Langchain Agent With MCP Servers

🚀 Using LangChain Agents with MCP Servers

A LangChain agent integrated with Model Context Protocol (MCP) servers via the MCP adapter.

🚀 Quick Start

This project demonstrates how to build a LangChain agent that uses the MCP protocol to interact with various services:

• Tavily Search: Web search and news search functionality.
• Weather: Simulated weather information retrieval.
• Math: Mathematical expression calculation.

The agent uses the ReAct agent pattern of LangGraph and dynamically selects and uses these tools based on the user's query.

✨ Features

• Graceful Shutdown Mechanism: All MCP servers implement perfect signal handling to ensure clean termination.
• Sub - process Management: The main agent tracks and manages all MCP server sub - processes.
• Error Handling: The entire application has powerful error - handling capabilities.
• Modular Design: Easy to expand and add new MCP servers.

🔧 Technical Details

Graceful Shutdown Mechanism

The project implements a comprehensive graceful shutdown system:

1. Signal Handling: Captures SIGINT and SIGTERM signals to initiate a graceful shutdown.
2. Process Tracking: The main agent maintains a registry of all sub - processes.
3. Clean - up Process: Ensures that all sub - processes are terminated on exit.
4. Shutdown Flag: Each MCP server has a shutdown flag to prevent new operations during shutdown.
5. Asynchronous Collaboration: Uses asyncio to allow ongoing operations to complete as much as possible.

📦 Installation

# Clone the repository
git clone https://github.com/yourusername/langchain-mcp.git
cd langchain-mcp

# Create a virtual environment
python -m venv .venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate

# Install dependencies
pip install -e .

📚 Documentation

Configuration

Create a .env file in the project root directory and add the following variables:

OPENAI_API_KEY=Your OpenAI API key
TAVILY_API_KEY=Your Tavily API key

Usage

Run the agent from the command line:

python src/agent.py

The agent will prompt for a query and then process it using appropriate tools.

Development

To add a new MCP server:

1. Create a new file in src/mcpserver/.
2. Implement the server with perfect signal handling.
3. Update src/mcpserver/__init__.py to expose the new server.
4. Add the new server configuration to src/agent.py.

📄 License

MIT