MCP Server for Cursor

A Model Context Protocol (MCP) server implementation for Cursor IDE integration, providing a modern web dashboard and tools through SSE (Server-Sent Events) and WebSocket connections.

Features

Real-time connection monitoring with WebSocket support
Modern web dashboard for server management
SSE-based communication with Cursor IDE
Built-in test tool for verifying connectivity
Google Drive integration for file management
Extensible architecture for adding new tools
Connection history visualization
Client management interface
Service configuration UI
Customizable server settings

Prerequisites

Python 3.8 or higher
Node.js 18 or higher
npm (Node.js package manager)
pip (Python package manager)
Cursor IDE

Installation

Clone the repository:

git clone <repository-url>
cd MCP-Server

Install dependencies:

# Install all dependencies (both backend and frontend)
npm run install-all

# Or install separately:
# Backend dependencies
pip install -r requirements.txt

# Frontend dependencies
cd frontend
npm install
cd ..

Development

Start the development servers:

# Start both backend and frontend in development mode
npm run dev   # This will start both servers concurrently

The servers will start at:

Backend: http://localhost:8765
Frontend: http://localhost:3000 (or 3001 if port 3000 is in use)

You can also start the servers separately:

# Backend (in one terminal)
python mcp_server.py

# Frontend (in another terminal)
cd frontend
npm run dev

The development servers provide:

Hot reloading for frontend changes
Automatic proxy of API requests to the backend
WebSocket connection handling
Concurrent backend and frontend development

If you're running other applications that use port 3000, the frontend will automatically try port 3001 and increment until it finds an available port. The actual URL will be displayed in the terminal when you run npm run dev.

Project Structure

MCP-Server/
├── frontend/                # Frontend application
│   ├── src/                # Source code
│   │   ├── api/           # API clients
│   │   ├── pages/         # React components
│   │   └── main.tsx       # Entry point
│   ├── package.json       # Frontend dependencies
│   └── vite.config.ts     # Vite configuration
├── mcp_server.py          # Main server implementation
├── services.py            # Service management
├── test_client.py         # Test client implementation
├── requirements.txt       # Python dependencies
├── package.json           # Root package.json
└── services_config.json   # Service configurations

Available Endpoints

HTTP Endpoints

/api/status - Get server status
/api/connections/history - Get connection history
/api/clients - Get active clients
/api/services/{service_id} - Service configuration
/api/settings - Server settings

WebSocket Endpoint

/ws/{client_id} - Real-time updates and ping/pong

SSE Endpoint

/sse - Server-Sent Events for Cursor IDE

Tool Endpoints

/invoke/test - Test tool
/invoke/google_drive - Google Drive operations

Production Deployment

Build and start the production server:

# Build frontend and start production server
npm run prod

# Or build frontend separately:
npm run build
python mcp_server.py

Configuration

Initial Setup

Copy the template configuration:

cp services_config.template.json services_config.json

Update services_config.json with your service credentials:
- For Google Drive:
  - Create a project in Google Cloud Console
  - Enable the Google Drive API
  - Create OAuth 2.0 credentials
  - Add your client_id and client_secret

Server Settings

Configure server settings through the web dashboard:

Debug mode
SSL settings
Connection limits
Ping timeout
Protocol settings

Testing

Test the server using the provided test client:

python test_client.py

The test client will:

Connect to both SSE and WebSocket endpoints
Send periodic pings
Test available tools
Monitor connection status

Troubleshooting

Connection Issues
- Check both backend and frontend logs
- Verify the WebSocket connection in browser DevTools
- Ensure the proxy settings in vite.config.ts are correct
- Check for port conflicts
Frontend Issues
- Clear browser cache
- Check browser console for errors
- Verify Node.js and npm versions
- Check for TypeScript compilation errors
Backend Issues
- Check mcp_server.log for errors
- Verify Python dependencies are installed
- Check port availability
- Ensure service configurations are valid

Security Notes

The server accepts all origins (CORS *) in development
WebSocket connections use client IDs for basic identification
Service credentials are stored in services_config.json
Environment variables should be used for sensitive data in production

Contributing

Fork the repository
Create a feature branch
Make your changes
Run tests
Submit a pull request

License

[Add your license information here]