Weather MCP Server and Client

A simple example of using fastmcp to create a weather MCP (Model Control Protocol) server and client.

Features

• Weather Server: Provides current weather and forecast data for major cities

• Weather Client: Demonstrates how to interact with the MCP server

• Tools Available:

  • get_weather(city) - Get current weather for a city

  • get_forecast(city, days) - Get weather forecast for 1-7 days

  • list_cities() - Get list of available cities

• Resources: Access to weather data via URI-based resources

Related MCP server: MCP Weather Server

Installation

1. Install dependencies:

pip install fastmcp asyncio-extras python-dateutil

2. Or install from requirements:

pip install -r requirements.txt

Usage

Running the Server

python weather_server.py

The server will start and provide weather tools for the following cities:

• New York

• London

• Tokyo

• Sydney

Running the Client

python weather_client.py

Choose between:

1. Demo mode - Shows example interactions

2. Interactive mode - Allows manual testing of commands

Available Cities

The server provides mock weather data for:

• New York

• London

• Tokyo

• Sydney

Example Usage

# Get current weather
weather = await get_weather("New York")

# Get 5-day forecast  
forecast = await get_forecast("London", 5)

# List all available cities
cities = await list_cities()

API Reference

Tools

get_weather(city: str)

Get current weather information for a city.

Returns:

• temperature (°C)

• humidity (%)

• condition (sunny, cloudy, rainy, etc.)

• wind_speed (km/h)

• pressure (hPa)

• timestamp

get_forecast(city: str, days: int = 3)

Get weather forecast for a city.

Parameters:

• city: City name

• days: Number of days (1-7)

Returns:

• List of daily forecasts with temperature highs/lows, conditions, etc.

list_cities()

Get list of available cities.

Returns:

• List of city names

Resources

• weather://current/{city} - Current weather data for a city

Extending the Example

To connect this to a real weather API:

1. Replace the MOCK_WEATHER_DATA with API calls

2. Add API key configuration

3. Handle API rate limits and errors

4. Add more cities and weather parameters

Notes

• This example uses mock data for demonstration

• In production, you'd connect to a real weather API like OpenWeatherMap

• The client examples show the structure but don't make actual MCP calls

• Refer to fastmcp documentation for complete MCP protocol implementation