Update to Wazu

{
  "name": "query-wazu-memory",
  "description": "Search, inspect, compare, trace, summarize, or map grounded evidence in the user's authorized Wazu personal or cohort memory. Use when a user asks what Wazu knows, requests recent or topic-based recall, needs full document or media evidence, wants related-item discovery through Hive Map, or asks for an answer grounded in Wazu data.",
  "included_files": [
    {
      "relative_path": "agents/openai.yaml",
      "size_in_bytes": 449
    },
    {
      "relative_path": "references/retrieval-routing.md",
      "size_in_bytes": 1748
    },
    {
      "relative_path": "references/wazu-data-model-and-scope.md",
      "size_in_bytes": 2126
    }
  ],
  "skill_md_contents": "---\nname: query-wazu-memory\ndescription: Search, inspect, compare, trace, summarize, or map grounded evidence in the user's authorized Wazu personal or cohort memory. Use when a user asks what Wazu knows, requests recent or topic-based recall, needs full document or media evidence, wants related-item discovery through Hive Map, or asks for an answer grounded in Wazu data.\n---\n\n# Query Wazu Memory\n\nUse Wazu as the evidence system of record. Keep scope explicit, choose the narrowest retrieval lane that matches the request, and hydrate evidence before making authoritative claims.\n\n## Load the Wazu contract\n\n- Read [references/wazu-data-model-and-scope.md](references/wazu-data-model-and-scope.md).\n- Read [references/retrieval-routing.md](references/retrieval-routing.md).\n\n## Execute the workflow\n\n1. Resolve scope.\n   - Call `account_context` first when the connected actor or active scope is unknown.\n   - Call `list_cohorts` when scope is missing or ambiguous.\n   - Personal memory uses `scope_kind=personal`; it never uses a cohort UUID.\n   - Cohort memory uses an authorized cohort UUID returned by Wazu.\n   - Report the returned `_scope`, especially when the request depends on the active cohort.\n2. Choose the smallest sufficient retrieval lane.\n   - Use `latest_search` for recent evidence.\n   - Use `search` for a topic, person, tag, or phrase.\n   - Use `full_search` only when deep vector or cross-modal recall is necessary.\n   - Use Hive Map tools for relationships and adjacent evidence.\n   - Use `get_item` for an exact item or to hydrate a search or map result.\n3. Hydrate evidence.\n   - Call `get_item` before treating a result as authoritative.\n   - Call `get_item_chunks` when the hydrated item is incomplete or chunk provenance matters.\n   - Fetch media only when the user needs to inspect it.\n4. Answer from the returned evidence.\n   - Separate facts, reasonable inference, and missing evidence.\n   - State the scope searched.\n   - Preserve useful item IDs, dates, and source locators.\n   - Summarize billing or metering status without exposing private internal identifiers.\n\n## Guardrails\n\n- Never invent a scope, cohort, item, sender, date, tag, or source.\n- Never infer authorization from a chunk's `cohort_id`; authorization comes from the resolved request scope.\n- Treat tags as classification, not proof of source.\n- Treat an empty result as a valid result, not permission to widen scope silently.\n- Do not silently switch retrieval lanes or fall back from personal to cohort scope.\n- Do not call write or ingestion tools from this skill.\n- Stop and report authorization, metering, or transport failures accurately.\n"
}