MCP Weather Server

A simple Model Context Protocol (MCP) server that provides real-time weather data to AI agents like GitHub Copilot.

Quick Start

Clone the repository:

git clone https://github.com/microsoft/lets-learn-mcp-javascript.git
cd mcp-weather-server

1. Install Dependencies

npm install

2. Run the Server

Test with MCP Inspector:

npx -y @modelcontextprotocol/inspector npx -y tsx main.ts

3. Use with VS Code

1. Open the mcp.jsonfile in .vscode folder
2. Click the start server button above line 4
3. Open Chat mode and select agent and choose a modal that supports MCPs such as Claude Sonnet
4. Type or speak into the chat and ask it what the weather is like in your city

Features

• 🌤️ Real-time weather data for any city
• 🌍 No API key required (uses Open-Meteo)
• 🤖 Works with GitHub Copilot and other MCP-compatible AI tools
• ⚡ Easy to test with MCP Inspector

Usage Examples

Ask GitHub Copilot:

• "What's the weather like in Tokyo?"
• "How's the weather in London today?"
• "Give me the current weather for Paris"

How It Works

The server provides a get-weather tool that:

1. Converts city names to coordinates using geocoding
2. Fetches current weather data from Open-Meteo API
3. Returns structured data that AI agents can format beautifully

Code Structure

// Creates MCP server with weather tool
const server = new McpServer({
  name: "Weather Server",
  version: "1.0.0"
});

// Defines the get-weather tool
server.tool('get-weather', 'Tool to get the weather of a city', ...);

// Connects via stdio transport
const transport = new StdioServerTransport();
server.connect(transport);

Dependencies

• @modelcontextprotocol/sdk - MCP server framework
• zod - Schema validation
• tsx - TypeScript execution (for development)

API Used

• Open-Meteo - Free weather API with no authentication required