Weather MCP

Weather Forecast Application Using Model Context Protocol (MCP)

overview

This application is an MCP server that uses the Open-Meteo API to retrieve and display weather forecast data. Through the MCP protocol, weather information can be accessed from LLM applications.

function

Displaying current weather information
View 3-day weather forecast
Get location-based weather information
Access via MCP protocol

set up

Install dependencies:

pip install -e .

Set environment variables: Create a .env file and set the required environment variables.

MCP_API_KEY=your_api_key_here
MCP_BASE_URL=http://localhost:8000

Start the MCP server:

python -m app.server

Run the client:

python -m app.client

MCP Tools

get_weather

Get current weather information and 3-day forecast for a given city.

Parameters:

city : The name of the city for which you want to get weather information (e.g. Tokyo, New York)

Example response:

{
  "location": {
    "name": "東京",
    "country": "日本",
    "latitude": 35.6895,
    "longitude": 139.6917,
    "timezone": "Asia/Tokyo"
  },
  "current": {
    "temperature": 22.5,
    "feels_like": 23.1,
    "humidity": 65,
    "wind_speed": 3.2,
    "wind_direction": 180,
    "precipitation": 0,
    "condition": "晴れ",
    "weather_code": 1
  },
  "forecast": [
    {
      "date": "2023-04-10",
      "max_temp": 24.5,
      "min_temp": 15.2,
      "precipitation": 0,
      "condition": "晴れ",
      "sunrise": "05:30",
      "sunset": "18:15"
    },
    ...
  ]
}

license

MIT