Show HN: Ratel, give agents unlimited tools and skills without context bloat
Summary
Ratel is a context engineering layer for AI agents that reduces tool and skill overload by indexing them into catalogs and selectively exposing only the needed capabilities per turn. It uses BM25-based retrieval to keep context lightweight and provides TypeScript and Python SDKs with examples. The project emphasizes open-source tooling, no vector DB, and benchmarks against tool utilization costs.