Fedora Docs RAG Database

A pre-built RAG (Retrieval-Augmented Generation) database of Fedora documentation, ready for use with local AI assistants.

What is this?

This repository provides a database dump containing vectorized Fedora documentation, suitable for semantic search and RAG-powered Q&A. Built using docs2db.

Features

• 🚀 Ready to use - Download the dump, restore it, and start querying
• 📚 Comprehensive - Includes Quick Docs, Sysadmin Guide, CoreOS, Silverblue, and more
• 🔄 Regularly updated - Rebuilt when upstream documentation changes
• 🔓 Open source - Same license as Fedora documentation

Quick Start

1. Download the database dump

# Download the latest release
curl -LO https://github.com/Lifto/FedoraDocsRAG/releases/latest/download/fedora-docs.sql

2. Restore and query

# Restore the dump (starts PostgreSQL via Podman automatically)
uvx docs2db db-start
uvx docs2db db-restore fedora-docs.sql

# Query the database
uvx docs2db-api query "How do I install packages on Fedora?"

Building from Source

If you want to build the database yourself:

Prerequisites

• Python 3.12
• uv
• Docker or Podman
• Git

Build

# Clone this repository
git clone https://github.com/Lifto/FedoraDocsRAG.git
cd FedoraDocsRAG

# Install dependencies and build
uv sync
uv run python build.py

The build script will:

1. Clone all Fedora documentation repositories
2. Build them with Antora (in a container)
3. Ingest, chunk, and embed using docs2db
4. Create a database dump in dist/fedora-docs.sql

Documentation Sources

This database includes documentation from:

Source	Description
Quick Docs	Common tasks and tutorials
Sysadmin Guide	Server administration
Release Notes	Version-specific changes
CoreOS	Container-focused OS
Silverblue	Immutable desktop
IoT	Internet of Things
And more...	See build.py for full list

License

Database Content (CC-BY-SA 4.0)

The database dump containing Fedora documentation is licensed under the Creative Commons Attribution-ShareAlike 4.0 International License.

This is a derivative work of Fedora Documentation, which is licensed under CC-BY-SA by the Fedora Project.

Build Scripts (Apache 2.0)

The build scripts and tooling in this repository are licensed under the Apache License 2.0.

Contributing

Contributions are welcome! Please:

1. Fork the repository
2. Create a feature branch
3. Submit a pull request

Related Projects

• docs2db - The ingestion pipeline
• docs2db-api - Query API for docs2db databases
• Fedora Docs - The upstream documentation