MCP Weather Server

A Model Context Protocol (MCP) server that provides hourly and daily weather forecasts using the Open-Meteo API.

Acknowledgment

This project is a fork of the original MCP Weather Server by Tim Luka Horstmann. The original version used AccuWeather API, and this fork has been migrated to use Open-Meteo API to provide free weather data access without API keys.

Related MCP server: WeatherMCP

NPM Package

Available on npm as weather-open-meteo-mcp.

Quick Start

No API key required! Open-Meteo provides free access to weather data for non-commercial use.

Simply run the MCP Weather server directly with:

npx -y weather-open-meteo-mcp

Or, for HTTP/REST access via supergateway:

npx -y supergateway --stdio "npx -y weather-open-meteo-mcp" \
  --port 4004 \
  --baseUrl http://127.0.0.1:4004 \
  --ssePath /messages \
  --messagePath /message \
  --cors "*"

MCP Server Config Example

For integration with Claude Desktop or other MCP-compatible clients, add this to your config (e.g. claude_desktop_config.json):

{
  "mcpServers": {
    "weather": {
      "command": "npx",
      "args": ["-y", "weather-open-meteo-mcp"]
    }
  }
}

Overview

This MCP server allows large language models (like Claude) to access real-time weather data. When integrated with an LLM, it enables the model to:

• Fetch accurate, up-to-date weather forecasts

• Provide hourly weather data for the next 12 hours

• Access daily weather forecasts for up to 15 days

• Display data in both metric (°C) and imperial (°F) units

• View temperature, conditions, precipitation information, and other weather details

Available Tools

Hourly Weather Forecast

• Tool name: weather-get_hourly

• Provides hourly forecasts for the next 12 hours

• Parameters:

  • location (required): City or location name

  • units (optional): "metric" (Celsius, default) or "imperial" (Fahrenheit)

Daily Weather Forecast

• Tool name: weather-get_daily

• Provides daily forecasts for up to 15 days

• Parameters:

  • location (required): City or location name

  • days (optional): Number of forecast days (1, 5, 10, or 15; default is 5)

  • units (optional): "metric" (Celsius, default) or "imperial" (Fahrenheit)

Prerequisites

• Node.js ≥18

• pnpm (recommended) or npm

Setup

1. Clone this repository:

   git clone https://github.com/adrianR84/Weather-Open-Meteo-MCP.git
   cd mcp-weather

2. Install dependencies:

   Using pnpm (recommended):

   pnpm install

   Or using npm:

   npm install

3. Build the project:

   npm run build

Usage with Claude Desktop

1. Configure Claude Desktop to use this MCP server:

   • Open Claude Desktop

   • Go to Settings > Developer > Edit Config

   • Add the following to your claude_desktop_config.json:

   {
     "mcpServers": {
       "weather": {
         "command": "npx",
         "args": ["-y", "weather-open-meteo-mcp"]
       }
     }
   }

2. Restart Claude Desktop

3. In a new conversation, enable the MCP server by clicking the plug icon and selecting "weather"

4. Now you can ask Claude for weather forecasts, such as:

   • "What's the hourly weather forecast for New York City?"

   • "Give me the 5-day forecast for London."

   • "What will the weather be like in Tokyo this week in Fahrenheit?"

   • "Will it rain in San Francisco tomorrow?"

Development

• Install dev dependencies: pnpm install

• Lint your code: pnpm run lint

• Build: pnpm run build

• Run tests: pnpm test

• Start in dev mode: pnpm run dev

(You can also use npm instead of pnpm if preferred)

Contributing

Contributions are welcome! Please feel free to submit a Pull Request to adrianR84/Weather-Open-Meteo-MCP.

Future Enhancements

We're always looking to improve the MCP Weather Server. Here are some features we're considering for future releases:

• Extended Hourly Forecasts: Beyond 12 hours, e.g., 24 or 48 hours.

• Weather Alerts: Integration with severe weather alerts.

• Location Autocomplete: Enhanced location searching with autocomplete suggestions.

• Historical Weather Data: Access to past weather conditions.

If you have ideas for other features, feel free to open an issue!

License

This project is licensed under the MIT License - see the LICENSE file for details.