Files

David Blanc Brioir ad2c29a777 Fix Library RAG MCP connection issue and add setup documentation

Problem:
- Library RAG MCP server was failing to connect with timeout error
- Backend status showed "connected: false" with "MCP error -32001: Request timed out"
- Documents uploaded via upload_document tool were never processed

Root Cause:
- MISTRAL_API_KEY was commented out in .env file
- MCP server requires this key for OCR and LLM processing
- Without the key, the Python subprocess fails to start
- This caused connection timeout in the Node.js backend

Solution:
- Uncommented MISTRAL_API_KEY in .env (not committed, in .gitignore)
- Added LIBRARY_RAG_SETUP.md with complete setup guide
- Updated .claude/settings.local.json with bash permissions

Changes:
- Added LIBRARY_RAG_SETUP.md (setup documentation)
- Updated .claude/settings.local.json (auto-approved bash commands)

Verified:
- MCP server now connects successfully
- Status endpoint shows "connected: true"
- All 7 Library RAG tools available (upload_document, search_library, etc.)

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>

2025-12-26 19:22:00 +01:00

2.9 KiB

Raw Blame History

Library RAG MCP Setup Guide

Quick Setup

To enable the Library RAG MCP server (document search with Weaviate), you need to configure the following in your .env file:

Required Environment Variables

# Library RAG MCP Configuration
LIBRARY_RAG_ENABLED=true
MCP_LIBRARY_RAG_SERVER_PATH=C:/GitHub/linear_coding_library_rag/generations/library_rag/mcp_server.py
LIBRARY_RAG_PYTHON_COMMAND=python
LIBRARY_RAG_CONNECTION_TIMEOUT=10000
LIBRARY_RAG_AUTO_RECONNECT=true
LIBRARY_RAG_MAX_RETRIES=3

# Weaviate Configuration
WEAVIATE_URL=http://localhost:8080

# REQUIRED: Mistral API Key (for OCR functionality)
MISTRAL_API_KEY=your_mistral_api_key_here

Why MISTRAL_API_KEY is Required

The Library RAG MCP server uses Mistral API for:

OCR with annotations: Extracting text from PDF images with layout annotations
LLM processing: Metadata extraction, table of contents generation, semantic chunking

Without this key, the MCP server will fail to start and the backend connection will timeout with error: MCP error -32001: Request timed out.

Getting Your Mistral API Key

Go to https://console.mistral.ai/
Sign up or log in
Navigate to API Keys section
Create a new API key
Copy the key and add it to your .env file

Troubleshooting

Issue: Library RAG shows "connected": false in status endpoint

Solution:

Check that MISTRAL_API_KEY is uncommented in .env
Verify the key is valid
Restart the backend server: npm run dev

Issue: Connection timeout error

Cause: The MCP server subprocess cannot start without the Mistral API key

Fix: Add the key to .env and restart

Verifying Connection

Check the connection status:

curl http://localhost:5175/api/library-rag/status | python -m json.tool

Expected response when connected:

{
  "success": true,
  "data": {
    "connected": true,
    "message": "Library RAG MCP server is connected and ready",
    "tools": [...],
    "error": null
  }
}

Available Tools

Once connected, you'll have access to:

upload_document - Upload and index PDF documents with full pipeline (OCR, metadata, chunking, Weaviate ingestion)
search_library - Semantic search over document chunks
hybrid_search - Search summaries and high-level content
list_collections - List all indexed documents
get_document - Retrieve document metadata
filter_search - Filter by author, work, or language
extract_text_from_image - OCR from image URLs

Make sure Weaviate Docker is running:

cd C:\GitHub\linear_coding_library_rag\generations\library_rag
docker-compose up -d

Check Weaviate status:

curl http://localhost:8080/v1/meta

Changelog

2025-12-26: Fixed MCP connection issue by adding MISTRAL_API_KEY requirement to documentation. The key must be present in .env for the MCP server to start successfully.

2.9 KiB Raw Blame History