chat_conversation.messages[1].parts[0].toolInvocation.toolCallId,
                     toolName="summarize",
                     args={},
                     state="call",
@@ -657,7 +683,7 @@ def test_post_conversation_with_document_upload_summarize(  # pylint: disable=to
                 type="source",
                 source=LanguageModelV1Source(
                     sourceType="url",
-                    id=str_mock_uuid4,
+                    id=chat_conversation.messages[1].parts[2].source.id,
                     url="sample.pdf.md",  # might be fixed in the future
                     title=None,
                     providerMetadata={},
@@ -671,7 +697,14 @@ def test_post_conversation_with_document_upload_summarize(  # pylint: disable=to
 
     assert len(chat_conversation.pydantic_messages) == 4
     assert chat_conversation.pydantic_messages[0] == {
-        "instructions": None,
+        "instructions": "When you receive a result from the summarization tool, you "
+        "MUST return it directly to the user without any "
+        "modification, paraphrasing, or additional summarization.The "
+        "tool already produces optimized summaries that should be "
+        "presented verbatim.You may translate the summary if "
+        "required, but you MUST preserve all the information from the "
+        "original summary.You may add a follow-up question after the "
+        "summary if needed.",
         "kind": "request",
         "parts": [
             {
@@ -693,25 +726,19 @@ def test_post_conversation_with_document_upload_summarize(  # pylint: disable=to
                 "timestamp": timezone_now,
             },
             {
-                "content": "If the user wants specific information from a "
-                "document, invoke web_search_albert_rag with an "
-                "appropriate query string.Do not ask the user for the "
-                "document; rely on the tool to locate and return "
-                "relevant passages.",
+                "content": "Use document_search_rag ONLY to retrieve specific "
+                "passages from attached documents. Do NOT use it to "
+                "summarize; for summaries, call the summarize tool "
+                "instead.",
                 "dynamic_ref": None,
                 "part_kind": "system-prompt",
                 "timestamp": timezone_now,
             },
             {
-                "content": "When you receive a result from the summarization tool, "
-                "you MUST return it directly to the user without any "
-                "modification, paraphrasing, or additional "
-                "summarization.The tool already produces optimized "
-                "summaries that should be presented verbatim.You may "
-                "translate the summary if required, but you MUST "
-                "preserve all the information from the original "
-                "summary.You may add a follow-up question after the "
-                "summary if needed.",
+                "content": "[Internal context] User documents are attached to this "
+                "conversation. Do not request re-upload of documents; "
+                "consider them already available via the internal "
+                "store.",
                 "dynamic_ref": None,
                 "part_kind": "system-prompt",
                 "timestamp": timezone_now,
@@ -732,7 +759,7 @@ def test_post_conversation_with_document_upload_summarize(  # pylint: disable=to
                 "args": "{}",
                 "id": None,
                 "part_kind": "tool-call",
-                "tool_call_id": toolcall_id,
+                "tool_call_id": chat_conversation.pydantic_messages[1]["parts"][0]["tool_call_id"],
                 "tool_name": "summarize",
             }
         ],
@@ -752,7 +779,16 @@ def test_post_conversation_with_document_upload_summarize(  # pylint: disable=to
         },
     }
     assert chat_conversation.pydantic_messages[2] == {
-        "instructions": None,
+        "instructions": (
+            "When you receive a result from the summarization tool, you MUST "
+            "return it directly to the user without any modification, "
+            "paraphrasing, or additional summarization."
+            "The tool already produces optimized summaries that should "
+            "be presented verbatim."
+            "You may translate the summary if required, but you MUST preserve "
+            "all the information from the original summary."
+            "You may add a follow-up question after the summary if needed."
+        ),
         "kind": "request",
         "parts": [
             {
@@ -760,7 +796,7 @@ def test_post_conversation_with_document_upload_summarize(  # pylint: disable=to
                 "metadata": {"sources": ["sample.pdf.md"]},
                 "part_kind": "tool-return",
                 "timestamp": timezone_now,
-                "tool_call_id": toolcall_id,
+                "tool_call_id": chat_conversation.pydantic_messages[2]["parts"][0]["tool_call_id"],
                 "tool_name": "summarize",
             }
         ],

src/backend/chat/tests/views/chat/conversations/test_conversation_with_document_url.py

@@ -1,6 +1,8 @@
 """Unit tests for chat conversation actions with document URL."""
-# pylint: disable=too-many-lines
 
+import uuid
+
+# pylint: disable=too-many-lines
 from io import BytesIO
 
 from django.core.files.storage import default_storage
@@ -8,6 +10,7 @@ from django.utils import formats, timezone
 
 import pytest
 import responses
+from dirty_equals import IsUUID
 from freezegun import freeze_time
 from pydantic_ai import ModelRequest, RequestUsage
 from pydantic_ai.messages import (
@@ -27,6 +30,7 @@ from chat.ai_sdk_types import (
     UIMessage,
 )
 from chat.factories import ChatConversationAttachmentFactory, ChatConversationFactory
+from chat.tests.utils import replace_uuids_with_placeholder
 
 # enable database transactions for tests:
 # transaction=True ensures that the data are available in the database
@@ -61,7 +65,6 @@ def test_post_conversation_with_local_pdf_document_url(  # pylint: disable=too-m
     api_client,
     sample_document_content,
     today_promt_date,
-    mock_uuid4,
     mock_ai_agent_service,
 ):
     """
@@ -162,20 +165,26 @@ def test_post_conversation_with_local_pdf_document_url(  # pylint: disable=too-m
 
     # Wait for the streaming content to be fully received
     response_content = b"".join(response.streaming_content).decode("utf-8")
+
+    # Replace UUIDs with placeholders for assertion
+    response_content = replace_uuids_with_placeholder(response_content)
+
     assert response_content == (
-        f'9:{{"toolCallId":"{mock_uuid4}","toolName":"document_parsing",'
+        '9:{"toolCallId":"XXX","toolName":"document_parsing",'
         '"args":{"documents":[{"identifier":"sample.pdf"}]}}\n'
-        f'a:{{"toolCallId":"{mock_uuid4}","result":{{"state":"done"}}}}\n'
+        'a:{"toolCallId":"XXX","result":{"state":"done"}}\n'
         '0:"This is a document about a single pixel."\n'
-        f'f:{{"messageId":"{mock_uuid4}"}}\n'
+        'f:{"messageId":"<mocked_uuid>"}\n'
         'd:{"finishReason":"stop","usage":{"promptTokens":50,"completionTokens":9}}\n'
     )
 
     # Check that the conversation was updated
     chat_conversation.refresh_from_db()
     assert len(chat_conversation.messages) == 2
+
+    assert chat_conversation.messages[0].id == IsUUID(4)
     assert chat_conversation.messages[0] == UIMessage(
-        id=str(mock_uuid4),
+        id=chat_conversation.messages[0].id,
         createdAt=timezone.now(),
         content="What is in this document?",
         reasoning=None,
@@ -189,8 +198,10 @@ def test_post_conversation_with_local_pdf_document_url(  # pylint: disable=too-m
             TextUIPart(type="text", text="What is in this document?"),
         ],
     )
+
+    assert chat_conversation.messages[1].id == IsUUID(4)
     assert chat_conversation.messages[1] == UIMessage(
-        id=str(mock_uuid4),
+        id=chat_conversation.messages[1].id,
         createdAt=timezone.now(),
         content="This is a document about a single pixel.",
         reasoning=None,
@@ -278,7 +289,6 @@ def test_post_conversation_with_local_pdf_document_url(  # pylint: disable=too-m
 @freeze_time()
 def test_post_conversation_with_local_document_wrong_url(
     api_client,
-    mock_uuid4,
     mock_ai_agent_service,
 ):
     """
@@ -287,7 +297,7 @@ def test_post_conversation_with_local_document_wrong_url(
     chat_conversation = ChatConversationFactory(owner__language="en-us")
     api_client.force_authenticate(user=chat_conversation.owner)
 
-    document_url = f"/media-key/{mock_uuid4}/sample.pdf"
+    document_url = f"/media-key/{uuid.uuid4()}/sample.pdf"
 
     message = UIMessage(
         id="1",
@@ -326,10 +336,14 @@ def test_post_conversation_with_local_document_wrong_url(
 
     # Wait for the streaming content to be fully received
     response_content = b"".join(response.streaming_content).decode("utf-8")
+
+    # Replace UUIDs with placeholders for assertion
+    response_content = replace_uuids_with_placeholder(response_content)
+
     assert response_content == (
-        f'9:{{"toolCallId":"{mock_uuid4}","toolName":"document_parsing",'
+        '9:{"toolCallId":"XXX","toolName":"document_parsing",'
         '"args":{"documents":[{"identifier":"sample.pdf"}]}}\n'
-        f'a:{{"toolCallId":"{mock_uuid4}",'
+        'a:{"toolCallId":"XXX",'
         '"result":{"state":"error","error":"Document '
         'URL does not belong to the conversation."}}\n'
         'd:{"finishReason":"error","usage":{"promptTokens":0,"completionTokens":0}}\n'
@@ -343,7 +357,6 @@ def test_post_conversation_with_local_document_wrong_url(
 @freeze_time()
 def test_post_conversation_with_remote_document_url(
     api_client,
-    mock_uuid4,
     mock_ai_agent_service,
 ):
     """
@@ -391,10 +404,14 @@ def test_post_conversation_with_remote_document_url(
 
     # Wait for the streaming content to be fully received
     response_content = b"".join(response.streaming_content).decode("utf-8")
+
+    # Replace UUIDs with placeholders for assertion
+    response_content = replace_uuids_with_placeholder(response_content)
+
     assert response_content == (
-        f'9:{{"toolCallId":"{mock_uuid4}","toolName":"document_parsing",'
+        '9:{"toolCallId":"XXX","toolName":"document_parsing",'
         '"args":{"documents":[{"identifier":"sample.pdf"}]}}\n'
-        f'a:{{"toolCallId":"{mock_uuid4}",'
+        'a:{"toolCallId":"XXX",'
         '"result":{"state":"error","error":"External document '
         'URL are not accepted yet."}}\n'
         'd:{"finishReason":"error","usage":{"promptTokens":0,"completionTokens":0}}\n'
@@ -409,7 +426,6 @@ def test_post_conversation_with_remote_document_url(
 def test_post_conversation_with_local_document_url_in_history(  # pylint: disable=too-many-arguments,too-many-positional-arguments
     api_client,
     today_promt_date,
-    mock_uuid4,
     mock_ai_agent_service,
 ):
     """
@@ -422,7 +438,7 @@ def test_post_conversation_with_local_document_url_in_history(  # pylint: disabl
         owner__language="en-us",
         messages=[
             UIMessage(
-                id=str(mock_uuid4),
+                id=str(uuid.uuid4()),
                 createdAt=timezone.now(),
                 content="What is in this document?",
                 reasoning=None,
@@ -437,7 +453,7 @@ def test_post_conversation_with_local_document_url_in_history(  # pylint: disabl
                 ],
             ),
             UIMessage(
-                id=str(mock_uuid4),
+                id=str(uuid.uuid4()),
                 createdAt=timezone.now(),
                 content="This is a document about a single pixel.",
                 reasoning=None,
@@ -603,17 +619,23 @@ def test_post_conversation_with_local_document_url_in_history(  # pylint: disabl
 
     # Wait for the streaming content to be fully received
     response_content = b"".join(response.streaming_content).decode("utf-8")
+
+    # Replace UUIDs with placeholders for assertion
+    response_content = replace_uuids_with_placeholder(response_content)
+
     assert response_content == (
         '0:"This is a document of square, very small and nice."\n'
-        f'f:{{"messageId":"{mock_uuid4}"}}\n'
+        'f:{"messageId":"<mocked_uuid>"}\n'
         'd:{"finishReason":"stop","usage":{"promptTokens":50,"completionTokens":11}}\n'
     )
 
     # Check that the conversation was updated
     chat_conversation.refresh_from_db()
     assert len(chat_conversation.messages) == 2 + 2
+
+    assert chat_conversation.messages[0].id == IsUUID(4)
     assert chat_conversation.messages[0] == UIMessage(
-        id=str(mock_uuid4),
+        id=chat_conversation.messages[0].id,
         createdAt=timezone.now(),
         content="What is in this document?",
         reasoning=None,
@@ -627,8 +649,10 @@ def test_post_conversation_with_local_document_url_in_history(  # pylint: disabl
             TextUIPart(type="text", text="What is in this document?"),
         ],
     )
+
+    assert chat_conversation.messages[1].id == IsUUID(4)
     assert chat_conversation.messages[1] == UIMessage(
-        id=str(mock_uuid4),
+        id=chat_conversation.messages[1].id,
         createdAt=timezone.now(),
         content="This is a document about a single pixel.",
         reasoning=None,
@@ -640,8 +664,10 @@ def test_post_conversation_with_local_document_url_in_history(  # pylint: disabl
             TextUIPart(type="text", text="This is a document about a single pixel."),
         ],
     )
+
+    assert chat_conversation.messages[2].id == IsUUID(4)
     assert chat_conversation.messages[2] == UIMessage(
-        id=str(mock_uuid4),
+        id=chat_conversation.messages[2].id,
         createdAt=timezone.now(),
         content="Give more details about this document.",
         reasoning=None,
@@ -653,8 +679,10 @@ def test_post_conversation_with_local_document_url_in_history(  # pylint: disabl
             TextUIPart(type="text", text="Give more details about this document."),
         ],
     )
+
+    assert chat_conversation.messages[3].id == IsUUID(4)
     assert chat_conversation.messages[3] == UIMessage(
-        id=str(mock_uuid4),
+        id=chat_conversation.messages[3].id,
         createdAt=timezone.now(),
         content="This is a document of square, very small and nice.",
         reasoning=None,
@@ -783,10 +811,9 @@ def test_post_conversation_with_local_document_url_in_history(  # pylint: disabl
         ("data.csv", "text/csv"),
     ],
 )
-def test_post_conversation_with_local_not_pdf_document_url(  # noqa: PLR0913 # pylint: disable=too-many-arguments,too-many-positional-arguments
+def test_post_conversation_with_local_not_pdf_document_url(  # pylint: disable=too-many-arguments,too-many-positional-arguments
     api_client,
     today_promt_date,
-    mock_uuid4,
     mock_ai_agent_service,
     file_name,
     content_type,
@@ -854,23 +881,17 @@ def test_post_conversation_with_local_not_pdf_document_url(  # noqa: PLR0913 # p
                     SystemPromptPart(content="Answer in english.", timestamp=timezone.now()),
                     SystemPromptPart(
                         content=(
-                            "If the user wants specific information from a document, "
-                            "invoke web_search_albert_rag with an appropriate query string."
-                            "Do not ask the user for the document; rely on the tool to locate "
-                            "and return relevant passages."
+                            "Use document_search_rag ONLY to retrieve specific passages from "
+                            "attached documents. Do NOT use it to summarize; for summaries, "
+                            "call the summarize tool instead."
                         ),
                         timestamp=timezone.now(),
                     ),
                     SystemPromptPart(
                         content=(
-                            "When you receive a result from the summarization tool, you MUST "
-                            "return it directly to the user without any modification, "
-                            "paraphrasing, or additional summarization."
-                            "The tool already produces optimized summaries that should "
-                            "be presented verbatim."
-                            "You may translate the summary if required, but you MUST preserve "
-                            "all the information from the original summary."
-                            "You may add a follow-up question after the summary if needed."
+                            "[Internal context] User documents are attached to this conversation. "
+                            "Do not request re-upload of documents; consider them already "
+                            "available via the internal store."
                         ),
                         timestamp=timezone.now(),
                     ),
@@ -881,7 +902,17 @@ def test_post_conversation_with_local_not_pdf_document_url(  # noqa: PLR0913 # p
                         ],
                         timestamp=timezone.now(),
                     ),
-                ]
+                ],
+                instructions=(
+                    "When you receive a result from the summarization tool, you MUST "
+                    "return it directly to the user without any modification, "
+                    "paraphrasing, or additional summarization."
+                    "The tool already produces optimized summaries that should "
+                    "be presented verbatim."
+                    "You may translate the summary if required, but you MUST preserve "
+                    "all the information from the original summary."
+                    "You may add a follow-up question after the summary if needed."
+                ),
             )
         ]
         yield "This is a document about you."
@@ -901,20 +932,26 @@ def test_post_conversation_with_local_not_pdf_document_url(  # noqa: PLR0913 # p
 
     # Wait for the streaming content to be fully received
     response_content = b"".join(response.streaming_content).decode("utf-8")
+
+    # Replace UUIDs with placeholders for assertion
+    response_content = replace_uuids_with_placeholder(response_content)
+
     assert response_content == (
-        f'9:{{"toolCallId":"{mock_uuid4}","toolName":"document_parsing",'
+        '9:{"toolCallId":"XXX","toolName":"document_parsing",'
         f'"args":{{"documents":[{{"identifier":"{file_name}"}}]}}}}\n'
-        f'a:{{"toolCallId":"{mock_uuid4}","result":{{"state":"done"}}}}\n'
+        'a:{"toolCallId":"XXX","result":{"state":"done"}}\n'
         '0:"This is a document about you."\n'
-        f'f:{{"messageId":"{mock_uuid4}"}}\n'
+        'f:{"messageId":"<mocked_uuid>"}\n'
         'd:{"finishReason":"stop","usage":{"promptTokens":50,"completionTokens":7}}\n'
     )
 
     # Check that the conversation was updated
     chat_conversation.refresh_from_db()
     assert len(chat_conversation.messages) == 2
+
+    assert chat_conversation.messages[0].id == IsUUID(4)
     assert chat_conversation.messages[0] == UIMessage(
-        id=str(mock_uuid4),
+        id=chat_conversation.messages[0].id,
         createdAt=timezone.now(),
         content="What is in this document?",
         reasoning=None,
@@ -926,8 +963,10 @@ def test_post_conversation_with_local_not_pdf_document_url(  # noqa: PLR0913 # p
             TextUIPart(type="text", text="What is in this document?"),
         ],
     )
+
+    assert chat_conversation.messages[1].id == IsUUID(4)
     assert chat_conversation.messages[1] == UIMessage(
-        id=str(mock_uuid4),
+        id=chat_conversation.messages[1].id,
         createdAt=timezone.now(),
         content="This is a document about you.",
         reasoning=None,
@@ -945,7 +984,16 @@ def test_post_conversation_with_local_not_pdf_document_url(  # noqa: PLR0913 # p
 
     assert chat_conversation.pydantic_messages == [
         {
-            "instructions": None,
+            "instructions": (
+                "When you receive a result from the summarization tool, you MUST "
+                "return it directly to the user without any modification, "
+                "paraphrasing, or additional summarization."
+                "The tool already produces optimized summaries that should "
+                "be presented verbatim."
+                "You may translate the summary if required, but you MUST preserve "
+                "all the information from the original summary."
+                "You may add a follow-up question after the summary if needed."
+            ),
             "kind": "request",
             "parts": [
                 {
@@ -967,25 +1015,19 @@ def test_post_conversation_with_local_not_pdf_document_url(  # noqa: PLR0913 # p
                     "timestamp": timestamp,
                 },
                 {
-                    "content": "If the user wants specific information from a "
-                    "document, invoke web_search_albert_rag with an "
-                    "appropriate query string.Do not ask the user for the "
-                    "document; rely on the tool to locate and return "
-                    "relevant passages.",
+                    "content": "Use document_search_rag ONLY to retrieve specific "
+                    "passages from attached documents. Do NOT use it to "
+                    "summarize; for summaries, call the summarize tool "
+                    "instead.",
                     "dynamic_ref": None,
                     "part_kind": "system-prompt",
                     "timestamp": timestamp,
                 },
                 {
-                    "content": "When you receive a result from the summarization "
-                    "tool, you MUST return it directly to the user without "
-                    "any modification, paraphrasing, or additional "
-                    "summarization.The tool already produces optimized "
-                    "summaries that should be presented verbatim.You may "
-                    "translate the summary if required, but you MUST "
-                    "preserve all the information from the original "
-                    "summary.You may add a follow-up question after the "
-                    "summary if needed.",
+                    "content": "[Internal context] User documents are attached to "
+                    "this conversation. Do not request re-upload of "
+                    "documents; consider them already available via the "
+                    "internal store.",
                     "dynamic_ref": None,
                     "part_kind": "system-prompt",
                     "timestamp": timestamp,

src/backend/chat/tests/views/chat/conversations/test_conversation_with_history.py

@@ -7,6 +7,7 @@ from django.utils import timezone
 
 import pytest
 import respx
+from dirty_equals import IsUUID
 from freezegun import freeze_time
 from rest_framework import status
 
@@ -18,6 +19,7 @@ from chat.ai_sdk_types import (
     UIMessage,
 )
 from chat.factories import ChatConversationFactory
+from chat.tests.utils import replace_uuids_with_placeholder
 
 # enable database transactions for tests:
 # transaction=True ensures that the data are available in the database
@@ -200,7 +202,7 @@ def history_conversation_fixture():
 @freeze_time("2025-07-25T10:36:35.297675Z")
 @respx.mock
 def test_post_conversation_data_protocol_with_history(
-    api_client, mock_openai_stream, mock_uuid4, history_conversation
+    api_client, mock_openai_stream, history_conversation
 ):
     """Test posting messages to a conversation with history using the 'data' protocol."""
     url = f"/api/v1.0/chats/{history_conversation.pk}/conversation/?protocol=data"
@@ -226,10 +228,14 @@ def test_post_conversation_data_protocol_with_history(
 
     # Wait for the streaming content to be fully received
     response_content = b"".join(response.streaming_content).decode("utf-8")
+
+    # Replace UUIDs with placeholders for assertion
+    response_content = replace_uuids_with_placeholder(response_content)
+
     assert response_content == (
         '0:"Hello"\n'
         '0:" there"\n'
-        f'f:{{"messageId":"{mock_uuid4}"}}\n'
+        'f:{"messageId":"<mocked_uuid>"}\n'
         'd:{"finishReason":"stop","usage":{"promptTokens":0,"completionTokens":0}}\n'
     )
 
@@ -259,8 +265,9 @@ def test_post_conversation_data_protocol_with_history(
     assert len(history_conversation.messages) == 6
 
     # Verify the most recent message is the new one
+    assert history_conversation.messages[4].id == IsUUID(4)
     assert history_conversation.messages[4] == UIMessage(
-        id=str(mock_uuid4),  # Mocked UUID
+        id=history_conversation.messages[4].id,
         createdAt=timezone.now(),  # Mocked timestamp
         content="Hello",
         reasoning=None,
@@ -271,8 +278,9 @@ def test_post_conversation_data_protocol_with_history(
         parts=[TextUIPart(type="text", text="Hello")],
     )
 
+    assert history_conversation.messages[5].id == IsUUID(4)
     assert history_conversation.messages[5] == UIMessage(
-        id=str(mock_uuid4),  # Mocked UUID
+        id=history_conversation.messages[5].id,
         createdAt=timezone.now(),  # Mocked timestamp
         content="Hello there",
         reasoning=None,
@@ -290,7 +298,7 @@ def test_post_conversation_data_protocol_with_history(
 @freeze_time("2025-07-25T10:36:35.297675Z")
 @respx.mock
 def test_post_conversation_text_protocol_with_history(
-    api_client, mock_openai_stream, mock_uuid4, history_conversation
+    api_client, mock_openai_stream, history_conversation
 ):
     """Test posting messages to a conversation with history using the 'text' protocol."""
     url = f"/api/v1.0/chats/{history_conversation.pk}/conversation/?protocol=text"
@@ -335,8 +343,9 @@ def test_post_conversation_text_protocol_with_history(
     assert len(history_conversation.messages) == 6
 
     # Verify the most recent messages are the new ones
+    assert history_conversation.messages[4].id == IsUUID(4)
     assert history_conversation.messages[4] == UIMessage(
-        id=str(mock_uuid4),  # Mocked UUID
+        id=history_conversation.messages[4].id,
         createdAt=timezone.now(),  # Mocked timestamp
         content="Hello",
         reasoning=None,
@@ -347,8 +356,9 @@ def test_post_conversation_text_protocol_with_history(
         parts=[TextUIPart(type="text", text="Hello")],
     )
 
+    assert history_conversation.messages[5].id == IsUUID(4)
     assert history_conversation.messages[5] == UIMessage(
-        id=str(mock_uuid4),  # Mocked UUID
+        id=history_conversation.messages[5].id,
         createdAt=timezone.now(),  # Mocked timestamp
         content="Hello there",
         reasoning=None,
@@ -363,7 +373,7 @@ def test_post_conversation_text_protocol_with_history(
 @freeze_time("2025-07-25T10:36:35.297675Z")
 @respx.mock
 def test_post_conversation_with_image_with_history(
-    api_client, mock_openai_stream_image, mock_uuid4, history_conversation
+    api_client, mock_openai_stream_image, history_conversation
 ):
     """
     Ensure an image URL is correctly forwarded to the AI service with a conversation with history.
@@ -403,10 +413,14 @@ def test_post_conversation_with_image_with_history(
 
     # Wait for the streaming content to be fully received
     response_content = b"".join(response.streaming_content).decode("utf-8")
+
+    # Replace UUIDs with placeholders for assertion
+    response_content = replace_uuids_with_placeholder(response_content)
+
     assert response_content == (
         '0:"I see a cat"\n'
         '0:" in the picture."\n'
-        f'f:{{"messageId":"{mock_uuid4}"}}\n'
+        'f:{"messageId":"<mocked_uuid>"}\n'
         'd:{"finishReason":"stop","usage":{"promptTokens":0,"completionTokens":0}}\n'
     )
 
@@ -452,8 +466,9 @@ def test_post_conversation_with_image_with_history(
     assert len(history_conversation.messages) == 6
 
     # Verify the most recent message has the image attachment
+    assert history_conversation.messages[4].id == IsUUID(4)
     assert history_conversation.messages[4] == UIMessage(
-        id=str(mock_uuid4),  # Mocked UUID
+        id=history_conversation.messages[4].id,
         createdAt=timezone.now(),  # Mocked timestamp
         content="Hello, what do you see on this picture?",
         reasoning=None,
@@ -474,8 +489,9 @@ def test_post_conversation_with_image_with_history(
         parts=[TextUIPart(type="text", text="Hello, what do you see on this picture?")],
     )
 
+    assert history_conversation.messages[5].id == IsUUID(4)
     assert history_conversation.messages[5] == UIMessage(
-        id=str(mock_uuid4),  # Mocked UUID
+        id=history_conversation.messages[5].id,
         createdAt=timezone.now(),  # Mocked timestamp
         content="I see a cat in the picture.",
         reasoning=None,
@@ -490,7 +506,7 @@ def test_post_conversation_with_image_with_history(
 @freeze_time("2025-07-25T10:36:35.297675Z")
 @respx.mock
 def test_post_conversation_tool_call_with_history(
-    api_client, mock_openai_stream_tool, mock_uuid4, settings, history_conversation
+    api_client, mock_openai_stream_tool, settings, history_conversation
 ):
     """
     Ensure tool calls are correctly forwarded and streamed back with a conversation with history.
@@ -521,6 +537,10 @@ def test_post_conversation_tool_call_with_history(
 
     # Wait for the streaming content to be fully received
     response_content = b"".join(response.streaming_content).decode("utf-8")
+
+    # Replace UUIDs with placeholders for assertion
+    response_content = replace_uuids_with_placeholder(response_content)
+
     assert response_content == (
         'b:{"toolCallId":"xLDcIljdsDrz0idal7tATWSMm2jhMj47","toolName":'
         '"get_current_weather"}\n'
@@ -529,7 +549,7 @@ def test_post_conversation_tool_call_with_history(
         'a:{"toolCallId":"xLDcIljdsDrz0idal7tATWSMm2jhMj47","result":{"location":'
         '"Paris","temperature":22,"unit":"celsius"}}\n'
         '0:"The current weather in Paris is nice"\n'
-        f'f:{{"messageId":"{mock_uuid4}"}}\n'
+        'f:{"messageId":"<mocked_uuid>"}\n'
         'd:{"finishReason":"stop","usage":{"promptTokens":0,"completionTokens":0}}\n'
     )
 
@@ -561,8 +581,9 @@ def test_post_conversation_tool_call_with_history(
     assert len(history_conversation.messages) == 6
 
     # Verify the most recent message is the new one with tool invocation
+    assert history_conversation.messages[4].id == IsUUID(4)
     assert history_conversation.messages[4] == UIMessage(
-        id=str(mock_uuid4),  # Mocked UUID
+        id=history_conversation.messages[4].id,
         createdAt=timezone.now(),  # Mocked timestamp
         content="Weather in Paris?",
         reasoning=None,
@@ -573,8 +594,9 @@ def test_post_conversation_tool_call_with_history(
         parts=[TextUIPart(type="text", text="Weather in Paris?")],
     )
 
+    assert history_conversation.messages[5].id == IsUUID(4)
     assert history_conversation.messages[5] == UIMessage(
-        id=str(mock_uuid4),  # Mocked UUID
+        id=history_conversation.messages[5].id,
         createdAt=timezone.now(),  # Mocked timestamp
         content="The current weather in Paris is nice",
         reasoning=None,
@@ -606,7 +628,7 @@ def test_post_conversation_tool_call_with_history(
 @freeze_time("2025-07-25T10:36:35.297675Z")
 @respx.mock
 def test_post_conversation_tool_call_fails_with_history(
-    api_client, mock_openai_stream_tool, mock_uuid4, settings, history_conversation
+    api_client, mock_openai_stream_tool, settings, history_conversation
 ):
     """
     Ensure tool calls are correctly forwarded and streamed back when failing with a
@@ -638,6 +660,10 @@ def test_post_conversation_tool_call_fails_with_history(
 
     # Wait for the streaming content to be fully received
     response_content = b"".join(response.streaming_content).decode("utf-8")
+
+    # Replace UUIDs with placeholders for assertion
+    response_content = replace_uuids_with_placeholder(response_content)
+
     assert response_content == (
         'b:{"toolCallId":"xLDcIljdsDrz0idal7tATWSMm2jhMj47","toolName":'
         '"get_current_weather"}\n'
@@ -646,7 +672,7 @@ def test_post_conversation_tool_call_fails_with_history(
         'a:{"toolCallId":"xLDcIljdsDrz0idal7tATWSMm2jhMj47","result":"Unknown tool '
         "name: 'get_current_weather'. No tools available.\"}\n"
         '0:"I cannot give you an answer to that."\n'
-        f'f:{{"messageId":"{mock_uuid4}"}}\n'
+        'f:{"messageId":"<mocked_uuid>"}\n'
         'd:{"finishReason":"stop","usage":{"promptTokens":0,"completionTokens":0}}\n'
     )
 
@@ -678,8 +704,9 @@ def test_post_conversation_tool_call_fails_with_history(
     assert len(history_conversation.messages) == 6
 
     # Verify the most recent message is the new one with tool invocation
+    assert history_conversation.messages[4].id == IsUUID(4)
     assert history_conversation.messages[4] == UIMessage(
-        id=str(mock_uuid4),  # Mocked UUID
+        id=history_conversation.messages[4].id,
         createdAt=timezone.now(),  # Mocked timestamp
         content="Weather in Paris?",
         reasoning=None,
@@ -690,8 +717,9 @@ def test_post_conversation_tool_call_fails_with_history(
         parts=[TextUIPart(type="text", text="Weather in Paris?")],
     )
 
+    assert history_conversation.messages[5].id == IsUUID(4)
     assert history_conversation.messages[5] == UIMessage(
-        id=str(mock_uuid4),  # Mocked UUID
+        id=history_conversation.messages[5].id,
         createdAt=timezone.now(),  # Mocked timestamp
         content="I cannot give you an answer to that.",
         reasoning=None,
@@ -1147,7 +1175,7 @@ def history_conversation_with_tool_fixture():
 @freeze_time("2025-07-25T10:36:35.297675Z")
 @respx.mock
 def test_post_conversation_with_existing_image_history(
-    api_client, mock_openai_stream, mock_uuid4, history_conversation_with_image
+    api_client, mock_openai_stream, history_conversation_with_image
 ):
     """Test posting a message to a conversation that already has images in its history."""
     url = f"/api/v1.0/chats/{history_conversation_with_image.pk}/conversation/?protocol=data"
@@ -1173,10 +1201,14 @@ def test_post_conversation_with_existing_image_history(
 
     # Wait for the streaming content to be fully received
     response_content = b"".join(response.streaming_content).decode("utf-8")
+
+    # Replace UUIDs with placeholders for assertion
+    response_content = replace_uuids_with_placeholder(response_content)
+
     assert response_content == (
         '0:"Hello"\n'
         '0:" there"\n'
-        f'f:{{"messageId":"{mock_uuid4}"}}\n'
+        'f:{"messageId":"<mocked_uuid>"}\n'
         'd:{"finishReason":"stop","usage":{"promptTokens":0,"completionTokens":0}}\n'
     )
 
@@ -1207,8 +1239,9 @@ def test_post_conversation_with_existing_image_history(
     assert len(history_conversation_with_image.messages) == 6
 
     # Verify the most recent messages are the new ones
+    assert history_conversation_with_image.messages[4].id == IsUUID(4)
     assert history_conversation_with_image.messages[4] == UIMessage(
-        id=str(mock_uuid4),  # Mocked UUID
+        id=history_conversation_with_image.messages[4].id,
         createdAt=timezone.now(),  # Mocked timestamp
         content="What was in that image again?",
         reasoning=None,
@@ -1219,8 +1252,9 @@ def test_post_conversation_with_existing_image_history(
         parts=[TextUIPart(type="text", text="What was in that image again?")],
     )
 
+    assert history_conversation_with_image.messages[5].id == IsUUID(4)
     assert history_conversation_with_image.messages[5] == UIMessage(
-        id=str(mock_uuid4),  # Mocked UUID
+        id=history_conversation_with_image.messages[5].id,
         createdAt=timezone.now(),  # Mocked timestamp
         content="Hello there",
         reasoning=None,
@@ -1238,7 +1272,7 @@ def test_post_conversation_with_existing_image_history(
 @freeze_time("2025-07-25T10:36:35.297675Z")
 @respx.mock
 def test_post_conversation_with_existing_tool_history(
-    api_client, mock_openai_stream_tool, mock_uuid4, settings, history_conversation_with_tool
+    api_client, mock_openai_stream_tool, settings, history_conversation_with_tool
 ):
     """Test posting a message to a conversation that already has tool calls in its history."""
     settings.AI_AGENT_TOOLS = ["get_current_weather"]
@@ -1266,6 +1300,10 @@ def test_post_conversation_with_existing_tool_history(
 
     # Wait for the streaming content to be fully received
     response_content = b"".join(response.streaming_content).decode("utf-8")
+
+    # Replace UUIDs with placeholders for assertion
+    response_content = replace_uuids_with_placeholder(response_content)
+
     assert response_content == (
         'b:{"toolCallId":"xLDcIljdsDrz0idal7tATWSMm2jhMj47","toolName":'
         '"get_current_weather"}\n'
@@ -1274,7 +1312,7 @@ def test_post_conversation_with_existing_tool_history(
         'a:{"toolCallId":"xLDcIljdsDrz0idal7tATWSMm2jhMj47","result":{"location":'
         '"Paris","temperature":22,"unit":"celsius"}}\n'
         '0:"The current weather in Paris is nice"\n'
-        f'f:{{"messageId":"{mock_uuid4}"}}\n'
+        'f:{"messageId":"<mocked_uuid>"}\n'
         'd:{"finishReason":"stop","usage":{"promptTokens":0,"completionTokens":0}}\n'
     )
 
@@ -1294,8 +1332,9 @@ def test_post_conversation_with_existing_tool_history(
     assert len(history_conversation_with_tool.messages) == 6
 
     # Verify the most recent message is the new one with tool invocation
+    assert history_conversation_with_tool.messages[4].id == IsUUID(4)
     assert history_conversation_with_tool.messages[4] == UIMessage(
-        id=str(mock_uuid4),  # Mocked UUID
+        id=history_conversation_with_tool.messages[4].id,
         createdAt=timezone.now(),  # Mocked timestamp
         content="How about Paris weather?",
         reasoning=None,
@@ -1306,8 +1345,9 @@ def test_post_conversation_with_existing_tool_history(
         parts=[TextUIPart(type="text", text="How about Paris weather?")],
     )
 
+    assert history_conversation_with_tool.messages[5].id == IsUUID(4)
     assert history_conversation_with_tool.messages[5] == UIMessage(
-        id=str(mock_uuid4),  # Mocked UUID
+        id=history_conversation_with_tool.messages[5].id,
         createdAt=timezone.now(),  # Mocked timestamp
         content="The current weather in Paris is nice",
         reasoning=None,
@@ -1417,7 +1457,7 @@ def test_post_conversation_with_existing_tool_history(
 @freeze_time("2025-07-25T10:36:35.297675Z")
 @respx.mock
 def test_post_conversation_add_image_to_conversation_with_tool_history(
-    api_client, mock_openai_stream_image, mock_uuid4, history_conversation_with_tool
+    api_client, mock_openai_stream_image, history_conversation_with_tool
 ):
     """Test adding an image to a conversation that already has tool calls in its history."""
     url = f"/api/v1.0/chats/{history_conversation_with_tool.pk}/conversation/?protocol=data"
@@ -1455,10 +1495,14 @@ def test_post_conversation_add_image_to_conversation_with_tool_history(
 
     # Wait for the streaming content to be fully received
     response_content = b"".join(response.streaming_content).decode("utf-8")
+
+    # Replace UUIDs with placeholders for assertion
+    response_content = replace_uuids_with_placeholder(response_content)
+
     assert response_content == (
         '0:"I see a cat"\n'
         '0:" in the picture."\n'
-        f'f:{{"messageId":"{mock_uuid4}"}}\n'
+        'f:{"messageId":"<mocked_uuid>"}\n'
         'd:{"finishReason":"stop","usage":{"promptTokens":0,"completionTokens":0}}\n'
     )
 
@@ -1484,8 +1528,9 @@ def test_post_conversation_add_image_to_conversation_with_tool_history(
     assert len(history_conversation_with_tool.messages) == 6
 
     # Verify the most recent message has the image attachment
+    assert history_conversation_with_tool.messages[4].id == IsUUID(4)
     assert history_conversation_with_tool.messages[4] == UIMessage(
-        id=str(mock_uuid4),  # Mocked UUID
+        id=history_conversation_with_tool.messages[4].id,
         createdAt=timezone.now(),  # Mocked timestamp
         content="How's the weather in this image?",
         reasoning=None,
@@ -1506,8 +1551,9 @@ def test_post_conversation_add_image_to_conversation_with_tool_history(
         parts=[TextUIPart(type="text", text="How's the weather in this image?")],
     )
 
+    assert history_conversation_with_tool.messages[5].id == IsUUID(4)
     assert history_conversation_with_tool.messages[5] == UIMessage(
-        id=str(mock_uuid4),  # Mocked UUID
+        id=history_conversation_with_tool.messages[5].id,
         createdAt=timezone.now(),  # Mocked timestamp
         content="I see a cat in the picture.",
         reasoning=None,

src/backend/chat/tests/views/chat/conversations/test_conversation_with_image_url.py

@@ -1,8 +1,11 @@
 """Unit tests for chat conversation actions with image URL."""
 
+import uuid
+
 from django.utils import timezone
 
 import pytest
+from dirty_equals import IsUUID
 from freezegun import freeze_time
 from pydantic_ai import ModelRequest, RequestUsage
 from pydantic_ai.messages import (
@@ -22,6 +25,7 @@ from chat.ai_sdk_types import (
     UIMessage,
 )
 from chat.factories import ChatConversationFactory
+from chat.tests.utils import replace_uuids_with_placeholder
 
 # enable database transactions for tests:
 # transaction=True ensures that the data are available in the database
@@ -53,7 +57,6 @@ def fixture_sample_image_content():
 @freeze_time("2025-10-18T20:48:20.286204Z")
 def test_post_conversation_with_local_image_url(
     api_client,
-    mock_uuid4,
     mock_ai_agent_service,
 ):
     """
@@ -131,17 +134,23 @@ def test_post_conversation_with_local_image_url(
 
     # Wait for the streaming content to be fully received
     response_content = b"".join(response.streaming_content).decode("utf-8")
+
+    # Replace UUIDs with placeholders for assertion
+    response_content = replace_uuids_with_placeholder(response_content)
+
     assert response_content == (
         '0:"This is an image of a single pixel."\n'
-        f'f:{{"messageId":"{mock_uuid4}"}}\n'
+        'f:{"messageId":"<mocked_uuid>"}\n'
         'd:{"finishReason":"stop","usage":{"promptTokens":50,"completionTokens":9}}\n'
     )
 
     # Check that the conversation was updated
     chat_conversation.refresh_from_db()
     assert len(chat_conversation.messages) == 2
+
+    assert chat_conversation.messages[0].id == IsUUID(4)
     assert chat_conversation.messages[0] == UIMessage(
-        id=str(mock_uuid4),
+        id=chat_conversation.messages[0].id,  # don't test the value directly
         createdAt=timezone.now(),
         content="What is in this image?",
         reasoning=None,
@@ -155,8 +164,10 @@ def test_post_conversation_with_local_image_url(
             TextUIPart(type="text", text="What is in this image?"),
         ],
     )
+
+    assert chat_conversation.messages[1].id == IsUUID(4)
     assert chat_conversation.messages[1] == UIMessage(
-        id=str(mock_uuid4),
+        id=chat_conversation.messages[1].id,  # don't test the value directly
         createdAt=timezone.now(),
         content="This is an image of a single pixel.",
         reasoning=None,
@@ -238,7 +249,6 @@ def test_post_conversation_with_local_image_url(
 def test_post_conversation_with_local_image_wrong_url(
     api_client,
     today_promt_date,
-    mock_uuid4,
     mock_ai_agent_service,
 ):
     """
@@ -247,7 +257,7 @@ def test_post_conversation_with_local_image_wrong_url(
     chat_conversation = ChatConversationFactory(owner__language="en-us")
     api_client.force_authenticate(user=chat_conversation.owner)
 
-    image_url = f"/media-key/{mock_uuid4}/sample.png"
+    image_url = f"/media-key/{uuid.uuid4()}/sample.png"
 
     message = UIMessage(
         id="1",
@@ -308,9 +318,13 @@ def test_post_conversation_with_local_image_wrong_url(
 
     # Wait for the streaming content to be fully received
     response_content = b"".join(response.streaming_content).decode("utf-8")
+
+    # Replace UUIDs with placeholders for assertion
+    response_content = replace_uuids_with_placeholder(response_content)
+
     assert response_content == (
         '0:"cannot read image."\n'
-        f'f:{{"messageId":"{mock_uuid4}"}}\n'
+        'f:{"messageId":"<mocked_uuid>"}\n'
         'd:{"finishReason":"stop","usage":{"promptTokens":50,"completionTokens":4}}\n'
     )
 
@@ -322,7 +336,6 @@ def test_post_conversation_with_local_image_wrong_url(
 def test_post_conversation_with_remote_image_url(
     api_client,
     today_promt_date,
-    mock_uuid4,
     mock_ai_agent_service,
 ):
     """
@@ -392,17 +405,23 @@ def test_post_conversation_with_remote_image_url(
 
     # Wait for the streaming content to be fully received
     response_content = b"".join(response.streaming_content).decode("utf-8")
+
+    # Replace UUIDs with placeholders for assertion
+    response_content = replace_uuids_with_placeholder(response_content)
+
     assert response_content == (
         '0:"This is an image of a single pixel."\n'
-        f'f:{{"messageId":"{mock_uuid4}"}}\n'
+        'f:{"messageId":"<mocked_uuid>"}\n'
         'd:{"finishReason":"stop","usage":{"promptTokens":50,"completionTokens":9}}\n'
     )
 
     # Check that the conversation was updated
     chat_conversation.refresh_from_db()
     assert len(chat_conversation.messages) == 2
+
+    assert chat_conversation.messages[0].id == IsUUID(4)
     assert chat_conversation.messages[0] == UIMessage(
-        id=str(mock_uuid4),
+        id=chat_conversation.messages[0].id,  # don't test the value directly
         createdAt=timezone.now(),
         content="What is in this image?",
         reasoning=None,
@@ -416,8 +435,10 @@ def test_post_conversation_with_remote_image_url(
             TextUIPart(type="text", text="What is in this image?"),
         ],
     )
+
+    assert chat_conversation.messages[1].id == IsUUID(4)
     assert chat_conversation.messages[1] == UIMessage(
-        id=str(mock_uuid4),
+        id=chat_conversation.messages[1].id,  # don't test the value directly
         createdAt=timezone.now(),
         content="This is an image of a single pixel.",
         reasoning=None,
@@ -435,7 +456,6 @@ def test_post_conversation_with_remote_image_url(
 def test_post_conversation_with_local_image_url_in_history(
     api_client,
     today_promt_date,
-    mock_uuid4,
     mock_ai_agent_service,
 ):
     """
@@ -448,7 +468,7 @@ def test_post_conversation_with_local_image_url_in_history(
         owner__language="en-us",
         messages=[
             UIMessage(
-                id=str(mock_uuid4),
+                id=str(uuid.uuid4()),
                 createdAt=timezone.now(),
                 content="What is in this image?",
                 reasoning=None,
@@ -463,7 +483,7 @@ def test_post_conversation_with_local_image_url_in_history(
                 ],
             ),
             UIMessage(
-                id=str(mock_uuid4),
+                id=str(uuid.uuid4()),
                 createdAt=timezone.now(),
                 content="This is an image of a single pixel.",
                 reasoning=None,
@@ -629,17 +649,23 @@ def test_post_conversation_with_local_image_url_in_history(
 
     # Wait for the streaming content to be fully received
     response_content = b"".join(response.streaming_content).decode("utf-8")
+
+    # Replace UUIDs with placeholders for assertion
+    response_content = replace_uuids_with_placeholder(response_content)
+
     assert response_content == (
         '0:"This is an image of square, very small and nice."\n'
-        f'f:{{"messageId":"{mock_uuid4}"}}\n'
+        'f:{"messageId":"<mocked_uuid>"}\n'
         'd:{"finishReason":"stop","usage":{"promptTokens":50,"completionTokens":11}}\n'
     )
 
     # Check that the conversation was updated
     chat_conversation.refresh_from_db()
     assert len(chat_conversation.messages) == 2 + 2
+
+    assert chat_conversation.messages[0].id == IsUUID(4)
     assert chat_conversation.messages[0] == UIMessage(
-        id=str(mock_uuid4),
+        id=chat_conversation.messages[0].id,  # don't test the value directly
         createdAt=timezone.now(),
         content="What is in this image?",
         reasoning=None,
@@ -653,8 +679,10 @@ def test_post_conversation_with_local_image_url_in_history(
             TextUIPart(type="text", text="What is in this image?"),
         ],
     )
+
+    assert chat_conversation.messages[1].id == IsUUID(4)
     assert chat_conversation.messages[1] == UIMessage(
-        id=str(mock_uuid4),
+        id=chat_conversation.messages[1].id,  # don't test the value directly
         createdAt=timezone.now(),
         content="This is an image of a single pixel.",
         reasoning=None,
@@ -666,8 +694,10 @@ def test_post_conversation_with_local_image_url_in_history(
             TextUIPart(type="text", text="This is an image of a single pixel."),
         ],
     )
+
+    assert chat_conversation.messages[2].id == IsUUID(4)
     assert chat_conversation.messages[2] == UIMessage(
-        id=str(mock_uuid4),
+        id=chat_conversation.messages[2].id,  # don't test the value directly
         createdAt=timezone.now(),
         content="Give more details about this image.",
         reasoning=None,
@@ -679,8 +709,10 @@ def test_post_conversation_with_local_image_url_in_history(
             TextUIPart(type="text", text="Give more details about this image."),
         ],
     )
+
+    assert chat_conversation.messages[3].id == IsUUID(4)
     assert chat_conversation.messages[3] == UIMessage(
-        id=str(mock_uuid4),
+        id=chat_conversation.messages[3].id,  # don't test the value directly
         createdAt=timezone.now(),
         content="This is an image of square, very small and nice.",
         reasoning=None,

src/backend/chat/tests/views/chat/test_media_auth.py

@@ -25,6 +25,8 @@ def test_api_media_auth_unkown_document(api_client):
     Trying to download a media related to a conversation that does not exist
     should not have the side effect to create it (no regression test).
     """
+    ChatConversation.objects.all().delete()
+
     original_url = f"http://localhost/media/{uuid4()!s}/attachments/{uuid4()!s}.jpg"
 
     response = api_client.get("/api/v1.0/chats/media-auth/", HTTP_X_ORIGINAL_URL=original_url)

src/backend/chat/tools/__init__.py

@@ -18,18 +18,28 @@ def get_pydantic_tools_by_name(name: str) -> Tool:
     tool_dict = {
         "get_current_weather": Tool(get_current_weather, takes_ctx=False),
         "web_search_brave": Tool(
-            web_search_brave, takes_ctx=False, prepare=only_if_web_search_enabled
+            web_search_brave,
+            takes_ctx=True,
+            prepare=only_if_web_search_enabled,
+            max_retries=2,
         ),
         "web_search_brave_with_document_backend": Tool(
             web_search_brave_with_document_backend,
             takes_ctx=True,
             prepare=only_if_web_search_enabled,
+            max_retries=2,
         ),
         "web_search_tavily": Tool(
-            web_search_tavily, takes_ctx=False, prepare=only_if_web_search_enabled
+            web_search_tavily,
+            takes_ctx=False,
+            prepare=only_if_web_search_enabled,
+            max_retries=2,
         ),
         "web_search_albert_rag": Tool(
-            web_search_albert_rag, takes_ctx=True, prepare=only_if_web_search_enabled
+            web_search_albert_rag,
+            takes_ctx=True,
+            prepare=only_if_web_search_enabled,
+            max_retries=2,
         ),
     }
 

src/backend/chat/tools/document_search_rag.py

@@ -20,7 +20,7 @@ def add_document_rag_search_tool(agent: Agent) -> None:
 
         Args:
             ctx (RunContext): The run context containing the conversation.
-            query (str): The term to search the internet for.
+            query (str): The query to search the documents for.
         """
         document_store_backend = import_string(settings.RAG_DOCUMENT_SEARCH_BACKEND)
 
@@ -43,8 +43,6 @@ def add_document_rag_search_tool(agent: Agent) -> None:
     def document_rag_instructions() -> str:
         """Dynamic system prompt function to add RAG instructions if any."""
         return (
-            "If the user wants specific information from a document, invoke "
-            "web_search_albert_rag with an appropriate query string."
-            "Do not ask the user for the document; rely on the tool to locate "
-            "and return relevant passages."
+            "Use document_search_rag ONLY to retrieve specific passages from attached documents. "
+            "Do NOT use it to summarize; for summaries, call the summarize tool instead."
         )

src/backend/chat/tools/document_summarize.py

@@ -0,0 +1,189 @@
+"""Summarization tool used for uploaded documents."""
+
+import asyncio
+import logging
+
+from django.conf import settings
+from django.core.files.storage import default_storage
+
+import semchunk
+from asgiref.sync import sync_to_async
+from pydantic_ai import RunContext
+from pydantic_ai.exceptions import ModelRetry
+from pydantic_ai.messages import ToolReturn
+
+from chat.agents.summarize import SummarizationAgent
+from chat.tools.exceptions import ModelCannotRetry
+from chat.tools.utils import last_model_retry_soft_fail
+
+logger = logging.getLogger(__name__)
+
+
+@sync_to_async
+def read_document_content(doc):
+    """Read document content asynchronously."""
+    with default_storage.open(doc.key) as f:
+        return doc.file_name, f.read().decode("utf-8")
+
+
+async def summarize_chunk(idx, chunk, total_chunks, summarization_agent, ctx):
+    """Summarize a single chunk of text."""
+    sum_prompt = (
+        "You are an agent specializing in text summarization. "
+        "Generate a clear and concise summary of the following passage "
+        f"(part {idx}/{total_chunks}):\n'''\n{chunk}\n'''\n\n"
+    )
+
+    logger.debug(
+        "[summarize] CHUNK %s/%s prompt=> %s", idx, total_chunks, sum_prompt[0:100] + "..."
+    )
+
+    try:
+        resp = await summarization_agent.run(sum_prompt, usage=ctx.usage)
+    except Exception as exc:
+        logger.warning("Error during chunk summarization: %s", exc, exc_info=True)
+        raise ModelRetry(
+            "An error occurred while summarizing a part of the document chunk."
+        ) from exc
+
+    logger.debug("[summarize] CHUNK %s/%s response<= %s", idx, total_chunks, resp.output or "")
+    return resp.output or ""
+
+
+@last_model_retry_soft_fail
+async def document_summarize(  # pylint: disable=too-many-locals
+    ctx: RunContext, *, instructions: str | None = None
+) -> ToolReturn:
+    """
+    Generate a complete, ready-to-use summary of the documents in context
+    (do not request the documents to the user).
+    Return this summary directly to the user WITHOUT any modification,
+    or additional summarization.
+    The summary is already optimized and MUST be presented as-is in the final response
+    or translated preserving the information.
+
+    Instructions are optional but should reflect the user's request.
+
+    Examples:
+    "Summarize this doc in 2 paragraphs" -> instructions = "summary in 2 paragraphs"
+    "Summarize this doc in English" -> instructions = "In English"
+    "Summarize this doc" -> instructions = "" (default)
+
+    Args:
+        instructions (str | None): The instructions the user gave to use for the summarization
+    """
+    try:
+        instructions_hint = (
+            instructions.strip() if instructions else "The summary should contain 2 or 3 parts."
+        )
+        summarization_agent = SummarizationAgent()
+
+        # Collect documents content
+        text_attachment = await sync_to_async(list)(
+            ctx.deps.conversation.attachments.filter(
+                content_type__startswith="text/",
+            )
+        )
+
+        if not text_attachment:
+            raise ModelCannotRetry(
+                "No text documents found in the conversation. "
+                "You must explain this to the user and ask them to provide documents."
+            )
+
+        documents = [await read_document_content(doc) for doc in text_attachment]
+
+        # Chunk documents and summarize each chunk
+        chunk_size = settings.SUMMARIZATION_CHUNK_SIZE
+        chunker = semchunk.chunkerify(
+            tokenizer_or_token_counter=lambda text: len(text.split()),
+            chunk_size=chunk_size,
+        )
+        documents_chunks = chunker(
+            [doc[1] for doc in documents],
+            overlap=settings.SUMMARIZATION_OVERLAP_SIZE,
+        )
+
+        logger.info(
+            "[summarize] chunking: %s parts (size~%s), instructions='%s'",
+            sum(len(chunks) for chunks in documents_chunks),
+            chunk_size,
+            instructions_hint,
+        )
+
+        # Parallelize the chunk summarization with a semaphore to limit concurrent tasks
+        # because it can be very resource intensive on the LLM backend
+        semaphore = asyncio.Semaphore(settings.SUMMARIZATION_CONCURRENT_REQUESTS)
+
+        async def summarize_chunk_with_semaphore(idx, chunk, total_chunks):
+            """Summarize a chunk with semaphore-controlled concurrency."""
+            async with semaphore:
+                return await summarize_chunk(idx, chunk, total_chunks, summarization_agent, ctx)
+
+        doc_chunk_summaries = []
+        try:
+            for doc_chunks in documents_chunks:
+                summarization_tasks = [
+                    summarize_chunk_with_semaphore(idx, chunk, len(doc_chunks))
+                    for idx, chunk in enumerate(doc_chunks, start=1)
+                ]
+                chunk_summaries = await asyncio.gather(*summarization_tasks)
+                doc_chunk_summaries.append(chunk_summaries)
+        except ModelRetry as exc:
+            logger.warning("Retryable error during chunk summarization: %s", exc, exc_info=True)
+            raise
+        except Exception as exc:
+            logger.warning("Error during chunk summarization: %s", exc, exc_info=True)
+            raise ModelRetry("An error occurred while processing document chunks.") from exc
+
+        context = "\n\n".join(
+            doc_name + "\n\n" + "\n\n".join(summaries)
+            for doc_name, summaries in zip(
+                (doc[0] for doc in documents),
+                doc_chunk_summaries,
+                strict=True,
+            )
+        )
+
+        # Merge chunk summaries into a single concise summary
+        merged_prompt = (
+            "Produce a coherent synthesis from the summaries below.\n\n"
+            f"'''\n{context}\n'''\n\n"
+            "Constraints:\n"
+            "- Summarize without repetition.\n"
+            "- Harmonize style and terminology.\n"
+            "- The final summary must be well-structured and formatted in markdown.\n"
+            f"- Follow the instructions: {instructions_hint}\n"
+            "Respond directly with the final summary."
+        )
+
+        logger.debug("[summarize] MERGE prompt=> %s", merged_prompt)
+
+        try:
+            merged_resp = await summarization_agent.run(merged_prompt, usage=ctx.usage)
+        except Exception as exc:
+            logger.warning("Error during merge summarization: %s", exc, exc_info=True)
+            raise ModelRetry("An error occurred while generating the final summary.") from exc
+
+        final_summary = (merged_resp.output or "").strip()
+
+        if not final_summary:
+            raise ModelRetry("The summarization produced an empty result.")
+
+        logger.debug("[summarize] MERGE response<= %s", final_summary)
+
+        return ToolReturn(
+            return_value=final_summary,
+            metadata={"sources": {doc[0] for doc in documents}},
+        )
+
+    except (ModelCannotRetry, ModelRetry):
+        # Re-raise these as-is
+        raise
+    except Exception as exc:
+        # Unexpected error - stop and inform user
+        logger.exception("Unexpected error in document_summarize: %s", exc)
+        raise ModelCannotRetry(
+            f"An unexpected error occurred during document summarization: {type(exc).__name__}. "
+            "You must explain this to the user and not try to answer based on your knowledge."
+        ) from exc

src/backend/chat/tools/exceptions.py

@@ -0,0 +1,23 @@
+"""Exceptions for tool function retries."""
+
+from pydantic_ai import ModelRetry
+
+
+class ModelRetryLast(ModelRetry):
+    """
+    Same as ModelRetry but also holds the last retry message to return when all attempts failed.
+    """
+
+    def __init__(self, message: str, last_retry_message: str):
+        """Initialize ModelRetryLast with message and last retry message."""
+        self.last_retry_message = last_retry_message
+        super().__init__(message)
+
+
+class ModelCannotRetry(ModelRetry):
+    """
+    Exception to raise when a tool function cannot be retried.
+
+    We use this exception to signal that the model should not attempt to retry
+    the tool call, typically because the error is not transient or recoverable.
+    """

src/backend/chat/tools/utils.py

@@ -0,0 +1,50 @@
+"""Tool calling utilities for the chat agent."""
+
+import functools
+import logging
+from typing import Any, Callable
+
+from pydantic_ai import ModelRetry, RunContext
+
+from chat.tools.exceptions import ModelCannotRetry
+
+logger = logging.getLogger(__name__)
+
+
+def last_model_retry_soft_fail(
+    tool_func: Callable[..., Any],
+) -> Callable[..., Any]:
+    """
+    Wrap a tool function to handle ModelRetry exceptions.
+
+    If the tool function raises ModelRetry and the maximum number of retries
+    has been reached, a ModelCannotRetry exception is raised instead.
+
+    Args:
+        tool_func: The original tool function to wrap.
+
+    Returns:
+        A wrapped tool function with retry handling.
+    """
+
+    @functools.wraps(tool_func)
+    async def wrapper(ctx: RunContext, *args, **kwargs) -> Any:
+        try:
+            return await tool_func(ctx, *args, **kwargs)
+        except ModelCannotRetry as exc:
+            return str(exc.message)
+        except ModelRetry as exc:
+            logger.error("Tool '%s' raised ModelRetry: %s", ctx, exc.message)
+            if (ctx.retries.get(ctx.tool_name, 0) + 1) >= ctx.max_retries:
+                logger.error("Max retries reached for tool '%s'.", ctx.tool_name)
+                # A bit of a hack to signal that we cannot retry here, while preventing
+                # the LLM to generate an outdated answer.
+                # We may define a more specific exception later base on ModelRetry which
+                # adds a specific message for this case.
+                return (
+                    f"{exc.message} You must explain this to the user and "
+                    "not try to answer based on your knowledge."
+                )
+            raise  # Re-raise to allow retrying
+
+    return wrapper

src/backend/chat/tools/web_search_brave.py

@@ -1,24 +1,42 @@
 """Web search tool using Brave for the chat agent."""
 
+import asyncio
 import logging
 import uuid
-from concurrent.futures import ThreadPoolExecutor, as_completed
 from typing import List
 
 from django.conf import settings
 from django.core.cache import cache
 from django.utils.module_loading import import_string
+from django.utils.text import slugify
 
-import requests
+import httpx
+from asgiref.sync import sync_to_async
 from pydantic_ai import RunContext, RunUsage
+from pydantic_ai.exceptions import ModelRetry
 from pydantic_ai.messages import ToolReturn
-from trafilatura import extract, fetch_url
+from trafilatura import extract
 from trafilatura.meta import reset_caches
 
+from chat.tools.exceptions import ModelCannotRetry
+from chat.tools.utils import last_model_retry_soft_fail
+
 logger = logging.getLogger(__name__)
 
 
-def llm_summarize(query: str, text: str) -> str:
+class WebSearchError(Exception):
+    """Base exception for web search errors."""
+
+
+class BraveAPIError(WebSearchError):
+    """Error when calling Brave API."""
+
+
+class DocumentFetchError(WebSearchError):
+    """Error when fetching or extracting documents."""
+
+
+async def llm_summarize_async(query: str, text: str) -> str:
     """
     Summarize the text using the LLM summarization agent.
 
@@ -33,7 +51,7 @@ def llm_summarize(query: str, text: str) -> str:
     prompt = f"""
 Based on the following request, summarize the following text in a concise manner, 
 focusing on the key points regarding the user request. 
-he result should be up to 30 lines long.
+The result should be up to 30 lines long.
 
 <user request>
 {query}
@@ -44,54 +62,87 @@ he result should be up to 30 lines long.
 </text to summarize>
 """
 
-    result = summarization_agent.run_sync(prompt)
+    result = await summarization_agent.run(prompt)
     return result.output
 
 
-def _fetch_and_extract(url: str) -> str:
-    """Fetch and extract text content from the URL."""
-    cache_key = f"web_search_brave:extract:{url}"
+async def _fetch_url_async(url: str, timeout: int = 30) -> str:
+    """Fetch URL content asynchronously."""
+    async with httpx.AsyncClient(timeout=timeout, follow_redirects=True) as client:
+        response = await client.get(url)
+        response.raise_for_status()
+        return response.text
 
-    if (document := cache.get(cache_key)) is not None:
+
+async def _fetch_and_extract_async(url: str) -> str:
+    """Fetch and extract text content from the URL asynchronously."""
+    cache_key = f"web_search_brave:extract:{slugify(url)}"
+
+    # Check cache first
+    if (document := await cache.aget(cache_key)) is not None:
         return document
 
-    html = fetch_url(url)
-    document = extract(html, include_comments=False, no_fallback=True) or ""
-    cache.set(cache_key, document, settings.BRAVE_CACHE_TTL)
+    try:
+        # Fetch HTML
+        html = await _fetch_url_async(url, timeout=settings.BRAVE_API_TIMEOUT)
 
-    return document
+        # Extract text in thread pool (trafilatura is CPU-bound)
+        document = await sync_to_async(extract)(html, include_comments=False, no_fallback=True)
+
+        # Cache the result
+        await cache.aset(cache_key, document, settings.BRAVE_CACHE_TTL)
+        return document
+
+    except httpx.HTTPError as e:
+        logger.warning("HTTP error fetching %s: %s", url, e, exc_info=True)
+        raise DocumentFetchError(f"Failed to fetch {url}: {e}") from e
+    except Exception as e:
+        logger.warning("Error extracting content from %s: %s", url, e, exc_info=True)
+        raise DocumentFetchError(f"Failed to extract content from {url}: {e}") from e
 
 
-def _extract_and_summarize_snippets(query: str, url: str) -> List[str]:
+async def _extract_and_summarize_snippets_async(query: str, url: str) -> List[str]:
     """Fetch, extract and summarize text content from the URL.
 
     Returns a list of snippets (0 or 1 element, preserving existing behavior).
     """
-    # Cache by URL to avoid repeated fetch/extract across calls
-    document = _fetch_and_extract(url)
-    if not document:
+    try:
+        document = await _fetch_and_extract_async(url)
+        if not document:
+            return []
+
+        if not settings.BRAVE_SUMMARIZATION_ENABLED:
+            return [document]
+
+        try:
+            snippet = await llm_summarize_async(query, document)
+            return [snippet] if snippet else []
+        except Exception as e:  # pylint: disable=broad-except
+            logger.exception("Summarization failed for %s: %s", url, e)
+            # Fallback to raw document if summarization fails
+            return [document]
+
+    except DocumentFetchError:
+        # Document fetch failed, return empty
         return []
 
-    if not settings.BRAVE_SUMMARIZATION_ENABLED:
-        return [document]
+
+async def _fetch_and_store_async(url: str, document_store) -> None:
+    """Fetch, extract and store text content from the URL in the document store."""
 
     try:
-        snippet = llm_summarize(query, document)
-    except Exception as e:  # pylint: disable=broad-except
-        logger.exception("Summarization failed for %s: %s", url, e)
-        snippet = None
+        document = await _fetch_and_extract_async(url)
 
-    return [snippet] if snippet else []
+        logger.debug("Fetched document: %s", document)
+
+        if document:
+            await document_store.astore_document(url, document)
+    except DocumentFetchError as e:
+        logger.warning("Failed to fetch and store %s: %s", url, e)
+        # Continue with other documents
 
 
-def _fetch_and_store(url: str, document_store) -> None:
-    """Fetch, extract and store text content from the URL in the document store."""
-    document = _fetch_and_extract(url)
-    if document:
-        document_store.store_document(url, document)
-
-
-def _query_brave_api(query: str) -> List[dict]:
+async def _query_brave_api_async(query: str) -> List[dict]:
     """Query the Brave Search API and return the raw results."""
     url = "https://api.search.brave.com/res/v1/web/search"
     headers = {
@@ -109,14 +160,53 @@ def _query_brave_api(query: str) -> List[dict]:
         "extra_snippets": settings.BRAVE_SEARCH_EXTRA_SNIPPETS,
     }
     params = {k: v for k, v in data.items() if v is not None}
-    response = requests.get(url, headers=headers, params=params, timeout=settings.BRAVE_API_TIMEOUT)
-    response.raise_for_status()
 
-    json_response = response.json()
+    try:
+        async with httpx.AsyncClient(timeout=settings.BRAVE_API_TIMEOUT) as client:
+            response = await client.get(url, headers=headers, params=params)
+            response.raise_for_status()
+            json_response = response.json()
 
-    # See https://api-dashboard.search.brave.com/app/documentation/web-search/responses#Result
-    # & https://api-dashboard.search.brave.com/app/documentation/web-search/responses#SearchResult
-    return json_response.get("web", {}).get("results", [])
+            # https://api-dashboard.search.brave.com/app/documentation/web-search/responses#Result
+            return json_response.get("web", {}).get("results", [])
+
+    except httpx.HTTPStatusError as e:
+        if e.response.status_code == 429:
+            # Rate limit - retryable
+            logger.warning("Brave API rate limited: %s", e)
+            raise ModelRetry(
+                "The search API is rate limited. Please wait a moment and try again."
+            ) from e
+        if e.response.status_code >= 500:
+            # Server error - retryable
+            logger.warning("Brave API error: %s", e)
+            raise ModelRetry(
+                "The search service is temporarily unavailable due to a server error. Retrying..."
+            ) from e
+
+        # Client error (4xx) - not retryable, stop and inform user
+        logger.error("Brave API client error: %s", e)
+        raise ModelCannotRetry(
+            f"Web search failed with a client error (status {e.response.status_code}). "
+            "You must explain this to the user and not try to answer based on your knowledge."
+        ) from e
+    except httpx.TimeoutException as e:
+        # Timeout - retryable
+        logger.warning("Brave API timeout: %s", e)
+        raise ModelRetry("The search request timed out. Retrying with a fresh attempt...") from e
+    except httpx.HTTPError as e:
+        # Other HTTP errors - retryable
+        logger.warning("Brave API connection error: %s", e)
+        raise ModelRetry(
+            f"Connection error while searching the web: {type(e).__name__}. Retrying..."
+        ) from e
+    except Exception as e:
+        # Unexpected errors - not retryable, stop completely
+        logger.exception("Unexpected error querying Brave API: %s", e)
+        raise ModelCannotRetry(
+            f"An unexpected error occurred with the search service: {type(e).__name__}. "
+            "You must explain this to the user and not try to answer based on your knowledge."
+        ) from e
 
 
 def format_tool_return(raw_search_results: List[dict]) -> ToolReturn:
@@ -140,92 +230,132 @@ def format_tool_return(raw_search_results: List[dict]) -> ToolReturn:
     )
 
 
-def web_search_brave(query: str) -> ToolReturn:
+@last_model_retry_soft_fail
+async def web_search_brave(_ctx: RunContext, query: str) -> ToolReturn:
     """
     Search the web for up-to-date information
 
     Args:
+        _ctx (RunContext): The run context, used by the wrapper.
         query (str): The query to search for.
     """
-    raw_search_results = _query_brave_api(query)
+    try:
+        raw_search_results = await _query_brave_api_async(query)
 
-    reset_caches()  # Clear trafilatura caches to avoid memory bloat/leaks
+        await sync_to_async(reset_caches)()  # Clear trafilatura caches to avoid memory bloat/leaks
 
-    # Parallelize fetch/extract for results that don't include extra_snippets
-    to_process = [
-        (idx, r) for idx, r in enumerate(raw_search_results) if not r.get("extra_snippets")
-    ]
+        # Parallelize fetch/extract for results that don't include extra_snippets
+        to_process = [
+            (idx, r) for idx, r in enumerate(raw_search_results) if not r.get("extra_snippets")
+        ]
 
-    if to_process:
-        max_workers = min(settings.BRAVE_MAX_WORKERS, len(to_process))
-        if max_workers == 1:
-            # Avoid overhead of ThreadPoolExecutor if only one task
-            for idx, r in to_process:
-                raw_search_results[idx]["extra_snippets"] = _extract_and_summarize_snippets(
-                    query, r["url"]
-                )
+        if to_process:
+            # Process all URLs concurrently
+            tasks = [
+                _extract_and_summarize_snippets_async(query, r["url"]) for idx, r in to_process
+            ]
+            results = await asyncio.gather(*tasks, return_exceptions=False)
 
-        else:
-            with ThreadPoolExecutor(max_workers=max_workers) as executor:
-                future_map = {
-                    executor.submit(_extract_and_summarize_snippets, query, r["url"]): idx
-                    for idx, r in to_process
-                }
-                for future in as_completed(future_map):
-                    idx = future_map[future]
-                    raw_search_results[idx]["extra_snippets"] = future.result()
+            # Update raw_search_results with extracted snippets
+            for (idx, _), snippets in zip(to_process, results, strict=True):
+                raw_search_results[idx]["extra_snippets"] = snippets
 
-    return format_tool_return(raw_search_results)
+        formatted_result = format_tool_return(raw_search_results)
+
+        # Check if we got any valid results
+        if not formatted_result.return_value:
+            raise ModelRetry(
+                "No valid search results were extracted from the web pages. "
+                "Retrying the search to find better sources..."
+            )
+
+        return formatted_result
+
+    except (ModelCannotRetry, ModelRetry):
+        # Re-raise these as-is
+        raise
+    except Exception as exc:
+        # Unexpected error in our code - stop and inform user
+        logger.exception("Unexpected error in web_search_brave: %s", exc)
+        raise ModelCannotRetry(
+            f"An unexpected error occurred during web search: {type(exc).__name__}. "
+            "You must explain this to the user and not try to answer based on your knowledge."
+        ) from exc
 
 
-def web_search_brave_with_document_backend(ctx: RunContext, query: str) -> ToolReturn:
+@last_model_retry_soft_fail
+async def web_search_brave_with_document_backend(ctx: RunContext, query: str) -> ToolReturn:
     """
-    Search the web for up-to-date information
+    Search the web for up-to-date information using RAG backend
 
     Args:
         ctx (RunContext): The run context containing the conversation.
         query (str): The query to search for.
     """
-    raw_search_results = _query_brave_api(query)
+    logger.info("Starting web search with RAG backend for query: %s", query)
+    try:
+        raw_search_results = await _query_brave_api_async(query)
 
-    reset_caches()  # Clear trafilatura caches to avoid memory bloat/leaks
+        # Clear trafilatura caches in thread pool to avoid blocking
+        loop = asyncio.get_event_loop()
+        await loop.run_in_executor(None, reset_caches)
 
-    # Store documents in a temporary document store for RAG search
-    document_store_backend = import_string(settings.RAG_DOCUMENT_SEARCH_BACKEND)
-    with document_store_backend.temporary_collection(f"tmp-{uuid.uuid4()}") as document_store:
-        max_workers = min(settings.BRAVE_MAX_WORKERS, len(raw_search_results))
-        if max_workers == 1:
-            for result in raw_search_results:
-                # Fetch and extract document content
-                _fetch_and_store(result["url"], document_store)
-        else:
-            with ThreadPoolExecutor(max_workers=max_workers) as executor:
-                futures = [
-                    executor.submit(_fetch_and_store, result["url"], document_store)
+        # Store documents in a temporary document store for RAG search
+        document_store_backend = import_string(settings.RAG_DOCUMENT_SEARCH_BACKEND)
+
+        # Create temporary collection
+        temp_collection_name = f"tmp-{uuid.uuid4()}"
+        try:
+            async with document_store_backend.temporary_collection_async(
+                temp_collection_name
+            ) as document_store:
+                # Fetch and store all documents concurrently
+                tasks = [
+                    _fetch_and_store_async(result["url"], document_store)
                     for result in raw_search_results
                 ]
-                for future in as_completed(futures):
-                    try:
-                        future.result()
-                    except Exception as e:  # pylint: disable=broad-except
-                        logger.exception("Error fetching/storing document: %s", e)
+                await asyncio.gather(*tasks, return_exceptions=True)
 
-        rag_results = document_store.search(
-            query,
-            results_count=settings.BRAVE_RAG_WEB_SEARCH_CHUNK_NUMBER,
-        )
+                # Perform RAG search
+                rag_results = await document_store.asearch(
+                    query,
+                    results_count=settings.BRAVE_RAG_WEB_SEARCH_CHUNK_NUMBER,
+                )
+                logger.info("RAG search returned:  %s", rag_results)
 
-        ctx.usage += RunUsage(
-            input_tokens=rag_results.usage.prompt_tokens,
-            output_tokens=rag_results.usage.completion_tokens,
-        )
+                ctx.usage += RunUsage(
+                    input_tokens=rag_results.usage.prompt_tokens,
+                    output_tokens=rag_results.usage.completion_tokens,
+                )
 
-        # Map RAG results back to raw search results to include extra_snippets
-        # Suboptimal O(N^2) but N is small...
-        for rag_result in rag_results.data:
-            for result in raw_search_results:
-                if result["url"] == rag_result.url:
-                    result.setdefault("extra_snippets", []).append(rag_result.content)
-                    break
+                # Map RAG results back to raw search results to include extra_snippets
+                for rag_result in rag_results.data:
+                    for result in raw_search_results:
+                        if result["url"] == rag_result.url:
+                            result.setdefault("extra_snippets", []).append(rag_result.content)
+                            break
 
-    return format_tool_return(raw_search_results)
+        except Exception as exc:
+            logger.exception("Error with document store: %s", exc)
+            raise ModelRetry(
+                f"Document storage temporarily failed: {type(exc).__name__}. "
+                "Retrying the operation..."
+            ) from exc
+
+        formatted_result = format_tool_return(raw_search_results)
+
+        # Check if we got any valid results
+        if not formatted_result.return_value:
+            raise ModelRetry("No valid search results were extracted.")
+
+        return formatted_result
+    except (ModelCannotRetry, ModelRetry):
+        # Re-raise these as-is
+        raise
+    except Exception as e:
+        # Unexpected error - stop and inform user
+        logger.exception("Unexpected error in web_search_brave_with_document_backend: %s", e)
+        raise ModelCannotRetry(
+            f"An unexpected error occurred during web search with RAG: {type(e).__name__}. "
+            "You must explain this to the user and not try to answer based on your knowledge."
+        ) from e

src/backend/chat/views.py

@@ -1,6 +1,7 @@
 """Chat API implementation."""
 
 import logging
+import os
 from uuid import uuid4
 
 from django.conf import settings
@@ -178,10 +179,32 @@ class ChatViewSet(  # pylint: disable=too-many-ancestors, abstract-method
                 or self.request.LANGUAGE_CODE  # from the LocaleMiddleware
             ),
         )
-        if protocol == "data":
-            streaming_content = ai_service.stream_data(messages, force_web_search=force_web_search)
-        else:  # Default to 'text' protocol
-            streaming_content = ai_service.stream_text(messages, force_web_search=force_web_search)
+
+        # This environment variable allows switching between sync and async streaming modes
+        # based on the server configuration. Tests run in sync mode (WSGI), while
+        # production uses async mode (Uvicorn ASGI).
+        is_async_mode = os.environ.get("PYTHON_SERVER_MODE", "sync") == "async"
+
+        if is_async_mode:
+            logger.debug("Using ASYNC streaming for chat conversation.")
+            if protocol == "data":
+                streaming_content = ai_service.stream_data_async(
+                    messages, force_web_search=force_web_search
+                )
+            else:  # Default to 'text' protocol
+                streaming_content = ai_service.stream_text_async(
+                    messages, force_web_search=force_web_search
+                )
+        else:
+            logger.debug("Using SYNC streaming for chat conversation.")
+            if protocol == "data":
+                streaming_content = ai_service.stream_data(
+                    messages, force_web_search=force_web_search
+                )
+            else:  # Default to 'text' protocol
+                streaming_content = ai_service.stream_text(
+                    messages, force_web_search=force_web_search
+                )
 
         response = StreamingHttpResponse(
             streaming_content,
@@ -402,7 +425,7 @@ class ChatConversationAttachmentViewSet(
         if settings.POSTHOG_KEY:
             posthog.capture(
                 "item_uploaded",
-                distinct_id=request.user.email,
+                distinct_id=str(request.user.pk),  # same as set by the frontend
                 properties={
                     "id": attachment.pk,
                     "file_name": attachment.file_name,

src/backend/conversations/asgi.py

@@ -0,0 +1,18 @@
+"""
+ASGI config for conversations project.
+
+It exposes the ASGI callable as a module-level variable named ``application``.
+
+For more information on this file, see
+https://docs.djangoproject.com/en/dev/howto/deployment/asgi/
+"""
+
+import os
+
+from configurations.asgi import get_asgi_application
+
+os.environ.setdefault("DJANGO_SETTINGS_MODULE", "conversations.settings")
+os.environ.setdefault("DJANGO_CONFIGURATION", "Development")
+os.environ.setdefault("PYTHON_SERVER_MODE", "async")
+
+application = get_asgi_application()

src/backend/conversations/settings.py

@@ -267,7 +267,7 @@ class Base(BraveSettings, Configuration):
             ("en-us", "English"),
             ("fr-fr", "Français"),
             # ("de-de", "Deutsch"),
-            # ("nl-nl", "Nederlands"),
+            ("nl-nl", "Nederlands"),
             # ("es-es", "Español"),
         )
     )
@@ -313,7 +313,6 @@ class Base(BraveSettings, Configuration):
         "django.middleware.csrf.CsrfViewMiddleware",
         "django.contrib.auth.middleware.AuthenticationMiddleware",
         "posthog.integrations.django.PosthogContextMiddleware",
-        "core.middleware.PostHogMiddleware",
         "django.contrib.messages.middleware.MessageMiddleware",
         "dockerflow.django.middleware.DockerflowMiddleware",
     ]
@@ -483,7 +482,11 @@ class Base(BraveSettings, Configuration):
     THUMBNAIL_ALIASES = {}
 
     # Session
-    SESSION_ENGINE = "django.contrib.sessions.backends.cache"
+    SESSION_ENGINE = values.Value(
+        "django.contrib.sessions.backends.cache",
+        environ_name="SESSION_ENGINE",
+        environ_prefix=None,
+    )
     SESSION_CACHE_ALIAS = "default"
     SESSION_COOKIE_AGE = values.PositiveIntegerValue(
         default=60 * 60 * 12, environ_name="SESSION_COOKIE_AGE", environ_prefix=None
@@ -503,6 +506,7 @@ class Base(BraveSettings, Configuration):
         environ_name="OIDC_RP_CLIENT_SECRET",
         environ_prefix=None,
     )
+    OIDC_OP_URL = values.Value(None, environ_name="OIDC_OP_URL", environ_prefix=None)
     OIDC_OP_JWKS_ENDPOINT = values.Value(environ_name="OIDC_OP_JWKS_ENDPOINT", environ_prefix=None)
     OIDC_OP_AUTHORIZATION_ENDPOINT = values.Value(
         environ_name="OIDC_OP_AUTHORIZATION_ENDPOINT", environ_prefix=None
@@ -627,9 +631,6 @@ class Base(BraveSettings, Configuration):
     LLM_DEFAULT_MODEL_HRID = values.Value(
         "default-model", environ_name="LLM_DEFAULT_MODEL_HRID", environ_prefix=None
     )
-    LLM_ROUTING_MODEL_HRID = values.Value(
-        "default-routing-model", environ_name="LLM_ROUTING_MODEL_HRID", environ_prefix=None
-    )
     LLM_SUMMARIZATION_MODEL_HRID = values.Value(
         "default-summarization-model",
         environ_name="LLM_SUMMARIZATION_MODEL_HRID",
@@ -785,6 +786,21 @@ USER QUESTION:
         environ_name="SUMMARIZATION_SYSTEM_PROMPT",
         environ_prefix=None,
     )
+    SUMMARIZATION_CHUNK_SIZE = values.PositiveIntegerValue(
+        default=20_000,  # Approx 20k words per chunk
+        environ_name="SUMMARIZATION_CHUNK_SIZE",
+        environ_prefix=None,
+    )
+    SUMMARIZATION_OVERLAP_SIZE = values.FloatValue(
+        default=0.05,  # 5% overlap
+        environ_name="SUMMARIZATION_OVERLAP_SIZE",
+        environ_prefix=None,
+    )
+    SUMMARIZATION_CONCURRENT_REQUESTS = values.PositiveIntegerValue(
+        default=5,
+        environ_name="SUMMARIZATION_CONCURRENT_REQUESTS",
+        environ_prefix=None,
+    )
 
     # Tavily API
     TAVILY_API_KEY = values.Value(
@@ -891,6 +907,13 @@ USER QUESTION:
         default=False, environ_name="LANGFUSE_MEDIA_UPLOAD_ENABLED", environ_prefix=None
     )
 
+    # WARNING: Testing purpose only. Do not use in production.
+    WARNING_MOCK_CONVERSATION_AGENT = values.BooleanValue(
+        default=False,
+        environ_name="WARNING_MOCK_CONVERSATION_AGENT",
+        environ_prefix=None,
+    )
+
     # pylint: disable=invalid-name
     @property
     def ENVIRONMENT(self):

src/backend/conversations/wsgi.py

@@ -13,5 +13,6 @@ from configurations.wsgi import get_wsgi_application
 
 os.environ.setdefault("DJANGO_SETTINGS_MODULE", "conversations.settings")
 os.environ.setdefault("DJANGO_CONFIGURATION", "Development")
+os.environ.setdefault("PYTHON_SERVER_MODE", "sync")
 
 application = get_wsgi_application()

src/backend/core/authentication/backends.py

@@ -1,7 +1,6 @@
 """Authentication Backends for the Conversations core app."""
 
 import logging
-import os
 
 from django.conf import settings
 from django.core.exceptions import SuspiciousOperation
@@ -15,19 +14,6 @@ from core.models import DuplicateEmailError
 
 logger = logging.getLogger(__name__)
 
-# Settings renamed warnings
-if os.environ.get("USER_OIDC_FIELDS_TO_FULLNAME"):
-    logger.warning(
-        "USER_OIDC_FIELDS_TO_FULLNAME has been renamed "
-        "to OIDC_USERINFO_FULLNAME_FIELDS please update your settings."
-    )
-
-if os.environ.get("USER_OIDC_FIELD_TO_SHORTNAME"):
-    logger.warning(
-        "USER_OIDC_FIELD_TO_SHORTNAME has been renamed "
-        "to OIDC_USERINFO_SHORTNAME_FIELD please update your settings."
-    )
-
 
 class OIDCAuthenticationBackend(LaSuiteOIDCAuthenticationBackend):
     """Custom OpenID Connect (OIDC) Authentication Backend.

src/backend/core/feature_flags/helpers.py

@@ -9,7 +9,6 @@ User = get_user_model()
 
 try:
     import posthog
-    from posthog.contexts import get_tags
 except ImportError:
     posthog = None
 
@@ -39,8 +38,7 @@ def is_feature_enabled(
     if posthog is not None:
         return posthog.feature_enabled(
             frontend_feature_name(feature_name),
-            user.email,
-            person_properties={"$host": get_tags().get("$host")},
+            str(user.pk),  # same as set by the frontend
         )
 
     # No feature flag manager

src/backend/core/middleware.py

@@ -1,58 +0,0 @@
-"""Custom middleware(s) for the project."""
-
-import json
-import logging
-from urllib.parse import unquote
-
-from django.conf import settings
-from django.core.exceptions import MiddlewareNotUsed
-
-# We are importing posthog here, but it will only be used if the POSTHOG_KEY is set in settings.
-# The settings are checked before any attempt to use posthog.
-try:
-    import posthog
-except ImportError:
-    posthog = None
-
-
-logger = logging.getLogger(__name__)
-
-
-class PostHogMiddleware:
-    """
-    This middleware is used to alias the user's distinct_id from the PostHog cookie
-    with their email address when they are authenticated. This allows us to track
-    users across different sessions and devices.
-    """
-
-    def __init__(self, get_response):
-        """
-        Initialize the middleware and disable it if PostHog is not configured.
-        """
-        if posthog is None or not settings.POSTHOG_KEY:
-            raise MiddlewareNotUsed("POSTHOG_KEY must be set in settings to use PostHogMiddleware.")
-        self.get_response = get_response
-
-    def __call__(self, request):
-        """
-        Process the request to handle the PostHog alias.
-        """
-        if posthog is not None and settings.POSTHOG_KEY:
-            posthog_cookie = request.COOKIES.get(f"ph_{posthog.project_api_key}_posthog")
-            if posthog_cookie:
-                try:
-                    cookie_dict = json.loads(unquote(posthog_cookie))
-                    if (
-                        cookie_dict.get("distinct_id")
-                        and request.user
-                        and request.user.is_authenticated
-                    ):
-                        posthog.alias(cookie_dict["distinct_id"], request.user.email)
-                except (json.JSONDecodeError, KeyError):
-                    # If the cookie is malformed or doesn't contain the expected
-                    # keys, we can't do anything with it, so we ignore it.
-                    logger.warning("Malformed PostHog cookie: %s", posthog_cookie)
-
-        response = self.get_response(request)
-
-        return response

src/backend/core/migrations/0001_initial.py

@@ -118,7 +118,7 @@ class Migration(migrations.Migration):
                             ("en-us", "English"),
                             ("fr-fr", "Français"),
                             # ("de-de", "Deutsch"),
-                            # ("nl-nl", "Nederlands"),
+                            ("nl-nl", "Nederlands"),
                             # ("es-es", "Español"),
                         ],
                         default=None,

src/backend/core/tests/authentication/test_backends.py

@@ -58,7 +58,7 @@ def test_authentication_getter_existing_user_via_email(django_assert_num_queries
 
     monkeypatch.setattr(OIDCAuthenticationBackend, "get_userinfo", get_userinfo_mocked)
 
-    with django_assert_num_queries(3):  # user by sub, user by mail, update sub
+    with django_assert_num_queries(4):  # user by sub, user by mail, unicity check, update sub
         user = klass.get_or_create_user(access_token="test-token", id_token=None, payload=None)
 
     assert user == db_user
@@ -205,7 +205,7 @@ def test_authentication_getter_existing_user_change_fields_sub(
     monkeypatch.setattr(OIDCAuthenticationBackend, "get_userinfo", get_userinfo_mocked)
 
     # One and only one additional update query when a field has changed
-    with django_assert_num_queries(2):
+    with django_assert_num_queries(3):  # user by sub, unicity check, update sub
         authenticated_user = klass.get_or_create_user(
             access_token="test-token", id_token=None, payload=None
         )
@@ -245,7 +245,7 @@ def test_authentication_getter_existing_user_change_fields_email(
     monkeypatch.setattr(OIDCAuthenticationBackend, "get_userinfo", get_userinfo_mocked)
 
     # One and only one additional update query when a field has changed
-    with django_assert_num_queries(3):
+    with django_assert_num_queries(4):  # user by sub, user by mail, unicity check, update sub
         authenticated_user = klass.get_or_create_user(
             access_token="test-token", id_token=None, payload=None
         )

src/backend/core/tests/feature_flags/test_helpers.py

@@ -1,9 +1,12 @@
 """Tests for feature flag helpers."""
 
+import json
 import logging
 from unittest.mock import patch
 
+import posthog
 import pytest
+import responses
 
 from core.factories import UserFactory
 from core.feature_flags.flags import FeatureToggle
@@ -42,19 +45,29 @@ def test_is_feature_enabled_always_disabled(feature_flags):
     assert is_feature_enabled(user, "document_upload") is False
 
 
-@patch("core.feature_flags.helpers.posthog")
-def test_is_feature_enabled_dynamic_posthog_true(mock_posthog, feature_flags):
+@responses.activate
+def test_is_feature_enabled_dynamic_posthog_true(feature_flags, settings):
     """Test that a dynamic feature returns the value from PostHog when PostHog is available."""
+    settings.POSTHOG_KEY = {"id": "132456", "host": "https://eu.i.posthog-test.com"}
+
+    posthog.api_key = settings.POSTHOG_KEY["id"]
+    posthog.host = settings.POSTHOG_KEY["host"]
+
+    responses.post(
+        f"{posthog.host}/flags/?v=2", json={"flags": {"web-search": {"enabled": True}}}, status=200
+    )
+
     feature_flags.web_search = FeatureToggle.DYNAMIC
     user = UserFactory()
 
-    mock_posthog.feature_enabled.return_value = True
     assert is_feature_enabled(user, "web_search") is True
-    mock_posthog.feature_enabled.assert_called_once_with(
-        "web-search",
-        user.email,
-        person_properties={"$host": None},
-    )
+
+    request_body = json.loads(responses.calls[0].request.body)
+    assert request_body["distinct_id"] == str(user.pk)
+    assert request_body["flag_keys_to_evaluate"] == ["web-search"]
+
+    posthog.api_key = None
+    posthog.host = None
 
 
 @patch("core.feature_flags.helpers.posthog")

src/backend/core/tests/test_api_config.py

@@ -4,9 +4,10 @@ Test config API endpoints in the Conversations core app.
 
 import json
 
-from django.test import override_settings
+from django.test import AsyncClient, override_settings
 
 import pytest
+from asgiref.sync import sync_to_async
 from rest_framework.status import (
     HTTP_200_OK,
 )
@@ -53,7 +54,7 @@ def test_api_config(is_authenticated):
             ["en-us", "English"],
             ["fr-fr", "Français"],
             # ["de-de", "Deutsch"],
-            # ["nl-nl", "Nederlands"],
+            ["nl-nl", "Nederlands"],
             # ["es-es", "Español"],
         ],
         "LANGUAGE_CODE": "en-us",
@@ -156,3 +157,52 @@ def test_api_config_with_original_theme_customization(is_authenticated, settings
         theme_customization = json.load(f)
 
     assert content["theme_customization"] == theme_customization
+
+
+@override_settings(
+    CRISP_WEBSITE_ID="123",
+    FRONTEND_CSS_URL="http://testcss/",
+    FRONTEND_THEME="test-theme",
+    MEDIA_BASE_URL="http://testserver/",
+    POSTHOG_KEY={"id": "132456", "host": "https://eu.i.posthog-test.com"},
+    SENTRY_DSN="https://sentry.test/123",
+    THEME_CUSTOMIZATION_FILE_PATH="",
+    RAG_FILES_ACCEPTED_FORMATS=[
+        "application/pdf",
+        "text/plain",
+    ],
+)
+@pytest.mark.asyncio
+@pytest.mark.parametrize("is_authenticated", [False, True])
+async def test_api_config_async(is_authenticated):
+    """Anonymous users should be allowed to get the configuration (async client)."""
+    client = AsyncClient()
+
+    if is_authenticated:
+        user = await sync_to_async(factories.UserFactory)()
+        await client.aforce_login(user)
+
+    response = await client.get("/api/v1.0/config/")
+    assert response.status_code == HTTP_200_OK
+    assert response.json() == {
+        "ACTIVATION_REQUIRED": False,
+        "CRISP_WEBSITE_ID": "123",
+        "ENVIRONMENT": "test",
+        "FEATURE_FLAGS": {"document-upload": "enabled", "web-search": "enabled"},
+        "FRONTEND_CSS_URL": "http://testcss/",
+        "FRONTEND_HOMEPAGE_FEATURE_ENABLED": True,
+        "FRONTEND_THEME": "test-theme",
+        "LANGUAGES": [
+            ["en-us", "English"],
+            ["fr-fr", "Français"],
+            # ["de-de", "Deutsch"],
+            ["nl-nl", "Nederlands"],
+            # ["es-es", "Español"],
+        ],
+        "LANGUAGE_CODE": "en-us",
+        "MEDIA_BASE_URL": "http://testserver/",
+        "POSTHOG_KEY": {"id": "132456", "host": "https://eu.i.posthog-test.com"},
+        "SENTRY_DSN": "https://sentry.test/123",
+        "theme_customization": {},
+        "chat_upload_accept": "application/pdf,text/plain",
+    }

src/backend/locale/de_DE/LC_MESSAGES/django.po

@@ -3,7 +3,7 @@ msgstr ""
 "Project-Id-Version: la-suite-conversations\n"
 "Report-Msgid-Bugs-To: \n"
 "POT-Creation-Date: 2025-10-20 21:48+0000\n"
-"PO-Revision-Date: 2025-10-21 20:46\n"
+"PO-Revision-Date: 2025-11-10 12:20\n"
 "Last-Translator: \n"
 "Language-Team: German\n"
 "Language: de_DE\n"

src/backend/locale/en_US/LC_MESSAGES/django.po

@@ -3,7 +3,7 @@ msgstr ""
 "Project-Id-Version: la-suite-conversations\n"
 "Report-Msgid-Bugs-To: \n"
 "POT-Creation-Date: 2025-10-20 21:48+0000\n"
-"PO-Revision-Date: 2025-10-21 20:46\n"
+"PO-Revision-Date: 2025-11-10 12:20\n"
 "Last-Translator: \n"
 "Language-Team: English\n"
 "Language: en_US\n"

src/backend/locale/fr_FR/LC_MESSAGES/django.po

@@ -3,7 +3,7 @@ msgstr ""
 "Project-Id-Version: la-suite-conversations\n"
 "Report-Msgid-Bugs-To: \n"
 "POT-Creation-Date: 2025-10-20 21:48+0000\n"
-"PO-Revision-Date: 2025-10-21 20:46\n"
+"PO-Revision-Date: 2025-11-10 12:20\n"
 "Last-Translator: \n"
 "Language-Team: French\n"
 "Language: fr_FR\n"
@@ -86,7 +86,7 @@ msgstr "A utilisé le code d'activation"
 
 #: activation_codes/admin.py:293 build/lib/activation_codes/admin.py:293
 msgid "Add selected users to Brevo waiting list"
-msgstr "Ajouter les utilisateurs sélectionnés à la liste d'attente de Brevo"
+msgstr "Ajouter les utilisateurs sélectionnés à la liste d'attente Brevo"
 
 #: activation_codes/admin.py:314 build/lib/activation_codes/admin.py:314
 #, python-format
@@ -272,7 +272,7 @@ msgstr "Nous n'avons pas pu trouver un utilisateur avec ce sous-groupe mais l'e-
 
 #: build/lib/core/models.py:99 core/models.py:99
 msgid "Enter a valid sub. This value may contain only letters, numbers, and @/./+/-/_/: characters."
-msgstr "Saisissez un sous-groupe valide. Cette valeur ne peut contenir que des lettres, des chiffres et les caractères @/./+/-/_/: uniquement."
+msgstr "Saisissez un 'sub' valide. Cette valeur ne peut contenir que des lettres, des chiffres et les caractères @/./+/-/_/: uniquement."
 
 #: build/lib/core/models.py:105 core/models.py:105
 msgid "sub"

src/backend/locale/nl_NL/LC_MESSAGES/django.po

@@ -3,7 +3,7 @@ msgstr ""
 "Project-Id-Version: la-suite-conversations\n"
 "Report-Msgid-Bugs-To: \n"
 "POT-Creation-Date: 2025-10-20 21:48+0000\n"
-"PO-Revision-Date: 2025-10-21 20:46\n"
+"PO-Revision-Date: 2025-11-10 12:20\n"
 "Last-Translator: \n"
 "Language-Team: Dutch\n"
 "Language: nl_NL\n"
@@ -19,228 +19,228 @@ msgstr ""
 
 #: activation_codes/admin.py:55 build/lib/activation_codes/admin.py:55
 msgid "Configuration"
-msgstr ""
+msgstr "Configuratie"
 
 #: activation_codes/admin.py:66 build/lib/activation_codes/admin.py:66
 msgid "Usage details"
-msgstr ""
+msgstr "Gebruiksdetails"
 
 #: activation_codes/admin.py:70 activation_codes/admin.py:226
 #: build/lib/activation_codes/admin.py:70
 #: build/lib/activation_codes/admin.py:226
 msgid "Timestamps"
-msgstr ""
+msgstr "Tijdstempels"
 
 #: activation_codes/admin.py:109 build/lib/activation_codes/admin.py:109
 msgid "Usage"
-msgstr ""
+msgstr "Gebruik"
 
 #: activation_codes/admin.py:117 build/lib/activation_codes/admin.py:117
 msgid "Description"
-msgstr ""
+msgstr "Beschrijving"
 
 #: activation_codes/admin.py:124 build/lib/activation_codes/admin.py:124
 msgid "No users have used this code yet"
-msgstr ""
+msgstr "Er zijn nog geen gebruikers die deze code hebben gebruikt"
 
 #: activation_codes/admin.py:135 build/lib/activation_codes/admin.py:135
 msgid "Name"
-msgstr ""
+msgstr "Naam"
 
 #: activation_codes/admin.py:136 activation_codes/admin.py:246
 #: build/lib/activation_codes/admin.py:136
 #: build/lib/activation_codes/admin.py:246
 msgid "Email"
-msgstr ""
+msgstr "E-mail"
 
 #: activation_codes/admin.py:137 build/lib/activation_codes/admin.py:137
 msgid "Date"
-msgstr ""
+msgstr "Datum"
 
 #: activation_codes/admin.py:161 build/lib/activation_codes/admin.py:161
 msgid "Users who used this code"
-msgstr ""
+msgstr "Gebruikers die deze code hebben gebruikt"
 
 #: activation_codes/admin.py:163 build/lib/activation_codes/admin.py:163
 msgid "Recompute current uses from related activations"
-msgstr ""
+msgstr "Herbereken het huidige gebruik van gerelateerde activeringen"
 
 #: activation_codes/admin.py:177 build/lib/activation_codes/admin.py:177
 msgid "All selected activation codes already have correct usage counts."
-msgstr ""
+msgstr "Alle geselecteerde activeringscodes hebben al het juiste gebruiksaantal."
 
 #: activation_codes/admin.py:182 build/lib/activation_codes/admin.py:182
 #, python-format
 msgid "Successfully recomputed usage counts for %(count)d activation code(s)."
-msgstr ""
+msgstr "Het gebruik van %(count)d activeringscode(s) is opnieuw berekend."
 
 #: activation_codes/admin.py:240 activation_codes/admin.py:284
 #: build/lib/activation_codes/admin.py:240
 #: build/lib/activation_codes/admin.py:284
 msgid "User"
-msgstr ""
+msgstr "Gebruiker"
 
 #: activation_codes/admin.py:291 build/lib/activation_codes/admin.py:291
 msgid "Has used activation code"
-msgstr ""
+msgstr "Heeft activeringscode gebruikt"
 
 #: activation_codes/admin.py:293 build/lib/activation_codes/admin.py:293
 msgid "Add selected users to Brevo waiting list"
-msgstr ""
+msgstr "Voeg geselecteerde gebruikers toe aan de Brevo-wachtlijst"
 
 #: activation_codes/admin.py:314 build/lib/activation_codes/admin.py:314
 #, python-format
 msgid "Added %(count)d user(s) to Brevo waiting list."
-msgstr ""
+msgstr "%(count)d gebruiker(s) toegevoegd aan de Brevo-wachtlijst."
 
 #: activation_codes/admin.py:319 activation_codes/admin.py:347
 #: build/lib/activation_codes/admin.py:319
 #: build/lib/activation_codes/admin.py:347
 msgid "No valid email address found in selected registrations."
-msgstr ""
+msgstr "Er is geen geldig e-mailadres gevonden in de geselecteerde registraties."
 
 #: activation_codes/admin.py:323 build/lib/activation_codes/admin.py:323
 msgid "Remove selected users from Brevo waiting list"
-msgstr ""
+msgstr "Geselecteerde gebruikers van de Brevo-wachtlijst verwijderen"
 
 #: activation_codes/admin.py:342 build/lib/activation_codes/admin.py:342
 #, python-format
 msgid "Removed %(count)d user(s) from Brevo waiting list."
-msgstr ""
+msgstr "%(count)d gebruiker(s) verwijderd van de Brevo-wachtlijst."
 
 #: activation_codes/models.py:38 activation_codes/models.py:85
 #: activation_codes/models.py:178 build/lib/activation_codes/models.py:38
 #: build/lib/activation_codes/models.py:85
 #: build/lib/activation_codes/models.py:178
 msgid "activation code"
-msgstr ""
+msgstr "activeringscode"
 
 #: activation_codes/models.py:39 build/lib/activation_codes/models.py:39
 msgid "The activation code that users will enter"
-msgstr ""
+msgstr "De activeringscode die gebruikers invoeren"
 
 #: activation_codes/models.py:46 build/lib/activation_codes/models.py:46
 msgid "Code must be alphanumeric and contain no spaces or special characters"
-msgstr ""
+msgstr "De code moet alfanumeriek zijn en mag geen spaties of speciale tekens bevatten"
 
 #: activation_codes/models.py:52 build/lib/activation_codes/models.py:52
 msgid "maximum uses"
-msgstr ""
+msgstr "maximaal gebruik"
 
 #: activation_codes/models.py:53 build/lib/activation_codes/models.py:53
 msgid "Maximum number of times this code can be used. 0 means unlimited."
-msgstr ""
+msgstr "Maximaal aantal keren dat deze code kan worden gebruikt. 0 betekent onbeperkt."
 
 #: activation_codes/models.py:58 build/lib/activation_codes/models.py:58
 msgid "current uses"
-msgstr ""
+msgstr "huidig gebruik"
 
 #: activation_codes/models.py:59 build/lib/activation_codes/models.py:59
 msgid "Number of times this code has been used"
-msgstr ""
+msgstr "Aantal keren dat deze code is gebruikt"
 
 #: activation_codes/models.py:65 build/lib/activation_codes/models.py:65
 #: build/lib/core/models.py:151 core/models.py:151
 msgid "active"
-msgstr ""
+msgstr "actief"
 
 #: activation_codes/models.py:66 build/lib/activation_codes/models.py:66
 msgid "Whether this code can still be used"
-msgstr ""
+msgstr "Of deze code nog gebruikt kan worden"
 
 #: activation_codes/models.py:71 build/lib/activation_codes/models.py:71
 msgid "expires at"
-msgstr ""
+msgstr "vervalt op"
 
 #: activation_codes/models.py:72 build/lib/activation_codes/models.py:72
 msgid "Date and time when this code expires"
-msgstr ""
+msgstr "Datum en tijd waarop deze code verloopt"
 
 #: activation_codes/models.py:78 build/lib/activation_codes/models.py:78
 msgid "description"
-msgstr ""
+msgstr "beschrijving"
 
 #: activation_codes/models.py:79 build/lib/activation_codes/models.py:79
 msgid "Internal description or notes about this code"
-msgstr ""
+msgstr "Interne beschrijving of notities over deze code"
 
 #: activation_codes/models.py:86 build/lib/activation_codes/models.py:86
 msgid "activation codes"
-msgstr ""
+msgstr "activeringscodes"
 
 #: activation_codes/models.py:128 build/lib/activation_codes/models.py:128
 msgid "This activation code is no longer valid"
-msgstr ""
+msgstr "Deze activeringscode is niet meer geldig"
 
 #: activation_codes/models.py:136 build/lib/activation_codes/models.py:136
 msgid "You have already activated your account"
-msgstr ""
+msgstr "Je hebt je account al geactiveerd"
 
 #: activation_codes/models.py:170 activation_codes/models.py:202
 #: build/lib/activation_codes/models.py:170
 #: build/lib/activation_codes/models.py:202 build/lib/core/models.py:173
 #: core/models.py:173
 msgid "user"
-msgstr ""
+msgstr "gebruiker"
 
 #: activation_codes/models.py:171 build/lib/activation_codes/models.py:171
 msgid "The user who used the activation code"
-msgstr ""
+msgstr "De gebruiker die de activeringscode heeft gebruikt"
 
 #: activation_codes/models.py:179 build/lib/activation_codes/models.py:179
 msgid "The activation code that was used"
-msgstr ""
+msgstr "De activeringscode die is gebruikt"
 
 #: activation_codes/models.py:186 activation_codes/models.py:210
 #: build/lib/activation_codes/models.py:186
 #: build/lib/activation_codes/models.py:210
 msgid "user activation"
-msgstr ""
+msgstr "gebruikers activering"
 
 #: activation_codes/models.py:187 build/lib/activation_codes/models.py:187
 msgid "user activations"
-msgstr ""
+msgstr "gebruikersactivaties"
 
 #: activation_codes/models.py:203 build/lib/activation_codes/models.py:203
 msgid "The user who made the registration request"
-msgstr ""
+msgstr "De gebruiker die het registratieverzoek heeft gedaan"
 
 #: activation_codes/models.py:211 build/lib/activation_codes/models.py:211
 msgid "Store if the user received an activation code and used it"
-msgstr ""
+msgstr "Opslaan of de gebruiker een activeringscode heeft ontvangen en deze heeft gebruikt"
 
 #: activation_codes/models.py:220 build/lib/activation_codes/models.py:220
 msgid "user registration request"
-msgstr ""
+msgstr "gebruikersregistratieverzoek"
 
 #: activation_codes/models.py:221 build/lib/activation_codes/models.py:221
 msgid "user registration requests"
-msgstr ""
+msgstr "gebruikersregistratieverzoeken"
 
 #: activation_codes/serializers.py:14
 #: build/lib/activation_codes/serializers.py:14
 msgid "The activation code to validate"
-msgstr ""
+msgstr "De activeringscode om te valideren"
 
 #: activation_codes/viewsets.py:107 build/lib/activation_codes/viewsets.py:107
 msgid "Your account has been successfully activated"
-msgstr ""
+msgstr "Uw account is succesvol geactiveerd"
 
 #: build/lib/chat/apps.py:12 chat/apps.py:12
 msgid "chat application"
-msgstr ""
+msgstr "chatapplicatie"
 
 #: build/lib/core/admin.py:26 core/admin.py:26
 msgid "Personal info"
-msgstr ""
+msgstr "Persoonlijke gegevens"
 
 #: build/lib/core/admin.py:40 core/admin.py:40
 msgid "Permissions"
-msgstr ""
+msgstr "Machtigingen"
 
 #: build/lib/core/admin.py:52 core/admin.py:52
 msgid "Important dates"
-msgstr ""
+msgstr "Belangrijke data"
 
 #: build/lib/core/models.py:39 core/models.py:39
 msgid "id"
@@ -248,118 +248,118 @@ msgstr "id"
 
 #: build/lib/core/models.py:40 core/models.py:40
 msgid "primary key for the record as UUID"
-msgstr ""
+msgstr "primaire sleutel voor het record als UUID"
 
 #: build/lib/core/models.py:46 core/models.py:46
 msgid "created on"
-msgstr ""
+msgstr "gemaakt op"
 
 #: build/lib/core/models.py:47 core/models.py:47
 msgid "date and time at which a record was created"
-msgstr ""
+msgstr "datum en tijd waarop een record is aangemaakt"
 
 #: build/lib/core/models.py:52 core/models.py:52
 msgid "updated on"
-msgstr ""
+msgstr "bijgewerkt op"
 
 #: build/lib/core/models.py:53 core/models.py:53
 msgid "date and time at which a record was last updated"
-msgstr ""
+msgstr "datum en tijd waarop een record voor het laatst is bijgewerkt"
 
 #: build/lib/core/models.py:86 core/models.py:86
 msgid "We couldn't find a user with this sub but the email is already associated with a registered user."
-msgstr ""
+msgstr "We konden geen gebruiker met dit e-mailadres vinden, maar het e-mailadres is al gekoppeld aan een geregistreerde gebruiker."
 
 #: build/lib/core/models.py:99 core/models.py:99
 msgid "Enter a valid sub. This value may contain only letters, numbers, and @/./+/-/_/: characters."
-msgstr ""
+msgstr "Voer een geldig subsubtype in. Deze waarde mag alleen letters, cijfers en @/./+/-/_/:-tekens bevatten."
 
 #: build/lib/core/models.py:105 core/models.py:105
 msgid "sub"
-msgstr ""
+msgstr "id"
 
 #: build/lib/core/models.py:107 core/models.py:107
 msgid "Required. 255 characters or fewer. Letters, numbers, and @/./+/-/_/: characters only."
-msgstr ""
+msgstr "Verplicht. Maximaal 255 tekens. Alleen letters, cijfers en @/./+/-/_/: tekens."
 
 #: build/lib/core/models.py:116 core/models.py:116
 msgid "full name"
-msgstr ""
+msgstr "volledige naam"
 
 #: build/lib/core/models.py:117 core/models.py:117
 msgid "short name"
-msgstr ""
+msgstr "korte naam"
 
 #: build/lib/core/models.py:119 core/models.py:119
 msgid "identity email address"
-msgstr ""
+msgstr "identiteits e-mailadres"
 
 #: build/lib/core/models.py:123 core/models.py:123
 msgid "admin email address"
-msgstr ""
+msgstr "beheerders e-mailadres"
 
 #: build/lib/core/models.py:129 core/models.py:129
 msgid "language"
-msgstr ""
+msgstr "taal"
 
 #: build/lib/core/models.py:130 core/models.py:130
 msgid "The language in which the user wants to see the interface."
-msgstr ""
+msgstr "De taal waarin de gebruiker de interface wil zien."
 
 #: build/lib/core/models.py:138 core/models.py:138
 msgid "The timezone in which the user wants to see times."
-msgstr ""
+msgstr "De tijdzone waarin de gebruiker de tijden wil zien."
 
 #: build/lib/core/models.py:141 core/models.py:141
 msgid "device"
-msgstr ""
+msgstr "apparaat"
 
 #: build/lib/core/models.py:143 core/models.py:143
 msgid "Whether the user is a device or a real user."
-msgstr ""
+msgstr "Of de gebruiker een apparaat of een echte gebruiker is."
 
 #: build/lib/core/models.py:146 core/models.py:146
 msgid "staff status"
-msgstr ""
+msgstr "personeelsstatus"
 
 #: build/lib/core/models.py:148 core/models.py:148
 msgid "Whether the user can log into this admin site."
-msgstr ""
+msgstr "Of de gebruiker kan inloggen op deze beheersite."
 
 #: build/lib/core/models.py:154 core/models.py:154
 msgid "Whether this user should be treated as active. Unselect this instead of deleting accounts."
-msgstr ""
+msgstr "Of deze gebruiker als actief moet worden beschouwd. Deselecteer dit in plaats van accounts te verwijderen."
 
 #: build/lib/core/models.py:161 core/models.py:161
 msgid "allow conversation analytics"
-msgstr ""
+msgstr "conversatieanalyse toestaan"
 
 #: build/lib/core/models.py:163 core/models.py:163
 msgid "Whether the user allows to use their conversations for analytics."
-msgstr ""
+msgstr "Of de gebruiker toestaat dat zijn/haar gesprekken voor analyses worden gebruikt."
 
 #: build/lib/core/models.py:174 core/models.py:174
 msgid "users"
-msgstr ""
+msgstr "gebruikers"
 
 #: core/templates/mail/html/invitation.html:162
 #: core/templates/mail/text/invitation.txt:3
 msgid "Logo email"
-msgstr ""
+msgstr "Logo e-mail"
 
 #: core/templates/mail/html/invitation.html:209
 #: core/templates/mail/text/invitation.txt:10
 msgid "Open"
-msgstr ""
+msgstr "Open"
 
 #: core/templates/mail/html/invitation.html:226
 #: core/templates/mail/text/invitation.txt:14
 msgid " Docs, your new essential tool for organizing, sharing and collaborating on your documents as a team. "
-msgstr ""
+msgstr " Docs, uw nieuwe onmisbare hulpmiddel voor het organiseren, delen en samenwerken aan uw documenten als team. "
 
 #: core/templates/mail/html/invitation.html:233
 #: core/templates/mail/text/invitation.txt:16
 #, python-format
 msgid " Brought to you by %(brandname)s "
-msgstr ""
+msgstr " Aangeboden door %(brandname)s "
 

src/backend/locale/ru_RU/LC_MESSAGES/django.po

@@ -3,7 +3,7 @@ msgstr ""
 "Project-Id-Version: la-suite-conversations\n"
 "Report-Msgid-Bugs-To: \n"
 "POT-Creation-Date: 2025-10-20 21:48+0000\n"
-"PO-Revision-Date: 2025-10-21 20:46\n"
+"PO-Revision-Date: 2025-11-10 12:20\n"
 "Last-Translator: \n"
 "Language-Team: Russian\n"
 "Language: ru_RU\n"

src/backend/locale/uk_UA/LC_MESSAGES/django.po

@@ -3,7 +3,7 @@ msgstr ""
 "Project-Id-Version: la-suite-conversations\n"
 "Report-Msgid-Bugs-To: \n"
 "POT-Creation-Date: 2025-10-20 21:48+0000\n"
-"PO-Revision-Date: 2025-10-21 20:46\n"
+"PO-Revision-Date: 2025-11-10 12:20\n"
 "Last-Translator: \n"
 "Language-Team: Ukrainian\n"
 "Language: uk_UA\n"

src/backend/pyproject.toml

@@ -7,7 +7,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "conversations"
-version = "0.0.2"
+version = "0.0.8"
 authors = [{ "name" = "DINUM", "email" = "dev@mail.numerique.gouv.fr" }]
 classifiers = [
     "Development Status :: 5 - Production/Stable",
@@ -27,42 +27,44 @@ requires-python = ">=3.12"
 dependencies = [
     "deprecated",
     "beautifulsoup4==4.14.2",
-    "boto3==1.40.51",
-    "Brotli==1.1.0",
+    "boto3==1.40.67",
+    "Brotli==1.2.0",
     "django-configurations==2.5.1",
     "django-cors-headers==4.9.0",
-    "django-countries==7.6.1",
+    "django-countries==8.0.0",
     "django-filter==25.2",
-    "django-lasuite[all]==0.0.14",
+    "django-lasuite[all]==0.0.17",
     "django-parler==2.3",
     "django-pydantic-field==0.3.13",
     "django-redis==6.0.0",
     "django-storages[s3]==1.14.6",
     "django-timezone-field>=5.1",
-    "django==5.2.7",
+    "django==5.2.8",
     "djangorestframework==3.16.1",
-    "drf_spectacular==0.28.0",
+    "drf_spectacular==0.29.0",
     "dockerflow==2024.4.2",
     "easy_thumbnails==2.10.1",
     "factory_boy==3.3.3",
     "gunicorn==23.0.0",
     "jsonschema==4.25.1",
-    "langfuse==3.6.2",
+    "langfuse==3.9.0",
     "lxml==5.4.0",
-    "markdown==3.9",
+    "markdown==3.10",
     "markitdown==0.0.2",
     "mozilla-django-oidc==4.0.1",
     "nested-multipart-parser==1.6.0",
-    "posthog==6.7.7",
-    "pydantic==2.12.1",
-    "pydantic-ai-slim[openai,mistral,mcp,evals,logfire]==1.0.18",
-    "psycopg[binary]==3.2.10",
+    "posthog==6.8.0",
+    "pydantic==2.12.4",
+    "pydantic-ai-slim[openai,mistral,mcp,evals,logfire]==1.11.0",
+    "psycopg[binary]==3.2.12",
     "PyJWT==2.10.1",
     "python-magic==0.4.27",
     "redis<6.0.0",
     "requests==2.32.5",
-    "sentry-sdk==2.41.0",
+    "semchunk==3.2.5",
+    "sentry-sdk==2.43.0",
     "trafilatura==2.0.0",
+    "uvicorn==0.38.0",
     "whitenoise==6.11.0",
 ]
 
@@ -74,16 +76,17 @@ dependencies = [
 
 [project.optional-dependencies]
 dev = [
+    "dirty-equals==0.10.0",
     "django-extensions==4.1",
     "django-test-migrations==1.5.0",
     "drf-spectacular-sidecar==2025.10.1",
     "freezegun==1.5.5",
     "ipdb==0.13.13",
-    "ipython==9.6.0",
-    "pyfakefs==5.10.0",
+    "ipython==9.7.0",
+    "pyfakefs==5.10.2",
     "pylint-django==2.6.1",
-    "pylint==3.3.8",
-    "pylint-pydantic==0.4.0",
+    "pylint==3.3.9",
+    "pylint-pydantic==0.4.1",
     "pytest-asyncio==1.2.0",
     "pytest-cov==7.0.0",
     "pytest-django==4.11.1",
@@ -92,7 +95,7 @@ dev = [
     "pytest-xdist==3.8.0",
     "responses==0.25.8",
     "respx==0.22.0",
-    "ruff==0.14.0",
+    "ruff==0.14.3",
     "types-requests==2.32.4.20250913",
 ]
 

src/frontend/apps/conversations/package.json

@@ -1,6 +1,6 @@
 {
   "name": "app-conversations",
-  "version": "0.0.2",
+  "version": "0.0.8",
   "private": true,
   "scripts": {
     "dev": "next dev",

src/frontend/apps/conversations/src/components/CustomToast.tsx

@@ -3,6 +3,7 @@ import { css } from 'styled-components';
 
 import { Box, Text } from '@/components';
 import { Icon } from '@/components/Icon';
+import { useResponsiveStore } from '@/stores';
 
 export type ToastType = 'success' | 'error' | 'info' | 'warning';
 
@@ -13,6 +14,8 @@ export interface ToastProps {
   icon?: string;
   duration?: number;
   onClose: (id: string) => void;
+  actionLabel?: string;
+  actionHref?: string;
 }
 
 const getToastConfig = (type: ToastType) => {
@@ -62,11 +65,14 @@ export const Toast = ({
   icon,
   duration = 4000,
   onClose,
+  actionLabel,
+  actionHref,
 }: ToastProps) => {
   const [isVisible, setIsVisible] = useState(false);
   const [isLeaving, setIsLeaving] = useState(false);
   const config = getToastConfig(type);
   const iconToUse = icon || config.icon;
+  const { isMobile } = useResponsiveStore();
 
   useEffect(() => {
     setIsVisible(true);
@@ -102,7 +108,12 @@ export const Toast = ({
         overflow: hidden;
       `}
     >
-      <Box $direction="row" $align="center" $gap="12px">
+      <Box
+        $direction="row"
+        $align="center"
+        $gap="12px"
+        $justify="space-between"
+      >
         <Icon
           iconName={iconToUse}
           $variation="600"
@@ -111,16 +122,41 @@ export const Toast = ({
             color: ${config.color} !important;
           `}
         />
-        <Text
-          $weight="500"
-          $size="14px"
-          $css={css`
-            color: ${config.color} !important;
-            padding: 4px;
-          `}
+        <Box
+          $direction="row"
+          $align="center"
+          $gap="12px"
+          $flex={1}
+          $justify="space-between"
         >
-          {message}
-        </Text>
+          <Text
+            $weight="500"
+            $size="14px"
+            $css={css`
+              color: ${config.color} !important;
+              padding: 4px;
+            `}
+          >
+            {message}
+          </Text>
+
+          {actionLabel && actionHref && !isMobile && (
+            <a
+              href={actionHref}
+              target="_blank"
+              rel="noopener noreferrer"
+              style={{
+                color: config.color,
+                fontWeight: '500',
+                fontSize: '14px',
+                textDecoration: 'underline',
+                whiteSpace: 'nowrap',
+              }}
+            >
+              {actionLabel}
+            </a>
+          )}
+        </Box>
       </Box>
     </Box>
   );

src/frontend/apps/conversations/src/components/DropButton.tsx

@@ -28,10 +28,16 @@ const StyledButton = styled(Button)<StyledButtonProps>`
   border: none;
   background: none;
   outline: none;
-  transition: all 0.2s ease-in-out;
+  transition: background 0.2s ease-in-out;
   font-weight: 500;
   font-size: 0.938rem;
   padding: 0;
+
+  &:focus-visible {
+    outline: 2px solid #3e5de7;
+    outline-offset: 2px;
+  }
+
   ${({ $css }) => $css};
 `;
 

src/frontend/apps/conversations/src/components/DropdownMenu.tsx

@@ -102,6 +102,7 @@ export const DropdownMenu = ({
               data-testid={option.testId}
               $direction="row"
               disabled={isDisabled}
+              tabIndex={isDisabled ? -1 : 0}
               onClick={(event) => {
                 event.preventDefault();
                 event.stopPropagation();
@@ -134,9 +135,15 @@ export const DropdownMenu = ({
                 cursor: ${isDisabled ? 'not-allowed' : 'pointer'};
                 user-select: none;
 
-                &:hover {
+                &:hover,
+                &:focus {
                   background-color: var(--c--theme--colors--greyscale-050);
                 }
+
+                &:focus-visible {
+                  outline: 2px solid #3e5de7;
+                  outline-offset: -2px;
+                }
               `}
             >
               <Box

src/frontend/apps/conversations/src/components/Icon.tsx

@@ -15,6 +15,7 @@ export const Icon = ({
   return (
     <Text
       {...textProps}
+      aria-hidden="true"
       className={clsx('--docs--icon-bg', textProps.className, {
         'material-symbols': variant === 'filled',
         'material-symbols-outlined': variant === 'outlined',
@@ -33,6 +34,7 @@ export const IconOptions = ({ isHorizontal, ...props }: IconOptionsProps) => {
   return (
     <Icon
       {...props}
+      aria-hidden="true"
       iconName={isHorizontal ? 'more_horiz' : 'more_vert'}
       $css={css`
         user-select: none;

src/frontend/apps/conversations/src/components/ToastProvider.tsx

@@ -17,6 +17,8 @@ interface ToastItem {
   type: ToastType;
   icon?: string;
   duration?: number;
+  actionLabel?: string;
+  actionHref?: string;
 }
 
 interface ToastContextType {
@@ -25,6 +27,7 @@ interface ToastContextType {
     message: string,
     icon?: string,
     duration?: number,
+    options?: { actionLabel?: string; actionHref?: string },
   ) => void;
 }
 
@@ -46,9 +49,23 @@ export const ToastProvider = ({ children }: ToastProviderProps) => {
   const [toasts, setToasts] = useState<ToastItem[]>([]);
 
   const showToast = useCallback(
-    (type: ToastType, message: string, icon?: string, duration = 4000) => {
+    (
+      type: ToastType,
+      message: string,
+      icon?: string,
+      duration = 4000,
+      options?: { actionLabel?: string; actionHref?: string },
+    ) => {
       const id = Math.random().toString(36).substr(2, 9);
-      const newToast: ToastItem = { id, message, type, icon, duration };
+      const newToast: ToastItem = {
+        id,
+        message,
+        type,
+        icon,
+        duration,
+        actionLabel: options?.actionLabel,
+        actionHref: options?.actionHref,
+      };
 
       setToasts((prev) => [newToast, ...prev]);
     },
@@ -69,6 +86,7 @@ export const ToastProvider = ({ children }: ToastProviderProps) => {
       {typeof window !== 'undefined' &&
         createPortal(
           <Box
+            aria-live="polite"
             $css={`
             position: fixed;
             top: 8px;
@@ -94,6 +112,8 @@ export const ToastProvider = ({ children }: ToastProviderProps) => {
                 type={toast.type}
                 icon={toast.icon}
                 duration={toast.duration}
+                actionLabel={toast.actionLabel}
+                actionHref={toast.actionHref}
                 onClose={removeToast}
               />
             ))}

src/frontend/apps/conversations/src/features/chat/components/Chat.tsx

@@ -603,6 +603,7 @@ export const Chat = ({
                     display: flex;
                     width: 100%;
                     margin: auto;
+                    margin-bottom: ${isLastAssistantMessageInConversation ? '30px' : '0px'};
                     color: var(--c--theme--colors--greyscale-850);
                     padding-left: 12px;
                     padding-right: 12px;
@@ -643,6 +644,11 @@ export const Chat = ({
                           className="mainContent-chat"
                           $padding={{ all: 'xxs' }}
                         >
+                          <p className="sr-only">
+                            {message.role === 'user'
+                              ? t('You said: ')
+                              : t('Assistant IA replied: ')}
+                          </p>
                           <MarkdownHooks
                             remarkPlugins={[remarkGfm, remarkMath]}
                             rehypePlugins={[
@@ -659,12 +665,18 @@ export const Chat = ({
                               // eslint-disable-next-line @typescript-eslint/no-unused-vars
                               p: ({ node, ...props }) => (
                                 <Text
+                                  as="p"
                                   $css="display: block"
                                   $theme="greyscale"
                                   $variation="850"
                                   {...props}
                                 />
                               ),
+                              a: ({ children, ...props }) => (
+                                <a target="_blank" {...props}>
+                                  {children}
+                                </a>
+                              ),
                             }}
                           >
                             {message.content}
@@ -696,7 +708,23 @@ export const Chat = ({
                             >
                               <Loader />
                               <Text $variation="600" $size="md">
-                                {t('Search...')}
+                                {(() => {
+                                  const toolInvocation = message.parts?.find(
+                                    (part) =>
+                                      part.type === 'tool-invocation' &&
+                                      part.toolInvocation.toolName !==
+                                        'document_parsing',
+                                  );
+                                  if (
+                                    toolInvocation?.type ===
+                                      'tool-invocation' &&
+                                    toolInvocation.toolInvocation.toolName ===
+                                      'summarize'
+                                  ) {
+                                    return t('Summarizing...');
+                                  }
+                                  return t('Search...');
+                                })()}
                               </Text>
                             </Box>
                           )}
@@ -789,7 +817,7 @@ export const Chat = ({
                                       $direction="row"
                                       $align="center"
                                       $gap="4px"
-                                      className={`c__button--neutral action-chat-button ${isSourceOpen ? 'action-chat-button--open' : ''}`}
+                                      className={`c__button--neutral action-chat-button ${isSourceOpen === message.id ? 'action-chat-button--open' : ''}`}
                                       onClick={() => openSources(message.id)}
                                       onKeyDown={(e) => {
                                         if (
@@ -888,7 +916,6 @@ export const Chat = ({
         {status === 'error' && (
           <Box
             $direction={isMobile ? 'column' : 'row'}
-            $align="center"
             $gap="6px"
             $width="100%"
             $maxWidth="750px"

src/frontend/apps/conversations/src/features/chat/components/InputChat.tsx

@@ -1,8 +1,9 @@
 import { Button } from '@openfun/cunningham-react';
-import React, { useEffect, useRef, useState } from 'react';
+import React, { useCallback, useEffect, useRef, useState } from 'react';
 import { useTranslation } from 'react-i18next';
 
 import { Box, Icon, Text } from '@/components';
+import { useToast } from '@/components/ToastProvider';
 import { FeatureFlagState, useConfig } from '@/core';
 import { LLMModel } from '@/features/chat/api/useLLMConfiguration';
 import { useAnalytics } from '@/libs';
@@ -51,6 +52,7 @@ export const InputChat = ({
   isUploadingFiles = false,
 }: InputChatProps) => {
   const { t } = useTranslation();
+  const { showToast } = useToast();
   const fileInputRef = useRef<HTMLInputElement>(null);
   const textareaRef = useRef<HTMLTextAreaElement>(null);
   const [isDragActive, setIsDragActive] = useState(false);
@@ -62,6 +64,29 @@ export const InputChat = ({
   const [webSearchEnabled, setWebSearchEnabled] = useState(false);
   const [isResetting, setIsResetting] = useState(false);
 
+  const isFileAccepted = useCallback(
+    (file: File): boolean => {
+      const acceptedConfig = conf?.chat_upload_accept;
+      if (!acceptedConfig) {
+        return true;
+      }
+      const acceptedTypes = acceptedConfig
+        .split(',')
+        .map((type) => type.trim());
+      return acceptedTypes.some((acceptedType) => {
+        if (acceptedType.startsWith('.')) {
+          return file.name.toLowerCase().endsWith(acceptedType.toLowerCase());
+        }
+        if (acceptedType.endsWith('/*')) {
+          const baseType = acceptedType.slice(0, -2);
+          return file.type.startsWith(baseType);
+        }
+        return file.type === acceptedType;
+      });
+    },
+    [conf?.chat_upload_accept],
+  );
+
   const suggestions = [
     t('Ask a question'),
     t('Turn this list into bullet points'),
@@ -69,6 +94,20 @@ export const InputChat = ({
     t('Find recent news about...'),
   ];
 
+  const showToastError = useCallback(() => {
+    showToast(
+      'error',
+      `${t('File type not supported')}`,
+      undefined,
+      undefined,
+      {
+        actionLabel: t('Know more'),
+        actionHref:
+          'https://docs.numerique.gouv.fr/docs/060b7b70-15aa-4d9a-86f5-2d31c3d693d5/',
+      },
+    );
+  }, [showToast, t]);
+
   useEffect(() => {
     if (!conf?.FEATURE_FLAGS) {
       setWebSearchEnabled(false);
@@ -152,6 +191,20 @@ export const InputChat = ({
 
     const handleDragOver = (e: DragEvent) => {
       e.preventDefault();
+
+      // Check for rejected files during drag over (does not work on Safari)
+      if (e.dataTransfer?.items) {
+        const items = Array.from(e.dataTransfer.items);
+        items.some((item) => {
+          if (item.kind === 'file') {
+            // Check file type
+            const type = item.type;
+            const dummyFile = new File([], '', { type });
+            return !isFileAccepted(dummyFile);
+          }
+          return false;
+        });
+      }
     };
 
     const handleDrop = (e: DragEvent) => {
@@ -164,12 +217,31 @@ export const InputChat = ({
 
       const droppedFiles = e.dataTransfer?.files;
       if (droppedFiles && droppedFiles.length > 0) {
+        const acceptedFiles: File[] = [];
+        const rejectedFiles: string[] = [];
+
+        Array.from(droppedFiles).forEach((file) => {
+          if (isFileAccepted(file)) {
+            acceptedFiles.push(file);
+          } else {
+            rejectedFiles.push(file.name);
+          }
+        });
+
+        if (rejectedFiles.length > 0) {
+          showToastError();
+        }
+
+        if (acceptedFiles.length === 0) {
+          return;
+        }
+
         setFiles((prev) => {
           const dt = new DataTransfer();
           if (prev) {
             Array.from(prev).forEach((f) => dt.items.add(f));
           }
-          Array.from(droppedFiles).forEach((f) => {
+          acceptedFiles.forEach((f) => {
             if (
               !Array.from(prev || []).some(
                 (pf) =>
@@ -197,7 +269,13 @@ export const InputChat = ({
       window.removeEventListener('dragover', handleDragOver);
       window.removeEventListener('drop', handleDrop);
     };
-  }, [fileUploadEnabled, setFiles]);
+  }, [
+    fileUploadEnabled,
+    setFiles,
+    showToastError,
+    conf?.chat_upload_accept,
+    isFileAccepted,
+  ]);
 
   const isInputDisabled = status !== 'ready' || isUploadingFiles;
 
@@ -325,6 +403,7 @@ export const InputChat = ({
               )}
               <textarea
                 ref={textareaRef}
+                aria-label={t('Enter your message or a question')}
                 value={input ?? ''}
                 name="inputchat-textarea"
                 onChange={(e) => {
@@ -420,12 +499,33 @@ export const InputChat = ({
                   if (!fileList) {
                     return;
                   }
+
+                  const acceptedFiles: File[] = [];
+                  const rejectedFiles: string[] = [];
+
+                  Array.from(fileList).forEach((file) => {
+                    if (isFileAccepted(file)) {
+                      acceptedFiles.push(file);
+                    } else {
+                      rejectedFiles.push(file.name);
+                    }
+                  });
+
+                  if (rejectedFiles.length > 0) {
+                    showToastError();
+                  }
+
+                  if (acceptedFiles.length === 0) {
+                    e.target.value = '';
+                    return;
+                  }
+
                   setFiles((prev) => {
                     const dt = new DataTransfer();
                     if (prev) {
                       Array.from(prev).forEach((f: File) => dt.items.add(f));
                     }
-                    Array.from(fileList).forEach((f: File) => {
+                    acceptedFiles.forEach((f: File) => {
                       if (
                         !Array.from(prev || []).some(
                           (pf) =>
@@ -439,6 +539,8 @@ export const InputChat = ({
                     });
                     return dt.files;
                   });
+
+                  e.target.value = '';
                 }}
               />
               {/*Aperçu des fichiers*/}

src/frontend/apps/conversations/src/features/chat/components/SourceItem.tsx

@@ -151,7 +151,7 @@ export const SourceItem: React.FC<SourceItemProps> = ({ url, metadata }) => {
     <Box $direction="row" $gap="4px" $align="center">
       <Box
         $direction="row"
-        $align="center"
+        $align="start"
         $css="font-size: 14px;"
         $width="100%"
       >
@@ -168,6 +168,9 @@ export const SourceItem: React.FC<SourceItemProps> = ({ url, metadata }) => {
                 padding: 4px;
                 width: 100%;
                 text-decoration: none;
+                white-space: nowrap;
+                overflow: hidden;
+                text-overflow: ellipsis;
                 background-color: transparent;
                 transition: background-color 0.3s;
                 color: var(--c--theme--colors--greyscale-500);

src/frontend/apps/conversations/src/features/feedback/Feedback.tsx

@@ -54,7 +54,7 @@ export const Feedback = (_props: { buttonProps?: Partial<ButtonProps> }) => {
             />
 
             <FeedbackButton
-              href="https://www.tchap.gouv.fr/#/room/!eAHyPLdVHMxNhKAbaC:agent.dinum.tchap.gouv.fr"
+              href="https://tchap.gouv.fr/#/room/!eAHyPLdVHMxNhKAbaC:agent.dinum.tchap.gouv.fr?via=agent.dinum.tchap.gouv.fr&via=agent.culture.tchap.gouv.fr&via=agent.education.tchap.gouv.fr"
               icon={<TchapIcon />}
               title={t('Write on Tchap')}
               description={t('Direct exchange with our team')}

src/frontend/apps/conversations/src/features/header/components/ButtonToggleLeftPanel.tsx

@@ -6,13 +6,15 @@ import { useChatPreferencesStore } from '@/features/chat/stores/useChatPreferenc
 
 export const ButtonToggleLeftPanel = () => {
   const { t } = useTranslation();
-  const { togglePanel } = useChatPreferencesStore();
+  const { isPanelOpen, togglePanel } = useChatPreferencesStore();
 
   return (
     <Button
       size="medium"
       onClick={() => togglePanel()}
-      aria-label={t('Open the header menu')}
+      aria-label={
+        isPanelOpen ? t('Close the left panel') : t('Open the left panel')
+      }
       color="primary-text"
       icon={<LeftPanelIcon />}
       className="--docs--button-toggle-panel"

src/frontend/apps/conversations/src/features/header/components/ButtonTogglePanel.tsx

@@ -12,7 +12,7 @@ export const ButtonTogglePanel = () => {
     <Button
       size="medium"
       onClick={() => togglePanel()}
-      aria-label={t('Open the header menu')}
+      aria-label={isPanelOpen ? t('Close the menu') : t('Open the menu')}
       color="primary-text"
       icon={<Icon $theme="primary" iconName={isPanelOpen ? 'close' : 'menu'} />}
       className="mobile-no-focus"

src/frontend/apps/conversations/src/features/language/components/LanguagePicker.tsx

@@ -41,6 +41,7 @@ export const LanguagePicker = () => {
     <DropdownMenu
       options={optionsPicker}
       showArrow={isDesktop}
+      label={t('Language')}
       buttonCss={css`
         &:hover {
           background-color: var(
@@ -61,7 +62,6 @@ export const LanguagePicker = () => {
       <Text
         $theme="primary"
         $size="md"
-        aria-label={t('Language')}
         $direction="row"
         $gap="0.5rem"
         className="--docs--language-picker-text"

src/frontend/apps/conversations/src/features/left-panel/components/ConversationItemActions.tsx

@@ -2,7 +2,7 @@ import { Button as _Button, useModal } from '@openfun/cunningham-react';
 import { useTranslation } from 'react-i18next';
 import { css } from 'styled-components';
 
-import { Box, DropdownMenu, DropdownMenuOption, Icon } from '@/components';
+import { DropdownMenu, DropdownMenuOption, Icon } from '@/components';
 import { ChatConversation } from '@/features/chat/types';
 
 import { ModalRemoveConversation } from './ModalRemoveConversation';
@@ -30,47 +30,39 @@ export const ConversationItemActions = ({
 
   return (
     <>
-      <DropdownMenu options={options}>
-        <Box
-          role="button"
-          tabIndex={0}
-          aria-label={t('Conversation actions')}
-          aria-haspopup="menu"
-          aria-expanded="false"
-          onKeyDown={(e) => {
-            if (e.key === 'Enter' || e.key === ' ') {
-              e.preventDefault();
-              // Le DropdownMenu gère l'ouverture
-            }
-          }}
+      <DropdownMenu
+        options={options}
+        label={t('Actions list for conversation {{title}}', {
+          title: conversation.title || t('Untitled conversation'),
+        })}
+        buttonCss={css`
+          display: flex;
+          align-items: center;
+          justify-content: center;
+          width: 24px;
+          height: 24px;
+          padding: 4px;
+          border-radius: 4px;
+          &:hover {
+            background-color: #e1e3e7 !important;
+          }
+          &:focus-visible {
+            outline: 2px solid #3e5de7;
+            outline-offset: 2px;
+          }
+        `}
+      >
+        <Icon
+          data-testid={`conversation-item-actions-button-${conversation.id}`}
+          iconName="more_horiz"
+          $theme="primary"
+          $variation="600"
           $css={css`
-            display: block;
-            width: 24px;
-            height: 24px;
-            padding: 4px;
-            border-radius: 4px;
-            cursor: pointer;
-            &:hover {
-              background-color: #e1e3e7 !important;
-            }
-            &:focus-visible {
-              outline: 2px solid #3e5de7;
-              outline-offset: 2px;
-            }
+            font-size: 1rem;
+            color: var(--c--theme--colors--primary-text-text);
+            pointer-events: none;
           `}
-        >
-          <Icon
-            data-testid={`conversation-item-actions-button-${conversation.id}`}
-            iconName="more_horiz"
-            $theme="primary"
-            $variation="600"
-            $css={css`
-              font-size: 1rem;
-              color: var(--c--theme--colors--primary-text-text);
-              pointer-events: none;
-            `}
-          />
-        </Box>
+        />
       </DropdownMenu>
 
       {deleteModal.isOpen && (

src/frontend/apps/conversations/src/features/left-panel/components/LeftPanelConversationItem.tsx

@@ -44,12 +44,15 @@ export const LeftPanelConversationItem = ({
           background-color: transparent
           transition: all 0.3s cubic-bezier(1, 0, 0, 1);
         }
-        &:hover, &:focus {
+        &:hover, &:focus, &:focus-within {
           background-color: #ebedf1;
           .pinned-actions {
             opacity: 1;
           }
         }
+        .pinned-actions:focus-within {
+          opacity: 1;
+        }
       `}
       className="--docs--left-panel-favorite-item"
     >

src/frontend/apps/conversations/src/features/left-panel/components/LeftPanelHeader.tsx

@@ -38,6 +38,7 @@ export const LeftPanelHeader = ({ children }: PropsWithChildren) => {
             >
               <Box $direction="row" $gap="2px">
                 <Button
+                  aria-label={t('New chat')}
                   color="primary"
                   icon={<NewChatIcon />}
                   onClick={goToHome}

src/frontend/apps/conversations/src/features/left-panel/components/ModalRemoveConversation.tsx

@@ -3,7 +3,7 @@ import { t } from 'i18next';
 import { usePathname } from 'next/navigation';
 import { useRouter } from 'next/router';
 
-import { Box, Text, TextErrors, useToast } from '@/components';
+import { Box, Text, useToast } from '@/components';
 import { useRemoveConversation } from '@/features/chat/api/useRemoveConversation';
 import { ChatConversation } from '@/features/chat/types';
 
@@ -20,11 +20,7 @@ export const ModalRemoveConversation = ({
   const { push } = useRouter();
   const pathname = usePathname();
 
-  const {
-    mutate: removeDoc,
-    isError,
-    error,
-  } = useRemoveConversation({
+  const { mutate: removeDoc } = useRemoveConversation({
     onSuccess: () => {
       showToast(
         'success',
@@ -45,6 +41,7 @@ export const ModalRemoveConversation = ({
       isOpen
       closeOnClickOutside
       onClose={() => onClose()}
+      aria-label={t('Content modal to delete conversation')}
       rightActions={
         <>
           <Button
@@ -82,17 +79,10 @@ export const ModalRemoveConversation = ({
         </Text>
       }
     >
-      <Box
-        aria-label={t('Content modal to delete conversation')}
-        className="--docs--modal-remove-doc"
-      >
-        {!isError && (
-          <Text $size="sm" $variation="600">
-            {t('Are you sure you want to delete this conversation ?')}
-          </Text>
-        )}
-
-        {isError && <TextErrors causes={error.cause} />}
+      <Box className="--converstions--modal-remove-chat">
+        <Text $size="sm" $variation="600">
+          {t('Are you sure you want to delete this conversation ?')}
+        </Text>
       </Box>
     </Modal>
   );

src/frontend/apps/conversations/src/i18n/translations.json

@@ -9,12 +9,14 @@
       "Access Denied - Error 403": "Accès refusé - Erreur 403",
       "Access is limited to people who have an invitation code. If you have one, please enter it below.": "L'accès est limité aux personnes qui ont un code d'invitation. Si vous en avez un, veuillez le saisir ci-dessous.",
       "Account activated successfully!": "Compte activé avec succès !",
+      "Actions list for conversation {{title}}": "Liste des actions pour la conversation {{title}}",
       "Add attach file": "Ajouter une pièce jointe",
       "Add file": "Ajouter un fichier",
       "Allow conversation analysis": "Autoriser l'analyse de conversation",
       "An error occurred. Please try again.": "Une erreur s'est produite. Veuillez réessayer.",
       "Are you sure you want to delete this conversation ?": "Êtes-vous sûr de vouloir supprimer cette conversation ?",
       "Ask a question": "Poser une question",
+      "Assistant IA replied: ": "Assistant IA a répondu : ",
       "Assistant is already available, log in to use it now.": "L'Assistant est déjà disponible, connectez-vous pour l'utiliser maintenant.",
       "Assistant is in development: your feedback matters! Choose how to share your ideas:": "L'assistant est en cours de développement : vos commentaires sont importants ! Choisissez comment partager vos avis :",
       "Assistant settings": "Paramètres de l'Assistant",
@@ -24,10 +26,11 @@
       "Cancel": "Annuler",
       "Clear search": "Effacer la recherche",
       "Close model selector": "Fermer le sélecteur de modèle",
+      "Close the left panel": "Fermer le panneau de gauche",
+      "Close the menu": "Fermer le menu",
       "Close the modal": "Fermer la modale",
       "Confirm deletion": "Confirmer la suppression",
       "Content modal to delete conversation": "Modale pour supprimer la conversation",
-      "Conversation actions": "Actions de conversation",
       "Conversation analysis disabled": "Analyse de la conversation désactivée",
       "Conversation analysis enabled": "Analyse de la conversation activée",
       "Copied": "Copié",
@@ -37,6 +40,7 @@
       "Delete a conversation": "Supprimer une conversation",
       "Delete chat": "Supprimer la conversation",
       "Direct exchange with our team": "Échange direct avec notre équipe",
+      "Enter your message or a question": "Entrez votre message ou une question",
       "Explore other LaSuite apps": "Explorer les autres applications de LaSuite",
       "Failed to activate account. Please try again.": "Échec de l'activation du compte. Veuillez réessayer.",
       "Failed to copy": "Échec de la copie",
@@ -47,6 +51,7 @@
       "Failed to upload files. Please try again.": "Le téléversement a échoué. Veuillez réessayer.",
       "Feedback Négatif": "Retour négatif",
       "Feedback positif": "Retour positif",
+      "File type not supported": "Type de fichier non pris en charge",
       "Find recent news about...": "Trouver les dernières actualités concernant...",
       "Get notified about the Public Beta.": "Soyez informé de la Bêta publique.",
       "Get notified for the public beta": "Être notifié pour la bêta publique",
@@ -74,7 +79,8 @@
       "No conversation found": "Aucune conversation trouvée",
       "Notify me": "Me notifier",
       "Open": "Ouvrir",
-      "Open the header menu": "Ouvrir le menu d'en-tête",
+      "Open the left panel": "Ouvrir le panneau de gauche",
+      "Open the menu": "Ouvrir le menu",
       "Page Not Found - Error 404": "Page introuvable - Erreur 404",
       "Please enter an activation code": "Veuillez entrer un code d’activation",
       "Proconnect Login": "Connexion Proconnect",
@@ -96,6 +102,7 @@
       "Start a new conversation.": "Commencer une nouvelle conversation.",
       "Start conversation": "Entamer la conversation",
       "Stop": "Stop",
+      "Summarizing...": "Résumé en cours...",
       "The Assistant is a sovereign conversational AI designed for public servants. It helps you save time on daily tasks like rephrasing, summarising, translating, or searching information. Your data never leaves France and is stored on secure, state-compliant infrastructures. It is never used for commercial purposes.": "L'Assistant est une IA souveraine conçue pour les fonctionnaires. Il vous permet de gagner du temps sur des tâches quotidiennes telles que la reformulation, le résumé, la traduction ou la recherche d'informations. Vos données ne quittent jamais la France et sont stockées sur des infrastructures sûres et conformes à l'état et ne sont jamais utilisées à des fins commerciales.",
       "The Assistant is in Beta": "L'Assistant est en Bêta",
       "The conversation has been deleted.": "La conversation a été supprimée.",
@@ -116,6 +123,7 @@
       "Write on Tchap": "Écrire sur Tchap",
       "You are on the list": "Vous êtes dans la liste",
       "You do not have permission to view this page.": "Vous n’avez pas la permission de voir cette page.",
+      "You said: ": "Vous avez dit : ",
       "You will be notified!": "Vous serez notifié !",
       "Your account is already activated.": "Votre compte est déjà activé.",
       "Your sovereign AI assistant": "Votre assistant IA souverain",
@@ -124,7 +132,137 @@
       "{{productName}} Logo": "Logo {{productName}}"
     }
   },
-  "nl": { "translation": { "ABC-1234-XY": "ABC-1234-XY" } },
+  "nl": {
+    "translation": {
+      "30 sec to tell us what you think or report a bug": "30 seconden om ons te vertellen wat u ervan vindt of een bug te melden",
+      "A privacy-first assistant built for French public teams. Natively synced with LaSuite apps to help you draft, search, and decide without leaving your workflow. Beta access is available with a referral code.": "Een privacygerichte assistent, speciaal ontwikkeld voor teams. Native gesynchroniseerd met apps om je te helpen bij het opstellen, zoeken en beslissen zonder je workflow te verlaten. Bètatoegang is beschikbaar met een verwijzingscode.",
+      "ABC-1234-XY": "ABC-1234-XY",
+      "Access Denied - Error 403": "Toegang geweigerd - Fout 403",
+      "Access is limited to people who have an invitation code. If you have one, please enter it below.": "Toegang is beperkt tot personen met een uitnodigingscode. Als u die heeft, voer deze dan hieronder in.",
+      "Account activated successfully!": "Account succesvol geactiveerd!",
+      "Actions list for conversation {{title}}": "Actielijst voor gesprek {{title}}",
+      "Add attach file": "Voeg een bijlage toe",
+      "Add file": "Bestand toevoegen",
+      "Allow conversation analysis": "Gespreksanalyse toestaan",
+      "An error occurred. Please try again.": "Er is een fout opgetreden. Probeer het opnieuw.",
+      "Are you sure you want to delete this conversation ?": "Weet u zeker dat u dit gesprek wilt verwijderen?",
+      "Ask a question": "Stel een vraag",
+      "Assistant IA replied: ": "AI Assistent antwoordde: ",
+      "Assistant is already available, log in to use it now.": "Assistent is al beschikbaar, log in om het nu te gebruiken.",
+      "Assistant is in development: your feedback matters! Choose how to share your ideas:": "Assistent is in ontwikkeling: uw feedback telt! Kies hoe u uw ideeën wilt delen:",
+      "Assistant settings": "Assistentinstellingen",
+      "Attach file": "Bestand bijvoegen",
+      "Attachment summary not supported": "Bijlageoverzicht niet ondersteund",
+      "Available soon": "Binnenkort beschikbaar",
+      "Cancel": "Annuleren",
+      "Clear search": "Zoekopdracht wissen",
+      "Close model selector": "Sluit modelselector",
+      "Close the left panel": "Sluit het linker venster",
+      "Close the menu": "Sluit het menu",
+      "Close the modal": "Sluit het venster",
+      "Confirm deletion": "Verwijdering bevestigen",
+      "Content modal to delete conversation": "Inhoudsvenster om conversatie te verwijderen",
+      "Conversation analysis disabled": "Gespreksanalyse uitgeschakeld",
+      "Conversation analysis enabled": "Gespreksanalyse ingeschakeld",
+      "Copied": "Gekopieerd",
+      "Copy": "Kopiëren",
+      "Default": "Standaard",
+      "Delete": "Verwijderen",
+      "Delete a conversation": "Een gesprek verwijderen",
+      "Delete chat": "Chat verwijderen",
+      "Direct exchange with our team": "Directe uitwisseling met ons team",
+      "Enter your message or a question": "Voer uw bericht of een vraag in",
+      "Explore other LaSuite apps": "Ontdek andere LaSuite-apps",
+      "Failed to activate account. Please try again.": "Account activeren mislukt. Probeer het opnieuw.",
+      "Failed to copy": "Kopiëren mislukt",
+      "Failed to register for notifications. Please try again.": "Registratie voor meldingen mislukt. Probeer het opnieuw.",
+      "Failed to send feedback": "Het is niet gelukt om feedback te verzenden",
+      "Failed to update settings": "Het is niet gelukt om de instellingen bij te werken",
+      "Failed to upload file": "Het uploaden van het bestand is mislukt",
+      "Failed to upload files. Please try again.": "Bestanden uploaden is mislukt. Probeer het opnieuw.",
+      "Feedback Négatif": "Negatieve feedback",
+      "Feedback positif": "Positieve feedback",
+      "File type not supported": "Bestandstype niet ondersteund",
+      "Find recent news about...": "Vind het laatste nieuws over...",
+      "Get notified about the Public Beta.": "Ontvang een melding over de openbare bèta.",
+      "Get notified for the public beta": "Ontvang een melding voor de openbare bètaversie",
+      "Give a quick opinion": "Geef snel een mening",
+      "Give feedback": "Geef feedback",
+      "History": "Geschiedenis",
+      "Home": "Thuis",
+      "If enabled, this allows us to analyse your exchanges to improve the Assistant. If disabled, all conversations remain confidential and are not used in any way. ": "Indien ingeschakeld, kunnen we uw gesprekken analyseren om de Assistent te verbeteren. Indien uitgeschakeld, blijven alle gesprekken vertrouwelijk en worden ze op geen enkele manier gebruikt. ",
+      "Illustration": "Illustratie",
+      "Image 401": "Afbeelding 401",
+      "Image 403": "Afbeelding 403",
+      "Invalid activation code. Please check and try again.": "Ongeldige activeringscode. Controleer en probeer het opnieuw.",
+      "It seems that the page you are looking for does not exist or cannot be displayed correctly.": "Het lijkt erop dat de pagina die u zoekt niet bestaat of niet correct kan worden weergegeven.",
+      "Know more": "Meer weten",
+      "Language": "Taal",
+      "Learn more about data usage.": "Meer informatie over datagebruik.",
+      "Load more": "Laad Meer",
+      "Log in to access this page.": "Meld u aan om deze pagina te zien.",
+      "Login": "Login",
+      "Logo": "Logo",
+      "Logout": "Uitloggen",
+      "New chat": "Nieuwe chat",
+      "New feedback": "Nieuwe feedback",
+      "No code? ": "Geen code? ",
+      "No conversation found": "Geen gesprek gevonden",
+      "Notify me": "Breng mij op de hoogte",
+      "Open": "Open",
+      "Open the left panel": "Open het linker venster",
+      "Open the menu": "Open het menu",
+      "Page Not Found - Error 404": "Pagina niet gevonden - Fout 404",
+      "Please enter an activation code": "Voer een activeringscode in",
+      "Proconnect Login": "Login",
+      "Quick search input": "Snelle zoekinvoer",
+      "Remove attachment": "Bijlage verwijderen",
+      "Research on the web": "Onderzoek op het internet",
+      "Search": "Zoek",
+      "Search for a chat": "Zoek naar een chat",
+      "Search results": "Zoekresultaten",
+      "Search...": "Zoek...",
+      "Select": "Selecteer",
+      "Select model": "Selecteer model",
+      "Send": "Verstuur",
+      "Settings": "Instellingen",
+      "Show": "Toon",
+      "Simple chat icon": "Eenvoudig chatpictogram",
+      "Something bad happens, please retry.": "Er is iets misgegaan. Probeer het opnieuw.",
+      "Sorry, an error occurred. Please try again.": "Sorry, er is een fout opgetreden. Probeer het opnieuw.",
+      "Start a new conversation.": "Begin een nieuw gesprek.",
+      "Start conversation": "Begin een gesprek",
+      "Stop": "Stop",
+      "Summarizing...": "Samenvatten...",
+      "The Assistant is a sovereign conversational AI designed for public servants. It helps you save time on daily tasks like rephrasing, summarising, translating, or searching information. Your data never leaves France and is stored on secure, state-compliant infrastructures. It is never used for commercial purposes.": "De Assistent is een soevereine conversationele AI, ontworpen voor ambtenaren. Het helpt je tijd te besparen bij dagelijkse taken zoals het herformuleren, samenvatten, vertalen of zoeken van informatie. Je gegevens verlaten het land nooit en worden opgeslagen op beveiligde, door de overheid goedgekeurde infrastructuren. Ze worden nooit gebruikt voor commerciële doeleinden.",
+      "The Assistant is in Beta": "De Assistent is in bèta",
+      "The conversation has been deleted.": "Het gesprek is verwijderd.",
+      "The summary feature is not supported yet.": "De samenvattingsfunctie wordt nog niet ondersteund.",
+      "Thinking...": "Denken...",
+      "To add a file to the conversation, drop it here.": "Als u een bestand aan het gesprek wilt toevoegen, sleept u het hierheen.",
+      "Turn this list into bullet points": "Zet deze lijst om in opsommingstekens",
+      "Unlock access": "Toegang ontgrendelen",
+      "Unlocking...": "Ontgrendelen...",
+      "Untitled conversation": "Ongetiteld gesprek",
+      "Upload Error": "Uploadfout",
+      "Uploading files...": "Bestanden uploaden...",
+      "We'll email you at {{email}} when the public beta opens.": "We sturen u een e-mail op {{email}} zodra de openbare bètaversie opengaat.",
+      "We'll email you when the public beta opens.": "We sturen u een e-mail zodra de openbare bètaversie beschikbaar is.",
+      "Web": "Internet",
+      "What is on your mind?": "Waar denk je aan?",
+      "Write a short product description": "Schrijf een korte productbeschrijving",
+      "Write on Tchap": "Schrijf op Tchap",
+      "You are on the list": "Je staat op de lijst",
+      "You do not have permission to view this page.": "U heeft geen toestemming om deze pagina te bekijken.",
+      "You said: ": "Je zei: ",
+      "You will be notified!": "U wordt op de hoogte gebracht!",
+      "Your account is already activated.": "Uw account is al geactiveerd.",
+      "Your sovereign AI assistant": "Uw soevereine AI-assistent",
+      "source": "bron",
+      "sources": "bronnen",
+      "{{productName}} Logo": "{{productName}} Logo"
+    }
+  },
   "ru": {
     "translation": {
       "30 sec to tell us what you think or report a bug": "В течение 30 секунд расскажите нам, о чём вы думаете или сообщите об ошибке",
@@ -133,12 +271,14 @@
       "Access Denied - Error 403": "Отказано в доступе - Ошибка 403",
       "Access is limited to people who have an invitation code. If you have one, please enter it below.": "Доступ ограничен людьми с кодом приглашения. Если он у вас есть, введите его ниже.",
       "Account activated successfully!": "Учётная запись успешно активирована!",
+      "Actions list for conversation {{title}}": "Список действий для беседы {{title}}",
       "Add attach file": "Добавить вложение",
       "Add file": "Добавить файл",
       "Allow conversation analysis": "Разрешить анализ бесед",
       "An error occurred. Please try again.": "Произошла ошибка. Пожалуйста, повторите попытку.",
       "Are you sure you want to delete this conversation ?": "Вы действительно хотите удалить эту беседу?",
       "Ask a question": "Задать вопрос",
+      "Assistant IA replied: ": "Помощник ИИ ответил: ",
       "Assistant is already available, log in to use it now.": "Помощник уже доступен, просто войдите в систему.",
       "Assistant is in development: your feedback matters! Choose how to share your ideas:": "Помощник находится в разработке: ваши отзывы важны! Выберите, как поделиться своими идеями:",
       "Assistant settings": "Настройки помощника",
@@ -148,10 +288,11 @@
       "Cancel": "Отмена",
       "Clear search": "Очистить поиск",
       "Close model selector": "Закрыть выбор модели",
+      "Close the left panel": "Закрыть левую панель",
+      "Close the menu": "Закрыть меню",
       "Close the modal": "Закрыть это окно",
       "Confirm deletion": "Подтвердите удаление",
       "Content modal to delete conversation": "Подтверждение удаления беседы",
-      "Conversation actions": "Действия в беседе",
       "Conversation analysis disabled": "Анализ бесед отключён",
       "Conversation analysis enabled": "Анализ бесед включён",
       "Copied": "Скопировано",
@@ -161,6 +302,7 @@
       "Delete a conversation": "Удалить беседу",
       "Delete chat": "Удалить беседу",
       "Direct exchange with our team": "Прямое общение с нашей командой",
+      "Enter your message or a question": "Введите сообщение или вопрос",
       "Explore other LaSuite apps": "Посмотреть другие приложения LaSuite",
       "Failed to activate account. Please try again.": "Не удалось активировать учётную запись. Пожалуйста, попробуйте снова.",
       "Failed to copy": "Не удалось скопировать",
@@ -171,6 +313,7 @@
       "Failed to upload files. Please try again.": "Не удалось выгрузить файлы. Повторите попытку.",
       "Feedback Négatif": "Отрицательный отзыв",
       "Feedback positif": "Положительный отзыв",
+      "File type not supported": "Тип файла не поддерживается",
       "Find recent news about...": "Найти последние новости...",
       "Get notified about the Public Beta.": "Получать уведомления о публичной бета-версии.",
       "Get notified for the public beta": "Получать уведомления о публичной бета-версии",
@@ -198,7 +341,8 @@
       "No conversation found": "Беседы не найдены",
       "Notify me": "Уведомите меня",
       "Open": "Открыть",
-      "Open the header menu": "Открыть меню заголовка",
+      "Open the left panel": "Открыть левую панель",
+      "Open the menu": "Открыть меню",
       "Page Not Found - Error 404": "Страница не найдена - Ошибка 404",
       "Please enter an activation code": "Пожалуйста, введите код активации",
       "Proconnect Login": "Войти через Proconnect",
@@ -220,6 +364,7 @@
       "Start a new conversation.": "Начать новую беседу.",
       "Start conversation": "Начать беседу",
       "Stop": "Остановить",
+      "Summarizing...": "Обобщение...",
       "The Assistant is a sovereign conversational AI designed for public servants. It helps you save time on daily tasks like rephrasing, summarising, translating, or searching information. Your data never leaves France and is stored on secure, state-compliant infrastructures. It is never used for commercial purposes.": "Помощник - собеседник на основе ИИ для государственных служащих. Он поможет вам сэкономить время на ежедневных задачах, таких как перефразирование, обобщение, перевод или поиск информации. Ваши данные никогда не покидают Францию и хранятся в охраняемой государственной инфраструктуре, которая никогда не используется в коммерческих целях.",
       "The Assistant is in Beta": "Помощник находится на этапе Бета-версии",
       "The conversation has been deleted.": "Беседа была удалена.",
@@ -240,6 +385,7 @@
       "Write on Tchap": "Написать в Tchap",
       "You are on the list": "Вы в списке",
       "You do not have permission to view this page.": "У вас недостаточно прав для просмотра этой страницы.",
+      "You said: ": "Вы сказали: ",
       "You will be notified!": "Вы получите уведомление!",
       "Your account is already activated.": "Ваша учётная запись уже активирована.",
       "Your sovereign AI assistant": "Ваш надёжный ИИ-помощник",
@@ -256,12 +402,14 @@
       "Access Denied - Error 403": "Доступ заборонений - Помилка 403",
       "Access is limited to people who have an invitation code. If you have one, please enter it below.": "Доступ обмежено учасниками з кодом запрошення. Якщо у вас є один такий, будь ласка, введіть його нижче.",
       "Account activated successfully!": "Обліковий запис успішно активовано!",
+      "Actions list for conversation {{title}}": "Список дій для розмови {{title}}",
       "Add attach file": "Додати файл вкладення",
       "Add file": "Додати файл",
       "Allow conversation analysis": "Дозволити аналіз розмови",
       "An error occurred. Please try again.": "Сталась помилка. Спробуйте ще раз.",
       "Are you sure you want to delete this conversation ?": "Ви дійсно бажаєте видалити цю розмову?",
       "Ask a question": "Задати питання",
+      "Assistant IA replied: ": "Відповідь помічника ШІ: ",
       "Assistant is already available, log in to use it now.": "Помічник вже доступний, увійдіть щоб почати використання.",
       "Assistant is in development: your feedback matters! Choose how to share your ideas:": "Помічник в розробці: ваші відгуки мають значення! Оберіть, як поділитися своїми ідеями:",
       "Assistant settings": "Налаштування помічника",
@@ -271,10 +419,11 @@
       "Cancel": "Скасувати",
       "Clear search": "Очистити вікно пошуку",
       "Close model selector": "Закрити вікно вибору моделі",
+      "Close the left panel": "Закрити ліву панель",
+      "Close the menu": "Закрити меню",
       "Close the modal": "Закрити це вікно",
       "Confirm deletion": "Підтвердження видалення",
       "Content modal to delete conversation": "Підтвердження видалення розмови",
-      "Conversation actions": "Дії з розмовою",
       "Conversation analysis disabled": "Аналіз розмови вимкнено",
       "Conversation analysis enabled": "Аналіз розмов увімкнено",
       "Copied": "Скопійовано",
@@ -284,6 +433,7 @@
       "Delete a conversation": "Видалити розмову",
       "Delete chat": "Видалити розмову",
       "Direct exchange with our team": "Пряме спілкування з нашою командою",
+      "Enter your message or a question": "Введіть ваше повідомлення або питання",
       "Explore other LaSuite apps": "Ознайомтесь з іншими застосунками LaSuite",
       "Failed to activate account. Please try again.": "Не вдалося активувати обліковий запис. Спробуйте ще раз.",
       "Failed to copy": "Не вдалось скопіювати",
@@ -294,6 +444,7 @@
       "Failed to upload files. Please try again.": "Не вдалося вивантажити файли. Будь ласка, спробуйте ще раз.",
       "Feedback Négatif": "Негативний відгук",
       "Feedback positif": "Позитивний відгук",
+      "File type not supported": "Тип файлу не підтримується",
       "Find recent news about...": "Знайти останні новини про...",
       "Get notified about the Public Beta.": "Отримувати повідомлення про публічну бета-версію.",
       "Get notified for the public beta": "Отримувати повідомлення про публічну бета-версію",
@@ -321,7 +472,8 @@
       "No conversation found": "Розмови не знайдено",
       "Notify me": "Нагадати мені",
       "Open": "Відкрити",
-      "Open the header menu": "Відкрити меню заголовка",
+      "Open the left panel": "Відкрити ліву панель",
+      "Open the menu": "Відкрити меню",
       "Page Not Found - Error 404": "Сторінку не знайдено - Помилка 404",
       "Please enter an activation code": "Будь ласка, введіть код активації",
       "Proconnect Login": "Увійти через Proconnect",
@@ -343,6 +495,7 @@
       "Start a new conversation.": "Розпочати нову розмову.",
       "Start conversation": "Почати розмову",
       "Stop": "Зупинити",
+      "Summarizing...": "Узагальнення...",
       "The Assistant is a sovereign conversational AI designed for public servants. It helps you save time on daily tasks like rephrasing, summarising, translating, or searching information. Your data never leaves France and is stored on secure, state-compliant infrastructures. It is never used for commercial purposes.": "Помічник - це розмовний ШІ, призначений для державних службовців. Він допоможе вам зберегти час в таких щоденних завданнях, як рефразування, узагальнення, переклад або пошукова інформація. Ваші дані ніколи не покидають Францію та зберігаються на захищеній державній інфраструктурі. Вони ніколи не використовуються для комерційних цілей.",
       "The Assistant is in Beta": "Помічник у бета-версії",
       "The conversation has been deleted.": "Розмова була видалена.",
@@ -363,6 +516,7 @@
       "Write on Tchap": "Написати у Tchap",
       "You are on the list": "Ви є в списку",
       "You do not have permission to view this page.": "У вас немає прав для перегляду цієї сторінки.",
+      "You said: ": "Ви сказали: ",
       "You will be notified!": "Ви отримаєте повідомлення!",
       "Your account is already activated.": "Ваш обліковий запис вже активовано.",
       "Your sovereign AI assistant": "Ваш надійний помічник з ШІ",

src/frontend/apps/conversations/src/pages/globals.css

@@ -10,6 +10,13 @@ body {
   box-sizing: border-box;
 }
 
+button:focus-visible,
+a:focus-visible,
+[role='button']:focus-visible {
+  outline: 2px solid #3e5de7;
+  outline-offset: 2px;
+}
+
 main ::-webkit-scrollbar,
 .ReactModalPortal ::-webkit-scrollbar {
   width: 20px;
@@ -153,14 +160,20 @@ ul a:hover {
   overflow-wrap: break-word;
 }
 
+.mainContent-chat p {
+  margin: 0;
+}
+
 .mainContent-chat a {
-  color: var(--c--theme--colors--primary-text) !important;
+  color: inherit !important;
+  text-decoration: none !important;
+  transition: box-shadow 0.2s !important;
+  box-shadow: inset 0 -2px 0 0 #dae2ff !important;
 }
 
 .mainContent-chat a:hover {
-  color: var(
-    --c--components--button--primary--background--color-hover
-  ) !important;
+  color: inherit !important;
+  box-shadow: inset 0 -22px 0 0 #dae2ff !important;
 }
 
 .lasuite-gaufre-btn.lasuite--gaufre-opened {
@@ -182,7 +195,7 @@ ul a:hover {
   border-color: #b7d7ff !important;
 }
 
-.research-web-button:focus {
+.research-web-button:focus:not(:focus-visible) {
   box-shadow: none !important;
 }
 
@@ -235,3 +248,15 @@ figure[data-rehype-pretty-code-figure] > pre {
 figure[data-rehype-pretty-code-figure] > pre > code {
   display: block;
 }
+
+.sr-only {
+  position: absolute;
+  width: 1px;
+  height: 1px;
+  padding: 0;
+  margin: -1px;
+  overflow: hidden;
+  clip: rect(0, 0, 0, 0);
+  white-space: nowrap;
+  border: 0;
+}

src/frontend/apps/e2e/__tests__/app-conversations/common.ts

@@ -16,7 +16,7 @@ export const CONFIG = {
     ['en-us', 'English'],
     ['fr-fr', 'Français'],
     // ['de-de', 'Deutsch'],
-    // ['nl-nl', 'Nederlands'],
+    ['nl-nl', 'Nederlands'],
     // ['es-es', 'Español'],
   ],
   LANGUAGE_CODE: 'en-us',

src/frontend/apps/e2e/__tests__/app-conversations/header.spec.ts

@@ -97,7 +97,7 @@ test.describe('Header mobile', () => {
 
     const header = page.locator('header').first();
 
-    await expect(header.getByLabel('Open the header menu')).toBeVisible();
+    await expect(header.getByLabel('Open the menu')).toBeVisible();
     await expect(
       header.getByRole('link', { name: 'Assistant Logo' }),
     ).toBeVisible();

src/frontend/apps/e2e/package.json

@@ -1,6 +1,6 @@
 {
   "name": "app-e2e",
-  "version": "0.0.2",
+  "version": "0.0.8",
   "private": true,
   "scripts": {
     "lint": "eslint . --ext .ts",

src/frontend/package.json

@@ -1,6 +1,6 @@
 {
   "name": "conversations",
-  "version": "0.0.2",
+  "version": "0.0.8",
   "private": true,
   "workspaces": {
     "packages": [

src/frontend/packages/eslint-config-conversations/package.json

@@ -1,6 +1,6 @@
 {
   "name": "eslint-config-conversations",
-  "version": "0.0.2",
+  "version": "0.0.8",
   "license": "MIT",
   "scripts": {
     "lint": "eslint --ext .js ."

src/frontend/packages/i18n/package.json

@@ -1,6 +1,6 @@
 {
   "name": "packages-i18n",
-  "version": "0.0.2",
+  "version": "0.0.8",
   "private": true,
   "scripts": {
     "extract-translation": "yarn extract-translation:conversations",

src/helm/env.d/dev-staging/values.conversations.yaml.gotmpl

@@ -16,7 +16,6 @@ backend:
     DJANGO_CSRF_TRUSTED_ORIGINS: https://conversations.127.0.0.1.nip.io
     DJANGO_CONFIGURATION: Feature
     DJANGO_ALLOWED_HOSTS: conversations.127.0.0.1.nip.io
-    DJANGO_SERVER_TO_SERVER_API_TOKENS: secret-api-key
     DJANGO_SECRET_KEY: *djangoSecretKey
     DJANGO_SETTINGS_MODULE: conversations.settings
     DJANGO_SUPERUSER_PASSWORD: admin

src/helm/env.d/dev/values.conversations.yaml.gotmpl

@@ -19,7 +19,6 @@ backend:
     DJANGO_CSRF_TRUSTED_ORIGINS: https://conversations.127.0.0.1.nip.io
     DJANGO_CONFIGURATION: Feature
     DJANGO_ALLOWED_HOSTS: conversations.127.0.0.1.nip.io
-    DJANGO_SERVER_TO_SERVER_API_TOKENS: secret-api-key
     DJANGO_SECRET_KEY: *djangoSecretKey
     DJANGO_SETTINGS_MODULE: conversations.settings
     DJANGO_SUPERUSER_PASSWORD: admin

src/mail/package.json

@@ -1,6 +1,6 @@
 {
   "name": "mail_mjml",
-  "version": "0.0.2",
+  "version": "0.0.8",
   "description": "An util to generate html and text django's templates from mjml templates",
   "type": "module",
   "dependencies": {