MCP Weather Server

A Model Context Protocol (MCP) server that demonstrates all three core MCP primitives: Tools, Resources, and Prompts. Provides weather information and clothing recommendations using the Open-Meteo API.

Features

Tools

• get_current_weather - Get current weather conditions for any location
• get_forecast - Get multi-day weather forecast (1-16 days)

Resources

• weather://clothing/{condition} - Clothing recommendations for different weather conditions:
  • hot, cold, mild, rainy, snowy, windy

Prompts

• what-to-wear - Combines current weather with clothing recommendations for personalized advice

Installation

npm install

Building

Build the TypeScript project and set executable permissions:

npm run build

The compiled output will be in the ./build directory, with build/index.js as the entry point.

Running the Server

The server communicates via stdio and is designed to be run by an MCP client (like Claude Desktop):

node build/index.js

Configuring with Claude Desktop

Add to your Claude Desktop configuration file:

macOS: ~/Library/Application Support/Claude/claude_desktop_config.json Windows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "weather": {
      "command": "node",
      "args": ["/absolute/path/to/test-mcp-server/build/index.js"]
    }
  }
}

Testing with MCP Inspector

The MCP Inspector provides a web interface for testing your server's tools, resources, and prompts.

Quick Start

# 1. Build your server
npm run build

# 2. Launch the Inspector
npx @modelcontextprotocol/inspector node build/index.js

This will:

• Start the Inspector web interface (typically at http://localhost:5173)
• Connect to your server via stdio
• Open your browser automatically

Testing Features

Once the Inspector is running, you can test:

1. Tools Tab:

   • Test get_current_weather with locations like "San Francisco, CA"
   • Test get_forecast with different day counts (1-16)

2. Resources Tab:

   • Browse clothing resources: weather://clothing/hot, weather://clothing/cold, etc.
   • View resource metadata and contents

3. Prompts Tab:

   • Try the what-to-wear prompt with different locations
   • See how prompts combine tools and resources

4. Notifications Pane:

   • Monitor server logs and debug messages
   • Watch for errors or API issues

Development Workflow

1. Make changes to your code
2. Rebuild: npm run build
3. Refresh the Inspector page to reconnect
4. Test your changes

Architecture

The project follows a feature-modular architecture:

src/
├── index.ts                     # Entry point - transport setup
├── server.ts                    # MCP server configuration
├── api/
│   └── open-meteo-client.ts     # Open-Meteo API client
├── tools/
│   ├── types.ts                 # Tool types & schemas
│   └── weather.ts               # Weather tool implementations
├── resources/
│   ├── types.ts                 # Resource types
│   ├── clothing.ts              # Clothing data
│   └── clothing-resources.ts    # Resource implementations
└── prompts/
    ├── types.ts                 # Prompt types & schemas
    └── what-to-wear.ts          # Prompt implementations

API

Uses the free and open-source Open-Meteo API:

• No authentication required
• Global coverage from 20+ national weather services
• Current conditions and forecasts up to 16 days

Development

See CLAUDE.md for detailed development guidance, including how to add new tools, resources, and prompts.

License

MIT