AI导哈

MCP服务器构建工具包

MCP Server Starter Kit是一个用于快速构建、测试和发布Model Context Protocol服务器的脚手架工具包，使开发者能够轻松创建可供AI代理使用的MCP服务器。

  • by AiAgentKarl
  • Python
  • 本地部署
  • 1 星标
  • 更新于
在 GitHub 查看 AiAgentKarl/mcp-starter-kit

使用说明

MCP Server Starter Kit

Build Your Own MCP Server in 5 Minutes

This kit gives you everything you need to build, test, and publish a production-ready MCP (Model Context Protocol) server that AI agents can use.

What's Inside

mcp-starter-kit/
├── template/           # Ready-to-use project template
│   ├── pyproject.toml  # Pre-configured build system
│   ├── src/
│   │   ├── server.py          # FastMCP server (fill in your tools)
│   │   ├── config.py          # Settings & environment variables
│   │   ├── tools/example.py   # Example tool with detailed comments
│   │   └── clients/example_client.py  # Example API client
│   ├── .gitignore
│   └── .mcp.json       # Claude Code integration config
├── examples/
│   ├── joke_server/    # Beginner: Random jokes API
│   ├── github_server/  # Intermediate: GitHub API wrapper
│   └── multi_api_server/ # Advanced: Multiple APIs combined
├── GUIDE.md            # Step-by-step tutorial
└── CHECKLIST.md        # Pre-launch checklist

Quick Start

1. Copy the template

cp -r template/ my-mcp-server/
cd my-mcp-server/

2. Rename placeholders

Open pyproject.toml and replace:

  • your-mcp-server with your server name
  • your_mcp_server with your Python package name
  • Your Name with your name
  • your-github-username with your GitHub username

3. Add your first tool

Edit src/tools/example.py — follow the comments to add your own tools.

4. Test locally

pip install -e .
your-mcp-server  # Runs via stdio

5. Publish to PyPI

pip install build twine
python -m build
python -m twine upload dist/*

That's it. Your MCP server is live and installable via pip install your-mcp-server.

Learning Path

Level Example What You'll Learn
Beginner joke_server/ Basic tool, simple API call
Intermediate github_server/ Auth, error handling, multiple tools
Advanced multi_api_server/ Multiple APIs, caching, complex tools

Requirements

  • Python 3.10+
  • pip install mcp (MCP SDK)

Support

License

MIT License — use this for anything, commercial or personal.

相关 MCP