mcp-basic-setup-anthropic

Learn MCP (Model Context Protocol)

A learning project demonstrating how to build and use Model Context Protocol (MCP) servers and clients. This project implements a weather service that provides real-time weather alerts and forecasts using the National Weather Service API.

🌟 Features

• Weather Alerts: Get active weather alerts for any US state
• Weather Forecasts: Retrieve detailed weather forecasts for specific coordinates
• MCP Integration: Demonstrates MCP server/client architecture
• AI Assistant: Uses Anthropic's Claude for natural language interaction with weather tools

🛠️ Tech Stack

• Python 3.13+
• FastMCP: For building the MCP server
• Anthropic Claude: For AI-powered interactions
• National Weather Service API: For weather data
• asyncio/httpx: For asynchronous HTTP requests

📦 Installation

1. Clone the repository:

   git clone <your-repo-url>
   cd learn-mcp
   

2. Install dependencies:

   pip install -e .
   

3. Set up environment variables:

   Create a .env file in the project root:

   ANTHROPIC_API_KEY=your_anthropic_api_key_here
   

🚀 Usage

Running the MCP Server

The server provides weather-related tools that can be accessed via MCP:

python server.py

Running the MCP Client

The client connects to the server and provides an interactive chat interface:

python client.py server.py

Available Tools

1. Get Weather Alerts

Get active weather alerts for a specified US state:

await get_alerts("CA")  # Get alerts for California

2. Get Weather Forecast

Get detailed weather forecast for specific coordinates:

await get_forecast(34.0522, -118.2437)  # Get forecast for Los Angeles

💬 Example Interactions

Once you start the client, you can ask natural language questions like:

• "What are the current weather alerts in Texas?"
• "Give me the forecast for New York City coordinates 40.7128, -74.0060"
• "Are there any severe weather warnings in Florida?"

🏗️ Project Structure

learn-mcp/
├── client.py          # MCP client implementation
├── server.py          # MCP server with weather tools
├── pyproject.toml     # Project configuration
├── README.md          # This file
├── LICENSE            # License file
└── uv.lock           # Dependency lock file

🔧 Key Components

Server (server.py)

• Implements MCP server using FastMCP
• Provides weather alert and forecast tools
• Integrates with National Weather Service API
• Handles asynchronous HTTP requests

Client (client.py)

• Connects to MCP server via stdio transport
• Integrates with Anthropic's Claude for AI interactions
• Provides interactive chat loop
• Handles tool calls and responses

📚 Learning Objectives

This project demonstrates:

1. MCP Architecture: Understanding server/client communication
2. Tool Definition: Creating tools that AI can use
3. API Integration: Working with external APIs (NWS)
4. Async Programming: Using asyncio for concurrent operations
5. AI Integration: Combining MCP with language models

🤝 Contributing

This is a learning project! Feel free to:

• Add new weather-related tools
• Improve error handling
• Add tests
• Enhance documentation
• Experiment with different MCP features

📝 License

This project is licensed under the terms specified in the LICENSE file.

🔗 Resources

• Model Context Protocol Documentation
• FastMCP Documentation
• National Weather Service API
• Anthropic Claude API

⚠️ Notes

• This project uses the National Weather Service API, which provides data for US locations only
• An Anthropic API key is required for the AI features
• The project is designed for learning purposes and may need additional error handling for production use.