mcp-rag-server is a Model Context Protocol (MCP) server that enables Retrieval Augmented Generation (RAG) capabilities. It empowers Large Language Models (LLMs) to answer questions based on your document content by indexing and retrieving relevant information efficiently.
Overview
what is mcp-rag-server?
The mcp-rag-server is a server that implements Retrieval Augmented Generation (RAG) to answer questions based on user-provided documents. It reads various document formats and utilizes embeddings to generate accurate responses.
how to use mcp-rag-server?
To use the mcp-rag-server, install the dependencies using bun install, then run the server with bun run index.ts. Configure the necessary environment variables for the LLM API and embedding model.
key features of mcp-rag-server?
- Supports multiple document formats (.json, .jsonl, .csv, .txt, .md)
- Efficient document reading and indexing using embeddings
- Combines user queries with relevant document chunks for comprehensive answers
- Configurable with various LLM APIs for response generation
use cases of mcp-rag-server?
- Answering questions based on specific documents or datasets.
- Assisting in research by providing relevant information from large document collections.
- Enhancing customer support by retrieving information from manuals or FAQs.
FAQ from mcp-rag-server?
- What document formats does mcp-rag-server support?
It supports .json, .jsonl, .csv, .txt, and .md formats.
- How does the server generate answers?
It uses a combination of document embeddings and a Large Language Model to generate answers based on user queries.
- Is there a limit to the number of document chunks selected?
By default, the server selects the top 15 relevant chunks based on similarity scores.