MCP Rag Server

@kwanLeeFrmVi

Basic information

About MCP Rag Server

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.

#rag #mcp-server

MCP Rag Server config

Add this server to your MCP-compatible client using the configuration below.

{
  "mcpServers": {
    "mcp-rag-server-kwanleefrmvi": {
      "command": "npx",
      "args": [
        "mcp-rag-server"
      ]
    }
  }
}

MCP Rag Server tools

Add documents from directory path or file path for RAG embedding and store to DB. Supported file types: .json, .jsonl, .txt, .md, .csv

Query indexed documents using RAG

Remove a specific document from the index by file path

Remove all documents from the index

List all document paths in the index

Overview

What is mcp-rag-server?

mcp-rag-server is a Model Context Protocol (MCP) server that enables Retrieval Augmented Generation (RAG). It indexes local documents and serves relevant context to Large Language Models, integrating with MCP-compatible clients.

How to use mcp-rag-server?

Install globally (npm install -g mcp-rag-server) or run via npx. Set environment variables (BASE_LLM_API, EMBEDDING_MODEL, VECTOR_STORE_PATH, CHUNK_SIZE) then start the server. Use MCP tools like embedding_documents and query_documents to index and retrieve content. An example MCP client configuration is provided in the README.

Key features of mcp-rag-server

Index documents in .txt, .md, .json, .jsonl, and .csv formats
Customizable chunk size for text splitting
Local vector store powered by SQLite via LangChain’s LibSQLVectorStore
Supports multiple embedding providers (OpenAI, Ollama, Granite, Nomic)
Exposes MCP tools and resources over stdio for seamless integration

Use cases of mcp-rag-server

Build a RAG pipeline for local document corpora
Provide LLMs with context from private or offline files
Index and query large collections of markdown or code documentation
Enable question-answering systems over custom datasets

Frequently asked questions

Which embedding providers are supported?

The server supports OpenAI, Ollama, Granite, and Nomic. Ollama with the `nomic-embed-text` model is recommended for best performance.

What file formats can be indexed?

It supports `.txt`, `.md`, `.json`, `.jsonl`, and `.csv` files.

How does the server store vectors?

Vectors are stored locally in a SQLite database via LangChain’s LibSQLVectorStore. The path is configured with the `VECTOR_STORE_PATH` environment variable.

What configuration variables are required?

Key variables are `BASE_LLM_API`, `EMBEDDING_MODEL`, `VECTOR_STORE_PATH`, and `CHUNK_SIZE`. An optional `LLM_API_KEY` can be set for providers that require it.

How do I run the server?

Install globally (`npm install -g mcp-rag-server`) and run `mcp-rag-server`, or use `npx mcp-rag-server`. Ensure all required environment variables are set before starting.

Comments

More Memory & Knowledge MCP servers

Codebase Memory

DeusData

High-performance code intelligence MCP server. Indexes codebases into a persistent knowledge graph — average repo in milliseconds. 162 languages, sub-ms queries, 99% fewer tokens. Single static binary, zero dependencies.

46.7K

Sensefold

SenseFold Labs

Personal context for every AI: search, read, and write back to your private Markdown library of articles, threads, PDFs, notes, and captured ChatGPT/Claude/Gemini/Grok conversations. Remote server, OAuth 2.1 paste-and-au

Stele

Stele-Dev

Shared memory for AI coding agents. Claude Code, Cursor and Codex read your project's decisions, lessons and tasks before they act.

Neither MCP

stonianua

Project context your AI can query through MCP — with source evidence.

AOCI-CODE

aoci-spec

Local MCP server and CLI that keeps a Git-versioned, model-written index of a codebase and its database schema, so coding agents start each task with the whole system in view.

1.7K