lightrag-mcp-server

netbrah/lightrag-mcp-server

3.1

If you are the rightful owner of lightrag-mcp-server and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to dayong@mcphub.com.

LightRAG MCP Server is a graph-based code search server that integrates with VS Code Copilot to enhance code retrieval and visualization.

LightRAG MCP Server

Graph-based code search MCP server that integrates LightRAG with VS Code Copilot.

Features

  • 🔍 Graph-based retrieval: Not just vector similarity, but entity relationships
  • 🌐 Multi-mode querying: Local (focused), Global (architectural), Hybrid (both)
  • 📊 Visual architecture: Generate Mermaid diagrams from code
  • ⚡ Incremental indexing: Real-time updates as code changes
  • 🗄️ Flexible storage: Development (NetworkX) and production (Neo4J + Milvus)
  • 🎯 IDE-optimized: Response formatting, keyword prioritization, and token control
  • 📋 Direct text insertion: Index clipboard/selection content without files

Quick Start

# Install dependencies
npm install
pip install -r python/requirements.txt

# Configure environment
cp .env.example .env
# Edit .env with your API keys

# Build
npm run build

# Run
node dist/index.js

Offline Installation

For air-gapped or offline environments, create a standalone installation bundle:

# Create offline bundle
npm run bundle:offline

# This creates: lightrag-mcp-server-{version}-offline-bundle.tar.gz
# Transfer this file to your offline environment and follow instructions in:
# - AIRGAP_DEPLOYMENT.md - Complete deployment guide
# - README-OFFLINE.md - Quick installation steps

VS Code Integration

Create .vscode/mcp.json:

{
  "servers": {
    "lightrag": {
      "command": "node",
      "args": ["path/to/lightrag-mcp-server/dist/index.js"],
      "env": {
        "LIGHTRAG_WORKING_DIR": "${workspaceFolder}/.lightrag",
        "OPENAI_API_KEY": "your-key"
      }
    }
  }
}

Documentation

See for complete documentation:

  • - NEW: IDE-specific features

License

MIT