Knowledge Base / RAG
Ground your agents in your own documents with curated retrieval and RAG plugins.
Ingestion
Parsing, chunking, and indexing plugins — reviewed for the formats your team actually uses.
Retrieval
Vector, hybrid, and metadata-filtered retrieval plugins, with honest notes on when each fits.
Evaluation
Checklists and recipes for measuring answer quality — the step most teams skip.
Getting started
1. Inventory your sources
List the document types and systems your agents must answer from — this decides your ingestion plugins.
2. Pick a retrieval strategy
Vector for semantic search, hybrid when exact facts matter, metadata filters for permissions.
3. Measure before you scale
Run the evaluation checklist on a small set before wiring agents into production.