GitHub MCP Server
A self-hosted MCP server that gives Claude access to your GitHub account — read files, browse repos, commit changes, and manage issues and pull requests, all from a conversation.
README
GitHub MCP Server
A self-hosted MCP server that gives Claude access to your GitHub account — read files, browse repos, commit changes, and manage issues and pull requests, all from a conversation.
You deploy it to your own Railway account under your own GitHub token, so your code and credentials stay yours.
Setup
→ SETUP.md — full walkthrough, about 15 minutes.
Short version:
- Create a fine-grained GitHub personal access token
- Fork this repo (already forked it? sync your fork first)
- Generate an auth token — a new random string you invent, not your GitHub token (
openssl rand -hex 32) - Deploy to Railway with both tokens as variables
- Add
https://your-app.up.railway.app/mcp?key=your-auth-tokenas a custom connector in Claude - Verify —
/mcpwithout the key should return403 Forbidden, with the key should returnok
What Claude can do once it's connected
- Read any file, directory, or full file tree in your repos
- Create and update files with real commits
- List and search repos, commits, and branches
- Create, read, update, and comment on issues
- List, read, and merge pull requests
- Search code, repos, and issues across your account
Claude shows these as 15 read-only tools and 5 write/delete tools, and lets you require approval per tool. Worth gating the write ones.
Security
The /mcp endpoint requires a shared secret passed as ?key=. Without it, requests are rejected with 403 Forbidden. The server will not start at all unless MCP_AUTH_TOKEN is set, so it can never run unprotected.
Generate your own auth token. Do not reuse one from another deployment. Your Railway URL is public, and this server holds a GitHub token with write access to your repos — the auth token is the only thing standing between that and anyone who happens to have your URL.
/health is intentionally open so Railway's healthcheck can reach it. It exposes no data and runs no tools.
Requirements
- A GitHub account
- A Railway account (the free tier is enough to start)
- A Claude plan that supports custom connectors
Stack
TypeScript, Express, and the official MCP SDK over streamable HTTP. Deployed via Dockerfile; Railway config is in railway.toml.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.