# Weather MCP Server

A Model Context Protocol (MCP) server that provides weather information and alerts using the National Weather Service (NWS) API.

## Features

- Get weather alerts for any US state
- Get detailed weather forecasts for any location using latitude and longitude
- Easy-to-use MCP interface
- Docker support for easy deployment

## Prerequisites

- Python 3.13 or higher
- Docker and Docker Compose (for containerized deployment)

## Installation

### Local Development

1. Clone the repository
2. Install dependencies using `uv`:
   ```bash
   pip install uv
   uv venv
   uv pip install .
   uv run src/main.py
   ```

### Docker Deployment

Simply use Docker Compose to build and run the server:

```bash
docker compose up -d
```

The server endpoint will be available at `http://localhost:8001/sse`.

## Available Tools

### get_alerts(state: str)

Get active weather alerts for a US state.

Parameters:
- `state`: Two-letter US state code (e.g., "CA" for California, "NY" for New York)

### get_forecast(latitude: float, longitude: float)

Get detailed weather forecast for a specific location.

Parameters:
- `latitude`: Latitude of the location
- `longitude`: Longitude of the location

## Development

The project uses Docker volumes for development, allowing code changes to be reflected without rebuilding the container. The source code is mounted at `/app/src` in the container.

## Technical Details

- Built with Python's MCP framework
- Uses httpx for async HTTP requests
- Implements proper error handling and request timeouts
- Uses uv for dependency management
- Containerized with Docker for easy deployment

## Environment Variables

- `PYTHONUNBUFFERED=1`: Ensures Python output is sent directly to the terminal without buffering

## Port

The server runs on port 8000 inside the container and is mapped to port 8001 on the host machine.