Python MCP Server Template
A template repository for creating custom Model Context Protocol (MCP) servers in Python that can be integrated with applications like Claude Desktop or Cursor.
README
py-mcp-server-template
This repository is a template to help you create your own MCP (Model Context Protocol) servers in Python. Fork this repository to get started.
Setup with uv
This project uses uv for Python packaging and virtual environment management. If you don't have uv installed, please refer to the official uv installation guide.
-
Clone your forked repository:
git clone https://github.com/YOUR_USERNAME/YOUR_REPOSITORY_NAME.git cd YOUR_REPOSITORY_NAME -
Create and activate the virtual environment:
uvtypically creates a.venvdirectory in your project root.uv venv source .venv/bin/activate # On macOS/Linux # .venv\Scripts\activate # On Windows -
Install dependencies: This project uses
pyproject.tomlto manage dependencies.uv pip install .If you add new dependencies, define them in your
pyproject.tomlfile and run this command again. If you are using arequirements.txtfile for some reason, you can install it withuv pip install -r requirements.txt.
Running the Server
The mcp_server.py script starts the MCP server.
To run the server directly:
uv run python mcp_server.py
Integrating with Claude Desktop or Cursor
To use this MCP server with an application like Claude Desktop or Cursor, you'll need to configure it in the application's settings. The configuration will typically involve specifying the command to run your server.
Here's an example configuration snippet. You'll need to replace /ABSOLUTE/PATH/TO/PARENT/FOLDER/YOUR_REPOSITORY_NAME with the actual absolute path to your project directory on your system.
{
"mcpServers": {
"my-custom-python-server": {
"command": "uv",
"args": [
"run",
"--python",
"/ABSOLUTE/PATH/TO/PARENT/FOLDER/YOUR_REPOSITORY_NAME/.venv/bin/python",
"/ABSOLUTE/PATH/TO/PARENT/FOLDER/YOUR_REPOSITORY_NAME/mcp_server.py"
],
"workingDirectory": "/ABSOLUTE/PATH/TO/PARENT/FOLDER/YOUR_REPOSITORY_NAME"
}
}
}
Explanation of the configuration:
"my-custom-python-server": This is a name you give to your server configuration."command": "uv": Specifiesuvas the command to execute."args": A list of arguments for theuvcommand:"run": Tellsuvto execute a command within its managed environment."--python": Specifies the Python interpreter to use. It's important to point this to the Python interpreter inside youruvvirtual environment (.venv/bin/python)."/ABSOLUTE/PATH/TO/PARENT/FOLDER/YOUR_REPOSITORY_NAME/mcp_server.py": The absolute path to your server script.
"workingDirectory": Specifies the working directory for the server process, which should be your project's root directory.
Important:
- Ensure the paths in the
argsandworkingDirectoryare correct for your system. - If the application cannot locate
uv, you might need to specify its full path in the"command"field. You can typically find this path by runningwhich uvin your terminal on macOS or Linux, orwhere uvon Windows. - The server listens on
stdioby default as configured inmcp_server.py(mcp.run(transport='stdio')), which is typically what applications like Cursor expect.
After configuring, the application should be able to communicate with your Python MCP server.
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。