MCP RAG Server
by karaage0703
MCP RAG Server is a Python server with RAG (Retrieval-Augmented Generation) functionality that complies with the Model Context Protocol (MCP). It indexes documents in multiple formats and retrieves relevant information using vector search.
Last updated: N/A
What is MCP RAG Server?
MCP RAG Server is a Python-based server that provides RAG capabilities, adhering to the Model Context Protocol (MCP). It allows indexing of documents in various formats (Markdown, Text, PowerPoint, PDF) and retrieving relevant information based on natural language queries using vector search.
How to use MCP RAG Server?
To use the MCP RAG Server, first install the dependencies and set up PostgreSQL with the pgvector extension. Configure the environment variables in the .env file. Then, start the MCP server using uv run python -m src.main. Use the command-line tool (python -m src.cli) to index documents. Finally, interact with the server using JSON-RPC requests to the search and get_document_count methods.
Key features of MCP RAG Server
MCP server implementation with JSON-RPC over stdio
RAG functionality with multi-format document indexing
Uses multilingual-e5-large model for embeddings
Vector database using PostgreSQL with pgvector
Supports incremental indexing for efficient updates
Provides tools for vector search and document count
Use cases of MCP RAG Server
Question answering system using indexed documents
Information retrieval from a knowledge base
Integrating RAG capabilities into AI tools like Cline/Cursor
Building a search engine for local documents
Creating a chatbot with access to a document repository
FAQ from MCP RAG Server
What document formats are supported?
What document formats are supported?
The server supports Markdown, Text, PowerPoint, and PDF files.
What embedding model is used?
What embedding model is used?
The multilingual-e5-large model is used for generating embeddings.
What is the database used for vector storage?
What is the database used for vector storage?
PostgreSQL with the pgvector extension is used as the vector database.
How do I perform a backup and restore?
How do I perform a backup and restore?
You can back up the PostgreSQL database using pg_dump and restore it using pg_restore. Optionally, you can also back up the processed documents and environment configuration.
How do I index new or changed files?
How do I index new or changed files?
Use the incremental indexing feature with the -i or --incremental flag when running the index command.