A basic implementation of a Model Context Protocol (MCP) server that demonstrates core functionality, including tools and resources. This guide will walk you through the steps to initialize, inspect, and integrate the server.
Before you begin, ensure you have the following installed:
- Python (Version 3.8 or later)
- uv CLI
To verify your installation, run:
python --version
uv --versionTo initialize the project, navigate to a local folder of your choice and launch your terminal (PowerShell or CMD). Then, run:
uv init mcp-server-basicThis will set up the project directory and install the necessary dependencies.
The MCP Inspector is a tool for debugging and monitoring your server. To start the inspector, use the following command:
uv run mcp dev server/weather.pyThis command will launch the inspector and allow you to analyze the server's behavior in real time.
To integrate Claude into your MCP server, execute the installation command:
uv run mcp install server/weather.pyThis will enable additional AI capabilities for your server.
-
uvCommand Not Found- Ensure
uvis installed and added to your system PATH. - Reinstall if necessary:
pip install uv-cli
- Ensure
-
Server Script Errors
- Verify the file path for
server/weather.pyis correct. - Check for missing dependencies.
- Verify the file path for
-
Permission Issues
- Run the terminal as an administrator (Windows) or use
sudo(Linux/Mac).
- Run the terminal as an administrator (Windows) or use
Make sure to create .env file and add the LLM API Keys for successful execution
Happy Coding! 🚀