MCP 服务器

github-analyzer-mcp

Enables Claude to analyze GitHub repositories with tools for health scoring, contributor analysis, issue tracking, code search, and more.

README

🔌 GitHub Repository Analyzer MCP

Banner

<p> <b>An MCP server that gives Claude 8 powerful GitHub analysis tools — analyze any repo, score its health, find issues, search code, and more — directly inside Claude Desktop.</b> </p>

</div>

🌟 What This Does

Once installed, you can ask Claude things like:

"Analyze the health of the langchain repo" "Who are the top contributors to pytorch/pytorch?" "Find all open bug issues in fastapi" "Search for TODO comments in microsoft/vscode" "What languages does the Next.js repo use?" "Show me recent commits to the main branch of tensorflow"

Claude uses this MCP server to fetch real GitHub data and give you deep, structured analysis.

🛠️ Available Tools

Tool	Description
`analyze_repo`	Full repo overview — stars, forks, metadata, activity
`get_contributors`	Top contributors with commit counts
`find_issues`	Open/closed issues with label filtering
`search_code`	Find code patterns across a repo
`get_languages`	Language breakdown with percentages
`analyze_commits`	Recent commit history and author patterns
`get_pull_requests`	Open/merged PRs with review metadata
`repo_health_score`	0-100 health score with improvement suggestions

📦 Installation

Option 1 — pip (recommended)

pip install github-analyzer-mcp

Option 2 — from source

git clone https://github.com/MidhunaMohanraj/github-analyzer-mcp
cd github-analyzer-mcp
pip install -e .

⚙️ Claude Desktop Configuration

Add to your Claude Desktop config file:

macOS: ~/Library/Application Support/Claude/claude_desktop_config.json Windows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "github-analyzer": {
      "command": "github-analyzer-mcp",
      "env": {
        "GITHUB_TOKEN": "your_github_token_here"
      }
    }
  }
}

Then restart Claude Desktop. You'll see the GitHub tools available in Claude's tool panel.

GitHub Token (optional but recommended)

Without a token: 60 requests/hour limit. With a token: 5,000 requests/hour + access to private repos.

Get a free token at: github.com/settings/tokens → Generate new token (classic) → check public_repo scope.

🧪 Test Locally

# Set token (optional)
export GITHUB_TOKEN=your_token_here

# Run the server directly
github-analyzer-mcp

# Or test with MCP inspector
npx @modelcontextprotocol/inspector github-analyzer-mcp

💬 Example Claude Conversations

Repo health check:

You: "Give me a health score for the huggingface/transformers repo" Claude: Uses repo_health_score → Returns 87/100, Grade A, with specific recommendations

Code search:

You: "Search for all uses of 'deprecated' in the requests library" Claude: Uses search_code → Returns file locations and code snippets

Contributor analysis:

You: "Who are the top 5 contributors to kubernetes/kubernetes?" Claude: Uses get_contributors → Returns ranked list with commit counts

📁 Project Structure

github-analyzer-mcp/
├── src/github_analyzer_mcp/
│   ├── __init__.py
│   └── server.py          # 8 MCP tools
├── pyproject.toml          # PyPI config
├── README.md
└── LICENSE