Remote MCP Server logo

Remote MCP Server

by chrisjpatty

This project provides a remote Model Context Protocol (MCP) server that can be deployed on Cloudflare Workers with OAuth login. It allows you to connect MCP clients like the MCP Inspector and Claude Desktop to a remote server.

View on GitHub

Last updated: N/A

What is Remote MCP Server?

This is a remote MCP server designed to run on Cloudflare Workers, enabling remote access to MCP APIs. It includes OAuth login for security and allows integration with tools like the MCP Inspector and Claude Desktop.

How to use Remote MCP Server?

The server can be deployed to Cloudflare Workers using the provided instructions. After deployment, you can connect MCP clients by configuring them to use the server's URL (e.g., workers.dev/sse). The README provides detailed steps for local development, deployment, and connecting various clients.

Key features of Remote MCP Server

  • Remote MCP server

  • Cloudflare Workers deployment

  • OAuth login

  • SSE transport

  • Integration with MCP Inspector

  • Integration with Claude Desktop

Use cases of Remote MCP Server

  • Providing remote access to AI tools

  • Integrating AI models with applications

  • Developing and testing MCP-based applications

  • Enabling Claude to use external tools

  • Building serverless AI workflows

FAQ from Remote MCP Server

How do I connect the MCP Inspector?

Start the inspector, switch the Transport Type to SSE, and enter the server's URL (e.g., http://localhost:8787/sse or worker-name.account-name.workers.dev/sse).

How do I connect Claude Desktop?

Update the Claude configuration file to point to the server's URL in the mcpServers section.

What if something goes wrong?

Try restarting Claude or connecting directly to the MCP server on the command line with npx mcp-remote http://localhost:8787/sse.

What is MCP?

MCP stands for Model Context Protocol. It is a protocol for connecting AI models with external tools and data sources.

How do I deploy to Cloudflare?

Create a KV namespace, add the namespace ID to wrangler.jsonc, and run npm run deploy.