Streamable HTTP MCP Server

# Complete Project Structure ``` mcp-server-project/ ├── mcp_server.py # FastAPI MCP server with SSE support ├── client.py # Azure OpenAI GPT-4o client ├── requirements.txt # Python dependencies ├── Dockerfile # Docker container configuration ├── docker-compose.yml # Docker orchestration ├── .env.example # Environment variables template ├── setup.sh # Setup and utility script └── README.md # Complete documentation ``` ## 📂 File Overview ### Core Files - **`mcp_server.py`**: Main MCP server implementing the MCP protocol with FastAPI and SSE - **`client.py`**: GPT-4o client that connects to the MCP server for tool usage - **`requirements.txt`**: All Python dependencies needed for the project ### Configuration Files - **`Dockerfile`**: Containerizes the application for easy deployment - **`docker-compose.yml`**: Orchestrates the MCP server and client services - **`.env.example`**: Template for Azure OpenAI configuration ### Utility Files - **`setup.sh`**: Convenience script for common operations - **`README.md`**: Complete documentation and usage guide ## 🚀 Quick Start Commands ```bash # 1. Setup environment cp .env.example .env # Edit .env with your Azure OpenAI credentials # 2. Start MCP server ./setup.sh start # 3. Run GPT-4o client ./setup.sh client # 4. Test the server ./setup.sh test ``` ## 🔄 How It Works 1. **MCP Server** (`mcp_server.py`) runs on port 8000 and provides: - Calculator tool for math operations - Weather tool for location-based data - Time tool for current timestamps 2. **GPT-4o Client** (`client.py`) connects to Azure OpenAI and: - Converts MCP tools to OpenAI function format - Handles tool calls from GPT-4o - Forwards requests to MCP server via SSE - Returns results back to GPT-4o for final response 3. **Communication Flow**: ``` User Question → GPT-4o → Tool Call → MCP Server → Tool Result → GPT-4o → Final Answer ``` ## 🛠️ Key Features - **MCP Protocol Compliance**: Fully implements MCP 2024-11-05 specification - **Streaming Support**: Uses Server-Sent Events for real-time communication - **Tool Integration**: Seamless tool calling between GPT-4o and MCP server - **Docker Ready**: Complete containerization for easy deployment - **Error Handling**: Comprehensive error handling and logging - **Extensible**: Easy to add new tools and capabilities This implementation provides a solid foundation for building more complex MCP-based applications with Azure OpenAI integration.