Git MCP Rust

Git MCP Rust is a high-performance Model Context Protocol (MCP) server that empowers AI Agents (Claude, Gemini, Cursor) to explore GitHub repositories in real-time without local cloning.

The Problem It Solves

AI models often have a "knowledge cutoff" and hallucinate library versions. This tool provides Real-Time Context:

Outdated Dependencies: AI can now fetch the absolute latest SemVer-sorted tags.
Context Gap: AI can "read" remote source code and file structures to understand libraries it wasn't trained on.

Features

Zero-Clone Exploration: Fetch trees, files, and metadata via GitHub API.
Smart Dependency Solving: Fetches and sorts tags by Semantic Versioning (SemVer), allowing AI to always use the latest library versions.
Secure & Rate-Limit Ready: Natively supports GITHUB_TOKEN to increase API rate limits from 60 to 5,000 requests/hour.
Multi-Arch Support: Native binaries for x86_64, aarch64 (ARM64), and armv7.

Tools Available for AI

| Tool | Description | |------|-------------| | get_tags | Returns latest tags/versions (supports limit & SemVer sorting). | | get_file_tree | Recursively lists files to reveal project architecture. | | get_file_content | Reads raw content of specific files from any branch/tag. | | get_readme | Fetches the default README for quick project overview. | | get_changelog | Compares two tags and returns a commit summary. |

Example Workflows

Here is how Git Remote MCP bridges the gap between an AI's outdated knowledge and the current state of a repository.

1. Solving the "Outdated Dependency" Problem

Scenario: A user wants to use the clap library in Rust, but the AI only knows about version v3.0 (which has a different API than the current v4.5).

User Prompt: "Create a Rust CLI using clap to parse an --input argument."
AI Internal Workflow:
1. Identify: The AI realizes it needs the latest clap version.
2. Verify: Calls get_tags(url="https://github.com/clap-rs/clap", limit=5).
3. Receive: Gets ["v4.5.59", "v4.5.58", ...].
4. Action: The AI reads the latest docs or simply uses the confirmed v4.5.59 version.
Result: The AI provides code using the modern derive API instead of the deprecated v3 builder pattern.

2. Remote Code Analysis (Without Cloning)

Scenario: You want to understand a specific logic in a massive repository without downloading it.

User Prompt: "How does the Zygisk-Loader project handle library injection?"
AI Internal Workflow:
1. Explore: Calls get_file_tree(url="...") to see the source structure.
2. Search: Calls search_repository(query="dlopen", url="...") to find the injection point.
3. Read: Calls get_file_content(path="src/main.cpp", url="...") to read the actual logic.
Result: The AI explains the exact C++ logic and function calls used for injection.

3. Reviewing Version Changes

Scenario: You need to know what changed between two versions of a library to decide if you should upgrade.

User Prompt: "What are the key changes in this tool between v1.0.0 and v1.2.0?"
AI Internal Workflow:
1. Fetch: Calls get_changelog(start_tag="v1.0.0", end_tag="v1.2.0", url="...").
2. Analyze: Summarizes the commit messages returned by the MCP.
Result: The AI lists the new features, bug fixes, and breaking changes accurately.

Visualizing the Data Flow

sequenceDiagram
    participant User
    participant AI as AI Agent (Claude/Gemini)
    participant MCP as Git-Remote-MCP
    participant GitHub as GitHub API

    User->>AI: "Use library X in my project"
    AI->>MCP: get_tags(url_X, limit=5)
    MCP->>GitHub: Request Tags Metadata
    GitHub-->>MCP: Raw Tags Data
    Note over MCP: Sorting via SemVer Logic
    MCP-->>AI: ["v2.1.0", "v2.0.9", ...]
    AI->>User: "Here is your code using the latest v2.1.0"

Installation

Quick Install (One-Liner)

Install the pre-compiled binary for your OS and Architecture (Linux x86_64/aarch64/armv7 or macOS Intel/Silicon):

curl -fsSL https://raw.githubusercontent.com/HanSoBored/git-mcp-rs/main/install.sh | bash

Build from Source (Developers)

Requires Rust and Cargo:

git clone https://github.com/HanSoBored/git-mcp-rs.git
cd git-mcp-rs
./build.sh

The build.sh script compiles the release and moves it to /usr/local/bin/ automatically.

Configuration

Add this to your MCP settings (e.g., Gemini Cli / Qwen Code config):

{
  "mcpServers": {
    "git-remote": {
      "command": "git_mcp",
      "env": {
        "GITHUB_TOKEN": "your_personal_access_token_here"
      }
    }
  }
}

Note: Providing a GITHUB_TOKEN increases your rate limit from 60 to 5,000 requests per hour.

System Prompt for AI Agents

To maximize the utility of this MCP, add this to your Agent's instructions:

- When asked about a library/dependency, ALWAYS use 'get_tags' with a limit of 5 to verify the latest version.
- Before suggesting code changes, use 'get_file_tree' and 'get_file_content' to understand the target repository's structure and existing logic.
- Do not guess file paths; use 'search_repository' to find definitions.

📄 License

MIT License. Feel free to use and contribute!

MCP Servers