Demonstrates integrating the Model Control Protocol (MCP) with LangGraph for building AI research assistants.
This project provides a practical demonstration of integrating the Model Control Protocol (MCP) with LangGraph, showcasing its benefits in standardizing tool integration for AI research assistants. It offers two implementations: one using direct API calls and another leveraging a simplified MCP client to wrap the Tavily API. By comparing these approaches, the project highlights how MCP can streamline and improve the modularity of AI agent development. The included scripts allow for side-by-side comparisons and individual executions, providing a hands-on experience with MCP in a LangGraph environment.