GitHub MCP Server
Official GitHub server that lets AI assistants read repos, manage issues and PRs, and check Actions runs.
- Stars
- 34.2k
- Forks
- 5.3k
- Licence
- MIT
- Language
- Go
- Last update
- 4 days ago
- Type
- Dev tool
What it is
The GitHub MCP Server connects AI tools such as assistants, chatbots, and agents directly to GitHub. Using MCP (Model Context Protocol), it lets those tools read repositories and code, manage issues and pull requests, analyze code, and work with workflows through natural-language requests. GitHub hosts a remote version and also publishes a local version that runs on your machine.
- Ask an AI assistant to summarize open issues or pull requests in a repository
- Have AI help triage bugs and review code changes
- Monitor GitHub Actions workflow runs and analyze build failures
- Review Dependabot alerts and security findings through natural language
Great for
- Developers who already use an MCP-compatible AI tool such as VS Code with Copilot, Claude Desktop, or Cursor
- Teams who want AI help triaging issues and reviewing pull requests
- People who want AI to look into failing GitHub Actions runs
- Developers building multi-step agent workflows that need GitHub context
Not ideal for
- People without an MCP-capable AI host, since the server has nothing to connect to without one
- GitHub Enterprise Server users, because the README says remote hosting is not supported there and the local setup is needed instead
- Anyone who wants a standalone GitHub app with its own interface, since this is a connector for other tools
How to get it
- 1
See the README
You'll need: A compatible MCP host with remote server support (VS Code 1.101+, Claude Desktop, Cursor, Windsurf, etc.)
Clone the repository
git clone https://github.com/github/github-mcp-server - 2
Run it
- Pick your AI host and open its install guide from the README list
- For VS Code, click the one-click install button or paste the remote JSON config into your host configuration
- Enable Agent mode in VS Code Copilot Chat so the server starts
- Authenticate with OAuth or a GitHub personal access token, then ask a GitHub question
- 3
Get the most out of it
- Use toolsets to limit which tools are exposed; the README says default toolsets are used when none are specified
- Try the insiders endpoint for early-access tools, via URL path or the X-MCP-Insiders header
- Read docs/remote-server.md for headers, toolsets, and advanced configuration
- Check docs/policies-and-governance.md if your organization restricts access
Downloads
Strengths
- Official project maintained by GitHub, with a hosted remote option that needs no local install
- Supports many AI hosts, with setup guides listed in the README
- Open source under the MIT license
Trade-offs
- Access depends on your GitHub permissions and the policies your organization enables, as the README notes
- Setup differs by AI host, and some hosts need a GitHub App or OAuth App for remote OAuth access
- The README is truncated here, so details of the local install and toolset options were not fully verified
How it compares
| Project | Stars | Type | Level | Updated | Best for |
|---|---|---|---|---|---|
| github-mcp-server (this) | 34.2k | Dev tool | intermediate | 4 days ago | Developers who already use an MCP-compatible AI tool such as VS Code with Copilot, Claude Desktop, or Cursor |
| Skill_Seekers | 15.3k | Dev tool | intermediate | 1 months ago | Developers who want an AI assistant to know a specific library's documentation |
| codebase-memory-mcp | 48k | Dev tool | intermediate | 4 days ago | Developers who use an MCP-capable AI coding agent on a large codebase |
| Trail-Code | 11k | Dev tool | beginner | 4 days ago | Developers who use Claude Code, Cursor, Codex or similar agents on a medium or large repository |
| context7 | 64.1k | AI model / agent | beginner | 4 days ago | Developers who use AI coding agents such as Claude, Cursor, or VS Code for libraries that change often |
| python-sdk | 24.7k | Library | intermediate | 4 days ago | Python developers who want to expose their own functions to an LLM app as MCP tools |
Stats synced from GitHub. The assessment above was drafted with AI from the project's README on 2026-10-29 and may miss recent changes — check the README before relying on a command. MIT licence: free to use, modify and distribute, including commercially, as long as you keep the copyright notice.