stratt://prompts/
5 units across 2 types: chain, task
| Slug | Type | Status | Fingerprint | Description |
|---|---|---|---|---|
| rag-evidence-answer | chain | draft | ●Verified | Orchestrates rag-query-rewriter, rag-chunk-reranker, rag-evidence-validator, and rag-context-synthesizer; refuses-with-gap on insufficient evidence. |
| rag-chunk-reranker | task | draft | ·Not fingerprinted | Reranks candidate chunks against a query, balancing relevance and MMR-style diversity; preserves input ids and emits scored ranking. |
| rag-context-synthesizer | task | draft | ·Not fingerprinted | Composes a grounded answer from id-keyed chunks; cites by id; refuses on insufficient evidence and reports coverage of all input chunks. |
| rag-evidence-validator | task | draft | ·Not fingerprinted | Audits a RAG answer against its source chunks; pairs each claim to supporting ids and flags unsupported spans with an overall verdict. |
| rag-query-rewriter | task | draft | ·Not fingerprinted | Rewrites a user query for downstream retrieval — expansion, decomposition, or HyDE — emitting a strict JSON plan with rewritten queries and rationale. |