Turf

Turf-MCP provides geographic analysis capabilities exposed as an MCP (Model Context Protocol) server. It enables you to perform spatial measurements, geometry operations, coordinate transformations, and more by connecting with an MCP client over standard MCP channels. This makes it easy to integrate Turf.js-based spatial tooling into your workflows and data processing pipelines.

How to use

You connect to Turf-MCP using an MCP client and choose either a local stdio server or a remote HTTP/SSE endpoint. Start a local stdio server to run Turf-MCP directly in your environment or use the remote HTTP/SSE endpoint if you prefer a hosted or network-accessible service. Once connected, you can perform a wide range of spatial operations including measuring distances, calculating areas, generating grids, transforming coordinates, and more.

How to install

Prerequisites you need before running Turf-MCP:

• Python 3.10+ to support the Turf-MCP library if you are integrating Python tools with MCP workflows.

• Node.js 14.x to provide the required runtime for the MCP server runner used in examples.

Steps to set up and run Turf-MCP locally using the stdio (local) mode are shown in the code examples. Follow the exact commands shown to ensure compatibility with the provided tooling.

Configuration for local stdio server

{
  "mcpServers": {
    "turf_mcp": {
      "command": "cmd",
      "args": ["/c", "uvx", "turf-mcp"],
      "type": "stdio"
    }
  }
}

Configuration for remote HTTP/SSE server

{
  "mcpServers": {
     "turf_mcp_server": {
      "url": "http://127.0.0.1:8000/sse"
    }
  }
}

How to start the remote server

Start the remote server by running the runtime command and selecting the SSE or HTTP mode. For example, you can start an SSE-based server that serves the MCP endpoint on port 8080, then connect your MCP client to the path provided (http://localhost:8080/sse for SSE or http://localhost:8080/mcp for HTTP).