Pypi MCP Server

🚀 PyPI MCP Server

Enable AI assistants to search and access PyPI package information through a simple MCP interface.

The PyPI MCP Server serves as a bridge for AI assistants to connect to the PyPI package repository via the Model Context Protocol (MCP). It allows AI models to programmatically search for Python packages and access their metadata, supporting functions such as retrieving package information, searching for packages, viewing version history, and download statistics.

✨ Features

• 🔎 Package Search: Query PyPI packages by keywords ✅
• 📊 Metadata Access: Obtain detailed metadata for specific packages ✅
• 📦 Version Information: Get all published versions of a package ✅
• 📈 Statistics: Retrieve package download statistics ✅
• 🚀 Efficient Retrieval: Quickly access package information ✅

🚀 Quick Start

Prerequisites

• Python 3.10+
• httpx
• BeautifulSoup4
• mcp-python-sdk
• typing-extensions

Installation

1. Clone the repository:

   git clone https://github.com/JackKuo666/PyPI-MCP-Server.git
   cd PyPI-MCP-Server
   

2. Install the required dependencies:

   pip install -r requirements.txt
   

Run the Server

python pypi_server.py

The server communicates with MCP clients via standard input/output (stdio) and does not require network port configuration.

💻 Usage Examples

Basic Usage

MCP Tools

Get Package Information

get_package_info(package_name: str, version: Optional[str] = None) -> Dict

Get detailed information about a specified package, with an optional version specification.

Search Packages

search_packages(query: str) -> List[Dict]

Search for PyPI packages by keywords.

Get Package Releases

get_package_releases(package_name: str) -> Dict

Get information about all published versions of a specified package.

Get Package Statistics

get_package_stats(package_name: str) -> Dict

Get the download statistics of a package.

🔧 Technical Details

The server uses the MCP protocol to communicate with clients via standard input/output (stdio) and does not require network port configuration.

📚 Documentation

Integration with AI Assistants

Using Claude Desktop

Add the following configuration to claude_desktop_config.json:

{
  "mcpServers": {
    "pypi": {
      "command": "python",
      "args": ["pypi_server.py"]
    }
  }
}

Usage Examples

In your AI assistant, you can call the PyPI MCP tools as follows:

Search for the Flask package using the PyPI tool:
@pypi search_packages("flask")

Get detailed information about a specific package:
@pypi get_package_info("requests")

Get information about a package with a specified version:
@pypi get_package_info("django", "4.2.0")

View all published package versions:
@pypi get_package_releases("numpy")

Get the download statistics of a package:
@pypi get_package_stats("pandas")

📄 License

MIT