MCP Server

A Model Context Protocol (MCP) server implementation using Node.js and Express.

What is MCP?

Model Context Protocol (MCP) is a standard that connects AI systems with external tools and data sources. This server provides a standardized way to expose functions/tools that can be called by AI systems.

Features

• Express-based HTTP API
• Built-in tool registration system
• Sample tool implementations
• Easy extensibility
• CORS support

Installation

# Clone the repository (if using git)
git clone [your-repo-url]
cd mcp-server

# Install dependencies
npm install

Configuration

The server uses environment variables for configuration:

• PORT: The port on which the server will run (default: 3000)
• NODE_ENV: The environment mode (development, production)

You can set these in the .env file.

Usage

Starting the server

# Start the server
npm start

# Start with nodemon (automatic restarts during development)
npm run dev

API Endpoints

• GET /health: Health check endpoint
• GET /functions: List all available tools/functions
• POST /execute: Execute a specific function with parameters

Example usage with curl:

# Get list of available functions
curl http://localhost:3000/functions

# Execute a function
curl -X POST http://localhost:3000/execute \
  -H "Content-Type: application/json" \
  -d '{"function_name": "echo", "parameters": {"message": "Hello, World!"}}'

Adding Custom Tools

You can add custom tools in the src/tools.js file following this format:

toolName: {
  description: "Description of what the tool does",
  execute: async (params) => {
    // Implementation
    return {
      // Result
    };
  }
}

Deploying to Windsurf

To deploy this MCP server to Windsurf:

1. Make sure you have the Windsurf CLI installed
2. Configure your Windsurf credentials
3. Run npm run deploy or follow the Windsurf deployment documentation

License

MIT