mcp-agentic-rag
The project is an MCP server and client designed to facilitate Retrieval-Augmented Generation applications. It provides tools for entity extraction, query refinement, and relevance checking, enhancing the performance of RAG systems.
Overview
This project implements an MCP (Model Context Protocol) server and client for building agentic RAG (Retrieval-Augmented Generation) applications. The server provides tools such as entity extraction, query refinement, and relevance checking to enhance RAG systems. The client demonstrates connection to the server and usage of its tools.
Server
The server uses the FastMCP class from the mcp library and offers tools for time retrieval, entity extraction, query refinement, and relevance checking.
Client
The client uses the ClientSession class from the mcp library to connect to the server and demonstrate tool usage.
Requirements
- Python 3.7+
- openai
- mcp
- dotenv
Installation
Clone the repository, install dependencies, and configure environment variables.
Usage
Start the MCP server and run the MCP client.
License
MIT License.