What problem does it solve?
Builds and queries a local knowledge base for mods using a Retrieval-Augmented Generation approach, combining vector search with structured data to answer questions about mod requirements and compatibility without relying on real-time Nexus Mods data.
Core Features & Use Cases
- Local-first RAG: vector search with ChromaDB paired with SQLite for structured metadata to power fast, offline knowledge lookup.
- Hybrid retrieval: semantic search results enriched with SQL-based metadata to improve precision on mod details.
- Mod compendium builder: create a persistent, personalized knowledge base of mods for quick reference and analysis.
- Use cases: answer mod requirements from descriptions, verify compatibility within a local MO2 repository, and discover similar or alternative mods.
Quick Start
Install chromadb and fastembed, then index your MO2 downloads into a local ChromaDB collection and a SQLite database to enable fast, offline semantic and structured searches.