RAG for your documents
Search a local folder of documents and check answers against them, with no API key
- Pricing
- Open source
- Platforms
- CLI
- Runs
- Locally
- Version
- 3.9.1
- Listing updated
- 3 days ago
About RAG for your documents
RAG for your documents is an open-source MCP server for knowledge and memory. It runs locally on your machine and is started by your AI client.
Point this server at a folder and it builds an in-memory index at startup using an offline TF-IDF embedder, so there is no database, no network call and no key. The assistant stays the language model; the server never calls one. It reads Markdown, HTML, CSV and text files.
Search returns ranked passages with a source file and chunk id, and passages sharing no terms with the query are dropped. The second tool is aimed at hallucination: it takes an answer, optionally with the question and a list of source passages, and checks the answer against them. Without sources it retrieves from your folder.
It belongs to the rag-pipeline-utils Node project, which also offers citations, evaluation and guardrails as a library. Start it with an absolute folder path via npx; it needs Node.js 22 or later.
Install
Run in a terminal
claude mcp add rag-mcp-server -- npx -y @devilsdev/rag-mcp-serverAdd to claude_desktop_config.json
{
"mcpServers": {
"rag-mcp-server": {
"command": "npx",
"args": [
"-y",
"@devilsdev/rag-mcp-server"
]
}
}
}Add to ~/.cursor/mcp.json
{
"mcpServers": {
"rag-mcp-server": {
"command": "npx",
"args": [
"-y",
"@devilsdev/rag-mcp-server"
]
}
}
}Add to .vscode/mcp.json
{
"servers": {
"rag-mcp-server": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"@devilsdev/rag-mcp-server"
]
}
}
}Add to ~/.codeium/windsurf/mcp_config.json
{
"mcpServers": {
"rag-mcp-server": {
"command": "npx",
"args": [
"-y",
"@devilsdev/rag-mcp-server"
]
}
}
}Add to your client's MCP settings (most use this shape)
{
"mcpServers": {
"rag-mcp-server": {
"command": "npx",
"args": [
"-y",
"@devilsdev/rag-mcp-server"
]
}
}
}Tools this server provides (2)
Taken from the project documentation; the running server may differ.
- search_documents
- Search the indexed folder and return ranked passages
- check_answer
- Check an answer against source passages
What it does
- Offline indexing
- Indexes a folder at startup with no key, database or network.
- Passage search
- Returns ranked passages with source file and chunk id.
- Answer checking
- Tests an answer against retrieved or supplied sources.
- Adjustable limits
- topK and question size can be set by input or flag.
- Stdio transport
- The client starts the server directly.
Good for
- Ask questions about a policy folder
- Verify that an assistant's answer is supported by your files
- Prototype document search without cloud services
- Keep sensitive documents local while using an AI client
Pricing
Open source
Free npm package; licence terms not shown
Prices change. Check devilsdev.github.io before you buy.
RAG for your documents FAQ
Does it call an LLM?
No. The connected assistant is the language model.
Does it need an API key?
No key, database or network is needed.
Which Node version is required?
Node.js 22 or later.
How do I add it to Claude Code?
Run claude mcp add docs -- npx -y @devilsdev/rag-mcp-server with your folder path.
Is RAG for your documents free?
Yes. RAG for your documents is open source, so you can use and self-host it for free.
How do I add RAG for your documents to Claude Code?
Run this command in your terminal: claude mcp add rag-mcp-server -- npx -y @devilsdev/rag-mcp-server
What are the best alternatives to RAG for your documents?
The closest alternatives to RAG for your documents are Codebase Memory, XERJ, Claude Code Ultimate Guide, Nautilus Compass Memory and note1.
Alternatives to RAG for your documents

Codebase Memory
MCP serverLocal MCP server that builds a code knowledge graph for AI coding agents
Runs locally- Open source
- 48.7K GitHub stars
- Knowledge and memory

XERJ
MCP serverLocal search engine for AI agents with cited answers over your code and docs
Runs locally- Open source
- 19.2K GitHub stars
- Knowledge and memory

Claude Code Ultimate Guide
MCP serverMCP server for searching the Claude Code Ultimate Guide from any MCP client
Runs locally- Open source
- 6.5K GitHub stars
- AI models and LLM tooling

Nautilus Compass Memory
MCP serverLocal-first long-term memory layer for AI agents, served over MCP
Not checked yet- 5 tools
- Remote
- Open source
- 1.3K GitHub stars
- Knowledge and memory

note1
MCP serverMCP server giving AI assistants search and summaries across your note1 meetings
- 4 tools
- Remote
- Free trial
- 1 GitHub star
- Productivity

ASTra
MCP serverLet agents read, edit, validate and undo Word .docx files with exact operations
Runs locally- 5 tools
- Freemium
- Files and documents