ednybracho/mcp-server-weather

Weather MCP Server

This repository contains an example server built with FastMCP to fetch weather data from the US National Weather Service. It exposes tools that can be invoked through an MCP client.

Requirements

• Python 3.12 (see .python-version)
• Dependencies listed in pyproject.toml and locked in uv.lock
  • httpx
  • mcp[cli]

Setup

Create a virtual environment and install the project in editable mode:

python -m venv .venv
source .venv/bin/activate
pip install -e .

If you use the uv package manager you can reproduce the locked environment:

uv pip install --system -r uv.lock

Running the server

Start the FastMCP server with:

python weather.py

This launches the server using the stdio transport. The following tools become available to an MCP client:

• get_alerts(state) – Fetch active alerts for a US state.
• get_forecast(latitude, longitude) – Retrieve a forecast for a geographic location.

For a quick sanity check you can also run python main.py, which prints a greeting.

Project layout

• weather.py – Server implementation and API calls.
• pyproject.toml – Project metadata and core dependencies.
• uv.lock – Lock file generated by the uv package manager.
• .python-version – Specifies the required Python version.
• main.py – Simple entry point used for testing the environment.