MCP server · Notes
Local RAG MCP server
by shinpr
Search your own PDF, Word, text and Markdown files from your AI chat, all on your own computer.
This is a small helper that lets your AI search through documents you keep on your own computer, like PDFs, Word files, plain text files and Markdown notes. It is handy if you have a pile of manuals, reports or notes and you want to ask questions about them without uploading anything to a website. Everything stays on your machine.
What is an MCP server? The 30-second version
On its own, your AI can only chat with you using what it already knows. An MCP server is a small helper program that gives your AI a new skill, in this case the ability to look inside your own document folder. Once it is connected, your AI can read and search the files you point it at, and then answer your questions using what it finds. This one runs quietly on your computer, so your documents never leave it.
What this MCP server does
You tell your AI which folder holds your documents, and it builds a searchable index of everything inside. When you ask a question, your AI sends it to this helper, the helper searches the index, and it sends back the most relevant pieces of text. It looks for both meaning and exact words, so it can find a topic even if you phrase it differently, and it can also find exact things like error codes or product names. Then your AI reads those pieces and writes you an answer in the chat.
Click to zoomWhat you can do with it
- Search across many PDFs, Word files, text files and Markdown notes at once
- Ask questions in plain language and get answers based on your own documents
- Find exact terms like error codes, class names or product names
- Read the surrounding paragraphs when a search result needs more context
- See which files are indexed and which are not
- Add a single new document or remove an old one from the index
- Keep everything local so nothing is uploaded to an outside service
Try asking your AI
- “Sync all documents in my folder and tell me when it is done.”
- “What does the API documentation say about authentication?”
- “Find the documented behavior of ERR_CONNECTION_REFUSED.”
- “Read the surrounding chunks for that authentication result.”
What it gives back to you
You get back short pieces of text from your documents, each with the file name, the page or section it came from, and a relevance score. Your AI then turns those pieces into a normal answer in the chat, often quoting the source. If you ask for more context, it can also show the paragraphs just before and after a match. For PDFs with images, it can include those images too if you turned that option on.
Before you start
What you need
- Node.js version 22 or newer installed on your computer
- A folder with the documents you want to search
- Internet access the first time, to download the small search model (about 90 MB)
- An MCP client that supports local servers, like Claude Code, Codex, Cursor or OpenCode
Good to know
It can add, replace and delete entries in its own index, so be careful when asking it to delete or re-sync, and remember it can only read files inside the folder you set as BASE_DIR.
Install it with your AI
Add Local RAG MCP server to your AI, no technical skills needed
You don't install anything by hand. You copy one prompt, paste it into an AI that can work on your computer, and it checks, installs and connects the server for you, asking you when it needs something.
Sign in to get the install prompt
Members get a ready-made prompt that lets the Claude desktop app check Local RAG MCP server, install it and connect it for them, step by step. You don't need any technical skills: you copy, paste and answer a few questions. Your connected AI can also find and install any of the 4,066 MCP servers here for you.
Who it's for
Anyone who works with a lot of PDFs, Word files or notes and wants to ask questions about them without sending them to an online service, especially people in jobs that handle private or confidential documents.
