Comment by eitanlebras

The hybrid search scoring is really well thought out. One question: are the 0.4/0.4/0.2 weights fixed or configurable? To me it feels like different use cases would want different balances (e.g. a coding assistant would want higher recency weights than a research tool would).