Weather MCP Server

A lightweight Model Context Protocol (MCP) server that gives AI assistants (like Claude) the ability to fetch real-time weather data for any city using the OpenWeatherMap API.

What It Does

This MCP server exposes a single tool — get_weather — that an AI assistant can call to retrieve the current weather conditions for any city in the world.

Example response:

The weather in London is light rain with 13°C.

Tech Stack

• Python — core language

• FastMCP — framework for building MCP servers quickly

• OpenWeatherMap API — weather data source

• Requests — HTTP client

Getting Started

1. Clone the Repository

git clone https://github.com/Tarun-004/weather-mcp.git
cd weather-mcp

2. Install Dependencies

pip install fastmcp requests

3. Add Your API Key

Get a free API key from OpenWeatherMap, then open server.py and replace the placeholder:

API_KEY = "your_openweathermap_api_key_here"

Never commit your API key to GitHub. Consider using a .env file and the python-dotenv package to keep it safe.

4. Run the Server

python server.py

MCP Configuration

To connect this server to an MCP-compatible client (e.g., Claude Desktop), add the following to your MCP config file:

{
  "mcpServers": {
    "weather": {
      "command": "python",
      "args": ["server.py"]
    }
  }
}

Available Tool

get_weather

Fetches the current weather for a given city.

Returns: A plain-text string with the weather description and temperature in Celsius.

Ideas for Future Improvements

• Add support for forecasts (5-day, hourly)

• Return humidity, wind speed, and UV index

• Support coordinates (lat/lon) in addition to city name

• Use environment variables for the API key

• Add error handling for invalid city names