MCP Autogen Sse Stdio

🚀 🤖 AutoGen MCP Server Example: Integration with AutoGen

This repository showcases a practical demonstration of integrating tools via the Model Context Protocol (MCP) within the AutoGen framework.

✨ Features

• Dual MCP Integration: Demonstrates how to connect an AutoGen agent to:
  • A local tool server (math_server.py) using the Stdio transport protocol.
  • A remote tool server (Apify's RAG web browser actor) using the Server - Sent Events (SSE) transport protocol.
• Local Tool Example: A simple calculator (add, multiply) running locally, implemented through math_server.py.
• Remote Tool Example: Utilizes Apify's RAG web browser actor for web search and content retrieval via its MCP server.
• AutoGen Agent: An AssistantAgent configured to use the above toolset to answer user questions.

🎯 Objectives

Showcase the flexibility provided by MCP in enabling AI agents to access hosted tools (both local and remote) through standardized communication protocols (Stdio and SSE).

📋 Scenarios

The example agent can answer two different types of questions:

1. A math question ((3 + 5) x 12?), expected to use the local math_server.py.
2. A request for the latest news ("Summarize the latest news on the negotiations between Iran and the United States..."), expected to use Apify's web - browsing tool.

📚 Libraries and Frameworks Used

• AutoGen: AI agent framework (autogen_agentchat, autogen_core, autogen_ext)
• MCP: Model Context Protocol for tool integration
• Python - dotenv: For environment variable management
• OpenAI API: Provides LLM functionality
• Apify API: Provides web - browsing functionality

📦 Installation

Carefully set up your environment by following these steps:

1. Prerequisites:
   • Ensure that Python 3.12 is installed.
   • Install uv if not already installed:

     pip install uv
     

2. Clone the Repository:

   git clone https://github.com/yourusername/autogen-mcp-example.git
   cd autogen-mcp-example
   

3. Install Dependencies:

   pip install -r requirements.txt
   

4. Configure Environment Variables:
   • Create a config.py file in the project root directory and add the following content:

     API_KEY = "your_api_key_here"
     

   • Set other environment variables as needed.

💻 Usage Examples

🚀 Quick Start

1. Start the Agent:

   python main.py
   

2. Use the Agent:
   • Send a math question:

     curl -X POST "http://localhost:8000/math" -d '{"expression": "(3+5)*2"}'
     

   • Send a news request:

     curl -X POST "http://localhost:8000/news" -d '{"query": "The latest negotiations between Iran and the United States"}'
     

🔧 Technical Details

User → AutoGen Agent → MCP Tools → Results → User

This example demonstrates how to easily integrate different types of tools into an agent using MCP!

⚠️ Important Note

If you encounter issues during installation or operation, please refer to the FAQ or contact the project maintainers.