Verified Recall (high precision)
recall_verifiedRead-only
Searches memory with cross-encoder reranking to verify relevance before answering, reducing false abstains on answerable questions.
Instructions
Like recall_gated, but reranks candidates with a cross-encoder and gates on the rerank relevance score (calibrated under a separate 'verified' domain) — higher precision on answerable questions at a slightly higher latency (~600ms). Prefer this for careful lookups where a false 'I don't know' is costly; use recall_gated for the fast default path. Returns the same answer | verify | abstain + sectioned briefing shape.
Input Schema
TableJSON Schema
| Name | Required | Description | Default |
|---|---|---|---|
| k | No | Max results (default 10) | |
| as_of | No | ISO-8601 date/time for point-in-time recall | |
| query | Yes | Natural-language question or topic to search memory for | |
| scope | No | Project scope id (default 'default') | default |
| space | No | Memory space routing: 'both' (default — private + team), 'private', or 'shared' | both |
Output Schema
TableJSON Schema
| Name | Required | Description | Default |
|---|---|---|---|
| results | No | Matching facts (fact_id, entity, key, value, rationale, source, confidence, valid_from, valid_to, recorded_at, score, kind, parent_id, full_text) | |
| decision | No | 'answer' | 'verify' | 'abstain' — the metamemory gate's verdict | |
| event_id | No | Pass to the memory_feedback tool once the user confirms or corrects the answer | |
| confidence | No | The gate's confidence in the recall | |
| context_block | No | Ready-to-use sectioned briefing: facts with validity ranges + provenance labels, usage instructions, and any whole-policy card bodies | |
| policy_applied | No | True when a governing team policy card superseded/bound this recall |