git-mcp
A Git MCP server that doesn't suck
Return to top
git-mcp is a production-grade Model Context Protocol server that gives AI agents full access to Git workflows. It uses simple-git under the hood and requires only a system Git installation — no other dependencies.
It is especially useful for coding agents running inside GitHub Copilot, Claude Code, Cursor, and OpenAI Codex, where reliable Git inspection and safe mutation matter more than vibes and shell roulette.
Prerequisites
- Node.js 20 or later
- Git installed and available on
PATH - An MCP-compatible AI agent or coding agent host (Claude Desktop, GitHub Copilot in VS Code, Cursor, OpenAI Codex, etc.)
Installation
npx (no install required)
Run git-mcp on demand without installing anything globally:
bash
npx @selfagency/git-mcp
Global install
bash
npm install -g @selfagency/git-mcp
git-mcp
From source
bash
git clone https://github.com/selfagency/git-mcp.git
cd git-mcp
pnpm install
pnpm build
pnpm start
Connecting to Claude Desktop
Edit your Claude Desktop configuration file:
- macOS:
~/Library/Application Support/Claude/claude_desktop_config.json - Windows:
%APPDATA%\Claude\claude_desktop_config.json
json
{
"mcpServers": {
"git": {
"command": "npx",
"args": ["-y", "@selfagency/git-mcp"],
"env": {
"GIT_REPO_PATH": "/path/to/your/repo"
}
}
}
}
Restart Claude Desktop after saving the file. You should see a hammer icon in the chat toolbar indicating MCP tools are available.
Connecting to VS Code (GitHub Copilot)
User settings
Add to your VS Code settings.json (Cmd+Shift+P → "Open User Settings (JSON)"):
json
{
"mcp": {
"servers": {
"git": {
"type": "stdio",
"command": "npx",
"args": ["-y", "@selfagency/git-mcp"],
"env": {
"GIT_REPO_PATH": "${workspaceFolder}"
}
}
}
}
}
Workspace .vscode/mcp.json
For per-project configuration, create .vscode/mcp.json in the root of your repository:
json
{
"servers": {
"git": {
"type": "stdio",
"command": "npx",
"args": ["-y", "@selfagency/git-mcp"],
"env": {
"GIT_REPO_PATH": "${workspaceFolder}"
}
}
}
}
Setting a Default Repository
The GIT_REPO_PATH environment variable sets the default repository path used by all tools when repo_path is not provided in the tool call:
bash
GIT_REPO_PATH=/home/user/myproject git-mcp
You can also pass it on the command line:
bash
npx @selfagency/git-mcp --repo-path /home/user/myproject
If neither is set, the AI must provide repo_path explicitly in every tool call.
Verifying the Installation
Ask your AI agent:
"What's the git status of this repo?"
"Show me the last 5 commits."
The AI agent will call git_status or git_log and stream results back. If you see an error about repo_path, set GIT_REPO_PATH in the server configuration.
Next Steps
- Configuration reference — all environment variables
- MCP Resources — URI-addressable read-only data
- Tool reference — complete parameter documentation for every tool
- Safety model — how git-mcp protects against destructive operations
Похожие серверы
Alpha Vantage MCP Server
спонсорAccess financial market data: realtime & historical stock, ETF, options, forex, crypto, commodities, fundamentals, technical indicators, & more
widemem.ai
Open-source AI memory layer with importance scoring, temporal decay, hierarchical memory, and YMYL prioritization
MCP Test Utils
Desktop UI automation for AI agents: screenshots, window management, mouse, keyboard, UI Automation tree, OCR
FastMCP
A fast, Pythonic framework for building MCP servers and clients.
Storybook MCP
A universal MCP server that connects to any Storybook site and extracts documentation in real-time using Playwright. Use it with any AI or client that supports MCP (Model Context Protocol)—Cursor, Claude Desktop, Windsurf, or other MCP hosts.
DeepSeek MCP Server
An MCP server for the DeepSeek API, providing code review, file management, and account management.
Bitrise
Manage apps, builds, and artifacts on Bitrise, a Continuous Integration and Delivery (CI/CD) platform.
MCP SSE Launcher
A Python management system for MCP servers that wraps stdio-based servers as SSE endpoints and includes a web inspector for testing.
kemdiCode MCP
kemdiCode MCP is a Model Context Protocol server that gives AI agents and IDE assistants access to 124 specialized tools for code analysis, generation, git operations, file management, AST-aware editing, project memory, cognition & self-improvement, multi-board kanban, and multi-agent coordination.
DeepSeek MCP Server
A server for code generation and completion using the DeepSeek API.
Local Context MCP
A collection of reference implementations for the Model Context Protocol (MCP), giving LLMs secure access to tools and data.