OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
Agent Memory & RAG

Autorag

AutoRAG 2.0 is a self-evolving librarian agent that indexes configured computer folders and helps agents find information in their files. Its `@autorag/librarian` package offers a CLI and TypeScript API for document search and question answering.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Autorag — OpenViST project profile in Agent Memory & RAG.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use AutoRAG when an agent or application needs to retrieve context from files in configured folders, or when you want a CLI workflow for indexing and searching documents.

Who it is for

  • People who want to search and ask questions across documents in selected folders.
  • TypeScript or JavaScript developers adding document retrieval to an agent or application.

Documented uses

  • Index a folder of supported documents for retrieval.
  • Ask a question about indexed files and inspect answers with source information.
  • Add document search to a TypeScript or JavaScript application.
  • Steer or stop a running streamed search through the TypeScript API.
Before you choose

Requirements & limits

The librarian package is published as `@autorag/librarian` and is identified as Apache-2.0 licensed. The CLI requires Node.js 24 or later, or Bun 1.2 or later.

  • The current AutoRAG 2.0 agent is TypeScript-based; the original Python RAG AutoML tool is maintained separately in the legacy directory.
  • Supported file types vary by parser; BMP and TIFF are not parsed.
  • OCR is disabled by default. Enabling it for photos and scanned PDF pages entails first-use model downloads based on configured languages.
  • On macOS and Linux, fsearch-cli must be installed separately for the faster filename-search path; without it, search falls back to a bounded filesystem walk.
  • The desktop file-manager app is a separate project that depends on the librarian package.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Set up and index a document folder

Configure a search path, build the document index and check indexing status.

Before you start

  • Node.js 24 or later, or Bun 1.2 or later.
  • A folder of documents to index.

Steps

  1. Install the global CLI with bun install -g @autorag/librarian or the command below.

    Documented command
    npm install -g @autorag/librarian

  2. Run autorag init --search-paths ~/Documents/research, replacing the example path with the folder to search.

  3. Run autorag refresh to parse documents and build the search indexes.

  4. Run autorag status to check indexing health.

Expected outcome

The selected folder is configured and indexed for document search, with indexing health available through the status command.

Guide limitations

  • The documented parser list does not include BMP or TIFF.
  • Image files and scanned PDF pages require OCR to be enabled; it is off by default.
  • On macOS and Linux, fsearch-cli is a separate install for faster filename search; without it, filename search uses a slower bounded filesystem walk.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
13 October 2026
Original website title
Marker-Inc-Korea/AutoRAG · Official homepage title · retained in its source language
How we review sourcesSuggest a correction