MCP Server Practice

🚀 MCP Server Practice

This is a modular server based on FastAPI, powered by FastMCP, which demonstrates the practical applications of AI tools, mathematical operations, web searches, audio response generation, and external API integrations.

🚀 Quick Start

This server showcases a variety of practical applications. You can quickly start exploring its features by running the server with the following command:

mcp.run(transport="stdio")

✨ Features

🔧 Math Tools

• add(a, b): Returns the sum of two integers.
• multiply(a, b): Returns the product of two integers.

🌤️ Weather API Integration

• get_weather(city): Retrieves real-time weather data for the specified city using the Weather API.

🌐 Web Search (via OpenAI Tools)

• web_search(query): Performs a real-time search using OpenAI's web search tool and returns concise results.

🔊 Audio Response Generation

• audio_query(text): Converts a text query into a voice response (in WAV format) using OpenAI's gpt-4o audio feature. The audio file is automatically saved in the /audio folder.

🙋♂️ Custom Resource Example

• greeting://{name}: Uses a resource-based route to return a personalized greeting string.

🔧 Technical Details

Technologies Used

• FastMCP: A server interface for tool-based workflows.
• FastAPI: A web framework for high-performance APIs.
• OpenAI: Provides AI-based search, chat, and audio generation capabilities.
• Weather API: An external API for real-time weather updates.
• Python + dotenv: Used for scripting and secure environment variable handling.

How It Works

Each tool is defined using the @mcp.tool() decorator, making it available as a callable interface to AI agents or CLI tools. When the server runs:

mcp.run(transport="stdio")