Weather MCP Server Lbv

🚀 Weather MCP Server

A Model Context Protocol (MCP) server that provides weather information, integrated into the desktop version of Claude. This server offers weather forecasts and alerts for any location within the United States through the National Weather Service API. It was created as part of the MIT MCP Hackathon (April 2025).

🚀 Quick Start

This Weather MCP Server offers weather - related services in the desktop version of Claude. It can provide forecasts and alerts for locations in the US.

✨ Features

• get_forecast: Retrieve a detailed weather forecast for any US location using latitude and longitude.
• get_alerts: Retrieve active weather alerts within a specific state based on the state code.

📦 Installation

Configuration Requirements

• Python 3.10 or higher
• uv package manager
• Desktop version of Claude
• MCP SDK 1.2.0 or higher

Installation Steps

1. Ensure uv is installed:

curl -LsSf https://astral.sh/uv/install.sh | sh

2. Configure the Python environment:

# Create and activate a virtual environment
uv venv
source .venv/bin/activate

# Install dependencies
uv add "mcp[cli]" httpx

📚 Documentation

Configuration Steps

1. Create or edit the configuration file ~/Library/Application Support/Claude/claude_desktop_config.json for the desktop version of Claude:

{
  "mcpServers": {
    "weather": {
      "command": "/path/to/uv",
      "args": [
        "--directory",
        "/absolute/path/to/weather",
        "run",
        "weather.py"
      ]
    }
  }
}

2. Replace /path/to/uv with the actual path to your uv executable file (find it using which uv).
3. Replace /absolute/path/to/weather with the absolute path of this project.

Usage Instructions

1. Launch the desktop version of Claude.
2. Check the hammer 🔨 icon to confirm that the server is connected.
3. Example queries:
   • "What's the weather like in Sacramento?"
   • "What active weather alerts are there in Texas?"

Troubleshooting

• Check Claude's log files at ~/Library/Logs/Claude/mcp*.log.
• Ensure that all paths in the configuration are absolute paths.
• Verify that the server can run independently (use uv run weather.py).
• Confirm that port 8001 is not occupied by other processes.

🔧 Technical Details

API Reference

This server uses the National Weather Service API:

• Base URL: https://api.weather.gov
• Endpoints used:
  • /alerts/active/area/{state} - Retrieve alert information based on the state.
  • /points/{lat},{lon} - Retrieve forecast grid data.
  • /forecast - Retrieve a detailed forecast.

About the MIT MCP Hackathon

This project is part of the MIT MCP Hackathon tutorial (April 2025). The hackathon aims to introduce developers to the Model Context Protocol and help projects or startups launch their own MCP servers. It is part of the MIT NANDA (Networked Agents and Decentralized AI) Hub initiative.

📄 License

This project follows the license statement of the original tutorial: Model Context Protocol