jaiswalraj03014/Open_Weather_Map

Weather MCP Server

A FastAPI-based Model Context Protocol (MCP) server that provides current weather and forecast information for cities worldwide using the OpenWeatherMap API.

Features

• Current Weather: Get real-time temperature and weather conditions for any city
• 5-Day Forecast: Retrieve weather forecast for the next 5 days
• Smart City Detection: Automatically extracts city names from natural language queries
• Async Implementation: Built with async/await for better performance

Installation

1. Clone or download the project files
2. Install required dependencies:

pip install fastapi uvicorn httpx python-dotenv

3. Set up your OpenWeatherMap API key:
   • Get a free API key from OpenWeatherMap
   • Create a .env file in the project root:

OPENWEATHER_API_KEY=your_api_key_here

Project Structure

weather-mcp/
├── main.py              # FastAPI application with MCP endpoints
├── weather_utils.py     # Weather API utilities
├── .env                 # Environment variables (create this)
└── README.md

Usage

Running the Server

uvicorn main:app --reload --port 8000

The server will start at http://localhost:8000

API Endpoints

POST /mcp

Main endpoint for weather queries. Accepts natural language requests.

Request Body:

{
  "context": "What's the weather in London?"
}

Example Responses:

{
  "response": "The current temperature in London is 15°C with scattered clouds."
}

{
  "response": "2024-01-15 12:00:00: 16°C, light rain\n2024-01-15 15:00:00: 17°C, overcast clouds\n..."
}

GET /metadata

Returns server metadata and endpoint information.

Response:

{
  "name": "Weather Tool",
  "description": "Provides current weather and forecast by city.",
  "endpoint": "http://localhost:8000/mcp"
}

Supported Query Types

The server understands natural language queries like:

• "What's the temperature in Paris?"
• "Show me the weather forecast for New York"
• "Will it rain in Tokyo tomorrow?"
• "How's the weather in Berlin?"

Example Usage

Using curl

# Current weather
curl -X POST "http://localhost:8000/mcp" \
  -H "Content-Type: application/json" \
  -d '{"context": "What is the weather in Tokyo?"}'

# Forecast
curl -X POST "http://localhost:8000/mcp" \
  -H "Content-Type: application/json" \
  -d '{"context": "Show me the forecast for London"}'

Using Python

import httpx
import asyncio

async def test_weather():
    async with httpx.AsyncClient() as client:
        # Current weather
        response = await client.post(
            "http://localhost:8000/mcp",
            json={"context": "What's the temperature in Paris?"}
        )
        print(response.json())
        
        # Forecast
        response = await client.post(
            "http://localhost:8000/mcp",
            json={"context": "Forecast for New York next days"}
        )
        print(response.json())

asyncio.run(test_weather())

Error Handling

• Returns user-friendly error messages when city is not found
• Handles API key authentication errors
• Provides fallback responses for unrecognized queries

Dependencies

• fastapi: Web framework for building APIs
• httpx: Async HTTP client for API requests
• python-dotenv: Environment variable management
• uvicorn: ASGI server for running the application

Configuration

• Default port: 8000
• Temperature units: Metric (°C)
• Forecast: 5 days with 3-hour intervals

Notes

• Requires a valid OpenWeatherMap API key
• City names are automatically extracted from queries
• Default fallback city: Delhi (when no city is detected)

License

This project is open source and available under the .