Interactive SQLite RAG Server with MCP Integration

A full-stack RAG (Retrieval-Augmented Generation) system that combines semantic search with LLM tool integration. Built with Node.js, React, SQLite vector search, and FastMCP for seamless AI agent integration.

• Document Management: CRUD operations for knowledge documents with automatic vector embeddings
• Semantic Search: Natural language search with metadata filtering and similarity scoring
• Vector Storage: SQLite with sqlite-vss extension for efficient vector similarity search
• MCP Integration: FastMCP server exposing rag_search tool for Claude, OpenAI, and other LLM frameworks
• Web Interface: Modern React UI for document management and search testing
• Health Monitoring: Real-time system health dashboard
• Multi-Database Support: Repository pattern enables PostgreSQL migration

• Node.js 18+ (LTS recommended)
• npm 9+ or yarn 1.22+
• OpenAI API Key (for embeddings)

1. Clone the repository

   git clone https://github.com/your-org/interactive-sqlite-rag-mcp.git
   cd interactive-sqlite-rag-mcp

2. Install backend dependencies

   cd backend
   npm install

3. Install frontend dependencies

   cd ../frontend
   npm install

1. Backend environment variables

   cd backend
   cp .env.example .env
   # Edit .env with your OpenAI API key

2. Frontend environment variables (optional, defaults should work for development)

   cd ../frontend
   cp .env.example .env

1. Start the backend server

   cd backend
   npm run dev

   Server will start on http://localhost:3000

2. Start the frontend (in another terminal)

   cd frontend
   npm run dev

   Frontend will start on http://localhost:5173

3. Access the application

   • Web interface: http://localhost:5173
   • API documentation: http://localhost:3000/api/docs (when implemented)
   • MCP server: http://localhost:3001 (for LLM integration)

• POST /api/documents - Create document with auto-embeddings
• GET /api/documents - List documents (paginated)
• GET /api/documents/:id - Get document by ID
• PUT /api/documents/:id - Update document
• DELETE /api/documents/:id - Delete document

• POST /api/search - Semantic search with filters

• GET /api/health - System health status

• GET /mcp/tools - List available MCP tools

The system exposes a rag_search tool for LLM frameworks:

{
  "name": "rag_search",
  "description": "Search the knowledge base using semantic similarity",
  "parameters": {
    "query": "string (required)",
    "filters": "object (optional)",
    "limit": "number (optional, default 10)"
  }
}

Add to your claude_desktop_config.json:

{
  "mcpServers": {
    "rag-server": {
      "command": "node",
      "args": ["/path/to/interactive-sqlite-rag-mcp/backend/src/mcp/server.js"],
      "env": {
        "OPENAI_API_KEY": "your-api-key"
      }
    }
  }
}

# Backend tests
cd backend
npm test
npm run test:coverage

# Frontend tests
cd frontend
npm test

# Backend
cd backend
npm run lint
npm run lint:fix

# Frontend
cd frontend
npm run lint
npm run lint:fix

# Run migrations
cd backend
npm run migrate

# Seed with sample data
npm run seed

backend/
├── src/
│   ├── models/           # Sequelize models
│   ├── repositories/     # Database abstraction layer
│   │   ├── adapters/     # SQLite/PostgreSQL adapters
│   ├── services/         # Business logic
│   ├── api/              # Express routes & middleware
│   └── mcp/              # FastMCP server
├── tests/                # Test suites
└── migrations/           # Database migrations

frontend/
├── src/
│   ├── components/       # React components
│   ├── pages/           # Page components
│   ├── services/        # API clients
│   └── lib/             # Utilities
└── tests/               # Component tests

• Search Latency: <200ms p95 for queries against <1K documents
• Document Creation: <500ms including embedding generation
• Memory Usage: <512MB for typical workloads
• Concurrent Users: Supports 10+ simultaneous search requests

1. Follow the project's Constitution
2. Write tests before implementation (TDD)
3. Ensure 80% test coverage minimum
4. Pass all linting checks
5. Update documentation for user-facing changes

MIT License - see LICENSE file for details.

• Issues: GitHub Issues
• Discussions: GitHub Discussions
• Documentation: See docs/ directory