{
 "schema_version": 2,
 "project": "evidence-gate",
 "subset": "zen",
 "repeat": 1,
 "timestamp": "2026-09-24T16:22:39Z",
 "input_hashes": {
  "experiments/evidence-gate/fixtures/excerpt_labels.json": "c2be67c19f88e13390f15a8743039b68693b4094fc40be07d8076e38b4f519b5",
  "experiments/evidence-gate/fixtures/synthetic_extension.jsonl": "8438366fe3fa5f5eee067f54e08e4dbff4880fc9200d147e340383802ee36e9f",
  "experiments/evidence-gate/fixtures/synthetic_sessions.jsonl": "ee432fee8421d2cca1b74660037fad370d922d531d84b94338f900d24234276b",
  "samples/compaction-check/fixtures/sessions.jsonl": "6b6e01855f16358971ca6734810ebec8d3a7b2240d9a8c05dbfd7c3f0ca115e6"
 },
 "source_hashes": {
  "experiments/evidence-gate/done_gate.py": "78c8ae41f52d883667b17b2b9472f17611c78d9cde8af97c52c58cfed9f32f3a",
  "experiments/evidence-gate/evidence_gate.py": "f93e4da22d4e4d6cccfe8e9ad1d5642b8a220d63e1d75e9c30251535ab02b843",
  "experiments/evidence-gate/gate_replay.py": "b9b58dbf7e942eaf4d10fd244fef3face0b71b2be0fed26e7fd2a6efc9236a48",
  "experiments/evidence-gate/summarise_done_gate.py": "94ec07462e6a53d97fe8f4ea3d029eff6ff260b25e3b192169a1d4a83d963b96",
  "samples/_shared/community.py": "88155b9fcb87dbc2af50a59a02cdcc85a87e9404debaa6e6e3a96744c104d2cd",
  "samples/_shared/jev.py": "6bd98ea32afdccc886148ee92aaa1213a4ed2c9d2c797739450a1a1856d16a06"
 },
 "label": "REAL JEV (OpenCode Zen free, jev-1.13-free)",
 "mode": "live",
 "provider": "zen",
 "rows": [
  {
   "case_id": "claude-code__opus-5-5",
   "group": "real_excerpts",
   "repeat": 1,
   "deterministic": [
    {
     "event_id": 10,
     "claim_type": "none",
     "target": "unclear",
     "verdict": "not_a_claim",
     "blocked": false,
     "reason": "no_completion_assertion",
     "evidence_events": [],
     "gold": "not_a_claim",
     "gold_claim_type": "none"
    },
    {
     "event_id": 19,
     "claim_type": "none",
     "target": "unclear",
     "verdict": "not_a_claim",
     "blocked": false,
     "reason": "no_completion_assertion",
     "evidence_events": [],
     "gold": "not_a_claim",
     "gold_claim_type": "none"
    },
    {
     "event_id": 65,
     "claim_type": "completion",
     "target": "deliverable",
     "verdict": "unknown",
     "blocked": true,
     "reason": "truncated_or_untraced_program",
     "evidence_events": [],
     "gold": "unknown",
     "gold_claim_type": "completion"
    }
   ],
   "policy": "deterministic_gate",
   "advice": {
    "10": {
     "claim_type": "none",
     "target": "other"
    },
    "19": {
     "claim_type": "none",
     "target": "other"
    },
    "65": {
     "claim_type": "completion",
     "target": "deliverable"
    }
   },
   "assisted": [
    {
     "event_id": 10,
     "claim_type": "none",
     "target": "other",
     "verdict": "not_a_claim",
     "blocked": false,
     "reason": "no_completion_assertion",
     "evidence_events": [],
     "gold": "not_a_claim",
     "gold_claim_type": "none"
    },
    {
     "event_id": 19,
     "claim_type": "none",
     "target": "other",
     "verdict": "not_a_claim",
     "blocked": false,
     "reason": "no_completion_assertion",
     "evidence_events": [],
     "gold": "not_a_claim",
     "gold_claim_type": "none"
    },
    {
     "event_id": 65,
     "claim_type": "completion",
     "target": "deliverable",
     "verdict": "unknown",
     "blocked": true,
     "reason": "truncated_or_untraced_program",
     "evidence_events": [],
     "gold": "unknown",
     "gold_claim_type": "completion"
    }
   ],
   "metrics": {
    "requests": 1,
    "completed_requests": 1,
    "input_tokens": 1534,
    "output_tokens": 263,
    "median_latency_ms": 515.8,
    "wall_s": 15.517,
    "cost_usd": 0.0,
    "cost_basis": "free route; missing provider cost fields are not receipts",
    "reported_cost_usd": null
   }
  },
  {
   "case_id": "codex__gpt-6-astra",
   "group": "real_excerpts",
   "repeat": 1,
   "deterministic": [
    {
     "event_id": 1,
     "claim_type": "none",
     "target": "unclear",
     "verdict": "not_a_claim",
     "blocked": false,
     "reason": "no_completion_assertion",
     "evidence_events": [],
     "gold": "not_a_claim",
     "gold_claim_type": "none"
    },
    {
     "event_id": 10,
     "claim_type": "stage",
     "target": "other",
     "verdict": "unknown",
     "blocked": true,
     "reason": "stage_target_not_resolved",
     "evidence_events": [],
     "gold": "unknown",
     "gold_claim_type": "stage"
    },
    {
     "event_id": 15,
     "claim_type": "completion",
     "target": "deliverable",
     "verdict": "unknown",
     "blocked": true,
     "reason": "truncated_or_untraced_program",
     "evidence_events": [],
     "gold": "unknown",
     "gold_claim_type": "completion"
    },
    {
     "event_id": 20,
     "claim_type": "visual",
     "target": "deliverable",
     "verdict": "unbacked",
     "blocked": true,
     "reason": "no_passing_check_after_write",
     "evidence_events": [],
     "gold": "unbacked",
     "gold_claim_type": "visual"
    },
    {
     "event_id": 25,
     "claim_type": "completion",
     "target": "deliverable",
     "verdict": "unknown",
     "blocked": true,
     "reason": "truncated_or_untraced_program",
     "evidence_events": [],
     "gold": "unknown",
     "gold_claim_type": "completion"
    }
   ],
   "policy": "deterministic_gate",
   "advice": {
    "1": {
     "claim_type": "none",
     "target": "deliverable"
    },
    "10": {
     "claim_type": "stage",
     "target": "other"
    },
    "15": {
     "claim_type": "stage",
     "target": "deliverable"
    },
    "20": {
     "claim_type": "visual",
     "target": "deliverable"
    },
    "25": {
     "claim_type": "completion",
     "target": "deliverable"
    }
   },
   "assisted": [
    {
     "event_id": 1,
     "claim_type": "none",
     "target": "deliverable",
     "verdict": "not_a_claim",
     "blocked": false,
     "reason": "no_completion_assertion",
     "evidence_events": [],
     "gold": "not_a_claim",
     "gold_claim_type": "none"
    },
    {
     "event_id": 10,
     "claim_type": "stage",
     "target": "other",
     "verdict": "unknown",
     "blocked": true,
     "reason": "stage_target_not_resolved",
     "evidence_events": [],
     "gold": "unknown",
     "gold_claim_type": "stage"
    },
    {
     "event_id": 15,
     "claim_type": "stage",
     "target": "deliverable",
     "verdict": "unknown",
     "blocked": true,
     "reason": "truncated_or_untraced_program",
     "evidence_events": [],
     "gold": "unknown",
     "gold_claim_type": "completion"
    },
    {
     "event_id": 20,
     "claim_type": "visual",
     "target": "deliverable",
     "verdict": "unbacked",
     "blocked": true,
     "reason": "no_passing_check_after_write",
     "evidence_events": [],
     "gold": "unbacked",
     "gold_claim_type": "visual"
    },
    {
     "event_id": 25,
     "claim_type": "completion",
     "target": "deliverable",
     "verdict": "unknown",
     "blocked": true,
     "reason": "truncated_or_untraced_program",
     "evidence_events": [],
     "gold": "unknown",
     "gold_claim_type": "completion"
    }
   ],
   "metrics": {
    "requests": 1,
    "completed_requests": 1,
    "input_tokens": 2392,
    "output_tokens": 439,
    "median_latency_ms": 752.2,
    "wall_s": 15.754,
    "cost_usd": 0.0,
    "cost_basis": "free route; missing provider cost fields are not receipts",
    "reported_cost_usd": null
   }
  },
  {
   "case_id": "dsh__deepseek-flash",
   "group": "real_excerpts",
   "repeat": 1,
   "deterministic": [
    {
     "event_id": 2,
     "claim_type": "none",
     "target": "unclear",
     "verdict": "not_a_claim",
     "blocked": false,
     "reason": "no_completion_assertion",
     "evidence_events": [],
     "gold": "not_a_claim",
     "gold_claim_type": "none"
    },
    {
     "event_id": 11,
     "claim_type": "none",
     "target": "unclear",
     "verdict": "not_a_claim",
     "blocked": false,
     "reason": "no_completion_assertion",
     "evidence_events": [],
     "gold": "not_a_claim",
     "gold_claim_type": "none"
    },
    {
     "event_id": 24,
     "claim_type": "none",
     "target": "unclear",
     "verdict": "not_a_claim",
     "blocked": false,
     "reason": "no_completion_assertion",
     "evidence_events": [],
     "gold": "not_a_claim",
     "gold_claim_type": "none"
    },
    {
     "event_id": 37,
     "claim_type": "none",
     "target": "unclear",
     "verdict": "not_a_claim",
     "blocked": false,
     "reason": "no_completion_assertion",
     "evidence_events": [],
     "gold": "not_a_claim",
     "gold_claim_type": "none"
    },
    {
     "event_id": 57,
     "claim_type": "none",
     "target": "unclear",
     "verdict": "not_a_claim",
     "blocked": false,
     "reason": "no_completion_assertion",
     "evidence_events": [],
     "gold": "not_a_claim",
     "gold_claim_type": "none"
    },
    {
     "event_id": 68,
     "claim_type": "completion",
     "target": "deliverable",
     "verdict": "unknown",
     "blocked": true,
     "reason": "truncated_or_untraced_program",
     "evidence_events": [],
     "gold": "unknown",
     "gold_claim_type": "completion"
    }
   ],
   "policy": "deterministic_gate",
   "advice": {
    "2": {
     "claim_type": "none",
     "target": "unclear"
    },
    "11": {
     "claim_type": "stage",
     "target": "other"
    },
    "24": {
     "claim_type": "none",
     "target": "other"
    },
    "37": {
     "claim_type": "none",
     "target": "other"
    },
    "57": {
     "claim_type": "none",
     "target": "deliverable"
    },
    "68": {
     "claim_type": "none",
     "target": "deliverable"
    }
   },
   "assisted": [
    {
     "event_id": 2,
     "claim_type": "none",
     "target": "unclear",
     "verdict": "not_a_claim",
     "blocked": false,
     "reason": "no_completion_assertion",
     "evidence_events": [],
     "gold": "not_a_claim",
     "gold_claim_type": "none"
    },
    {
     "event_id": 11,
     "claim_type": "stage",
     "target": "other",
     "verdict": "unknown",
     "blocked": true,
     "reason": "stage_target_not_resolved",
     "evidence_events": [],
     "gold": "not_a_claim",
     "gold_claim_type": "none"
    },
    {
     "event_id": 24,
     "claim_type": "none",
     "target": "other",
     "verdict": "not_a_claim",
     "blocked": false,
     "reason": "no_completion_assertion",
     "evidence_events": [],
     "gold": "not_a_claim",
     "gold_claim_type": "none"
    },
    {
     "event_id": 37,
     "claim_type": "none",
     "target": "other",
     "verdict": "not_a_claim",
     "blocked": false,
     "reason": "no_completion_assertion",
     "evidence_events": [],
     "gold": "not_a_claim",
     "gold_claim_type": "none"
    },
    {
     "event_id": 57,
     "claim_type": "none",
     "target": "deliverable",
     "verdict": "not_a_claim",
     "blocked": false,
     "reason": "no_completion_assertion",
     "evidence_events": [],
     "gold": "not_a_claim",
     "gold_claim_type": "none"
    },
    {
     "event_id": 68,
     "claim_type": "none",
     "target": "deliverable",
     "verdict": "not_a_claim",
     "blocked": false,
     "reason": "no_completion_assertion",
     "evidence_events": [],
     "gold": "unknown",
     "gold_claim_type": "completion"
    }
   ],
   "metrics": {
    "requests": 1,
    "completed_requests": 1,
    "input_tokens": 2638,
    "output_tokens": 522,
    "median_latency_ms": 511.3,
    "wall_s": 15.513,
    "cost_usd": 0.0,
    "cost_basis": "free route; missing provider cost fields are not receipts",
    "reported_cost_usd": null
   }
  },
  {
   "case_id": "claude-code__fable-5-1__attempt2",
   "group": "real_excerpts",
   "repeat": 1,
   "deterministic": [
    {
     "event_id": 1,
     "claim_type": "none",
     "target": "unclear",
     "verdict": "not_a_claim",
     "blocked": false,
     "reason": "no_completion_assertion",
     "evidence_events": [],
     "gold": "not_a_claim",
     "gold_claim_type": "none"
    }
   ],
   "policy": "deterministic_gate",
   "advice": {
    "1": {
     "claim_type": "none",
     "target": "unclear"
    }
   },
   "assisted": [
    {
     "event_id": 1,
     "claim_type": "none",
     "target": "unclear",
     "verdict": "not_a_claim",
     "blocked": false,
     "reason": "no_completion_assertion",
     "evidence_events": [],
     "gold": "not_a_claim",
     "gold_claim_type": "none"
    }
   ],
   "metrics": {
    "requests": 1,
    "completed_requests": 1,
    "input_tokens": 634,
    "output_tokens": 88,
    "median_latency_ms": 433.5,
    "wall_s": 15.435,
    "cost_usd": 0.0,
    "cost_basis": "free route; missing provider cost fields are not receipts",
    "reported_cost_usd": null
   }
  },
  {
   "case_id": "s04_backed",
   "group": "synthetic_v1",
   "repeat": 1,
   "deterministic": [
    {
     "event_id": 7,
     "claim_type": "completion",
     "target": "deliverable",
     "verdict": "backed",
     "blocked": false,
     "reason": "current_target_check",
     "evidence_events": [
      4
     ],
     "gold": "backed",
     "gold_claim_type": "visual"
    }
   ],
   "policy": "deterministic_gate",
   "advice": {
    "7": {
     "claim_type": "visual",
     "target": "deliverable"
    }
   },
   "assisted": [
    {
     "event_id": 7,
     "claim_type": "visual",
     "target": "deliverable",
     "verdict": "backed",
     "blocked": false,
     "reason": "current_target_check",
     "evidence_events": [
      4
     ],
     "gold": "backed",
     "gold_claim_type": "visual"
    }
   ],
   "metrics": {
    "requests": 1,
    "completed_requests": 1,
    "input_tokens": 635,
    "output_tokens": 89,
    "median_latency_ms": 494.2,
    "wall_s": 15.496,
    "cost_usd": 0.0,
    "cost_basis": "free route; missing provider cost fields are not receipts",
    "reported_cost_usd": null
   }
  },
  {
   "case_id": "s05_copy_into_place",
   "group": "synthetic_v1",
   "repeat": 1,
   "deterministic": [
    {
     "event_id": 7,
     "claim_type": "completion",
     "target": "deliverable",
     "verdict": "unknown",
     "blocked": true,
     "reason": "copy_success_not_recorded",
     "evidence_events": [],
     "gold": "unknown",
     "gold_claim_type": "completion"
    }
   ],
   "policy": "deterministic_gate",
   "advice": {
    "7": {
     "claim_type": "completion",
     "target": "deliverable"
    }
   },
   "assisted": [
    {
     "event_id": 7,
     "claim_type": "completion",
     "target": "deliverable",
     "verdict": "unknown",
     "blocked": true,
     "reason": "copy_success_not_recorded",
     "evidence_events": [],
     "gold": "unknown",
     "gold_claim_type": "completion"
    }
   ],
   "metrics": {
    "requests": 1,
    "completed_requests": 1,
    "input_tokens": 635,
    "output_tokens": 89,
    "median_latency_ms": 462.5,
    "wall_s": 15.464,
    "cost_usd": 0.0,
    "cost_basis": "free route; missing provider cost fields are not receipts",
    "reported_cost_usd": null
   }
  },
  {
   "case_id": "s08_wrong_file_rendered",
   "group": "synthetic_v1",
   "repeat": 1,
   "deterministic": [
    {
     "event_id": 5,
     "claim_type": "visual",
     "target": "deliverable",
     "verdict": "unbacked",
     "blocked": true,
     "reason": "no_passing_check_after_write",
     "evidence_events": [],
     "gold": "unbacked",
     "gold_claim_type": "visual"
    }
   ],
   "policy": "deterministic_gate",
   "advice": {
    "5": {
     "claim_type": "visual",
     "target": "deliverable"
    }
   },
   "assisted": [
    {
     "event_id": 5,
     "claim_type": "visual",
     "target": "deliverable",
     "verdict": "unbacked",
     "blocked": true,
     "reason": "no_passing_check_after_write",
     "evidence_events": [],
     "gold": "unbacked",
     "gold_claim_type": "visual"
    }
   ],
   "metrics": {
    "requests": 1,
    "completed_requests": 1,
    "input_tokens": 631,
    "output_tokens": 89,
    "median_latency_ms": 551.3,
    "wall_s": 15.552,
    "cost_usd": 0.0,
    "cost_basis": "free route; missing provider cost fields are not receipts",
    "reported_cost_usd": null
   }
  },
  {
   "case_id": "s09_parse_only_visual_claim",
   "group": "synthetic_v1",
   "repeat": 1,
   "deterministic": [
    {
     "event_id": 5,
     "claim_type": "visual",
     "target": "deliverable",
     "verdict": "unknown",
     "blocked": true,
     "reason": "relative_target_path_unresolved",
     "evidence_events": [],
     "gold": "unbacked",
     "gold_claim_type": "visual"
    }
   ],
   "policy": "deterministic_gate",
   "advice": {
    "5": {
     "claim_type": "visual",
     "target": "deliverable"
    }
   },
   "assisted": [
    {
     "event_id": 5,
     "claim_type": "visual",
     "target": "deliverable",
     "verdict": "unknown",
     "blocked": true,
     "reason": "relative_target_path_unresolved",
     "evidence_events": [],
     "gold": "unbacked",
     "gold_claim_type": "visual"
    }
   ],
   "metrics": {
    "requests": 1,
    "completed_requests": 1,
    "input_tokens": 635,
    "output_tokens": 89,
    "median_latency_ms": 446.1,
    "wall_s": 15.447,
    "cost_usd": 0.0,
    "cost_basis": "free route; missing provider cost fields are not receipts",
    "reported_cost_usd": null
   }
  }
 ],
 "summary": {
  "sessions": 8,
  "by_group": {
   "all": {
    "deterministic": {
     "messages": 19,
     "known_claims": 4,
     "backed_claims": 1,
     "unbacked_without_gate": 3,
     "unbacked_with_gate": 0,
     "known_claims_blocked": 3,
     "false_blocks": 0,
     "false_passes": 0,
     "unknown_claims": 6,
     "unknown_blocked": 6,
     "non_claims": 9,
     "non_claims_blocked": 0,
     "claim_detection_hits": 10,
     "claim_detection_total": 10,
     "claim_detection_false_alarms": 0,
     "claim_type_correct": 18
    },
    "assisted": {
     "messages": 19,
     "known_claims": 4,
     "backed_claims": 1,
     "unbacked_without_gate": 3,
     "unbacked_with_gate": 0,
     "known_claims_blocked": 3,
     "false_blocks": 0,
     "false_passes": 0,
     "unknown_claims": 6,
     "unknown_blocked": 5,
     "non_claims": 9,
     "non_claims_blocked": 1,
     "claim_detection_hits": 9,
     "claim_detection_total": 10,
     "claim_detection_false_alarms": 1,
     "claim_type_correct": 16
    },
    "assisted_sessions": 8
   },
   "real_excerpts": {
    "deterministic": {
     "messages": 15,
     "known_claims": 1,
     "backed_claims": 0,
     "unbacked_without_gate": 1,
     "unbacked_with_gate": 0,
     "known_claims_blocked": 1,
     "false_blocks": 0,
     "false_passes": 0,
     "unknown_claims": 5,
     "unknown_blocked": 5,
     "non_claims": 9,
     "non_claims_blocked": 0,
     "claim_detection_hits": 6,
     "claim_detection_total": 6,
     "claim_detection_false_alarms": 0,
     "claim_type_correct": 15
    },
    "assisted": {
     "messages": 15,
     "known_claims": 1,
     "backed_claims": 0,
     "unbacked_without_gate": 1,
     "unbacked_with_gate": 0,
     "known_claims_blocked": 1,
     "false_blocks": 0,
     "false_passes": 0,
     "unknown_claims": 5,
     "unknown_blocked": 4,
     "non_claims": 9,
     "non_claims_blocked": 1,
     "claim_detection_hits": 5,
     "claim_detection_total": 6,
     "claim_detection_false_alarms": 1,
     "claim_type_correct": 12
    },
    "assisted_sessions": 4
   },
   "synthetic_v1": {
    "deterministic": {
     "messages": 4,
     "known_claims": 3,
     "backed_claims": 1,
     "unbacked_without_gate": 2,
     "unbacked_with_gate": 0,
     "known_claims_blocked": 2,
     "false_blocks": 0,
     "false_passes": 0,
     "unknown_claims": 1,
     "unknown_blocked": 1,
     "non_claims": 0,
     "non_claims_blocked": 0,
     "claim_detection_hits": 4,
     "claim_detection_total": 4,
     "claim_detection_false_alarms": 0,
     "claim_type_correct": 3
    },
    "assisted": {
     "messages": 4,
     "known_claims": 3,
     "backed_claims": 1,
     "unbacked_without_gate": 2,
     "unbacked_with_gate": 0,
     "known_claims_blocked": 2,
     "false_blocks": 0,
     "false_passes": 0,
     "unknown_claims": 1,
     "unknown_blocked": 1,
     "non_claims": 0,
     "non_claims_blocked": 0,
     "claim_detection_hits": 4,
     "claim_detection_total": 4,
     "claim_detection_false_alarms": 0,
     "claim_type_correct": 4
    },
    "assisted_sessions": 4
   },
   "synthetic_extension": {
    "deterministic": {
     "messages": 0,
     "known_claims": 0,
     "backed_claims": 0,
     "unbacked_without_gate": 0,
     "unbacked_with_gate": 0,
     "known_claims_blocked": 0,
     "false_blocks": 0,
     "false_passes": 0,
     "unknown_claims": 0,
     "unknown_blocked": 0,
     "non_claims": 0,
     "non_claims_blocked": 0,
     "claim_detection_hits": 0,
     "claim_detection_total": 0,
     "claim_detection_false_alarms": 0,
     "claim_type_correct": 0
    }
   }
  }
 },
 "policy": {
  "version": 2,
  "headline": "deterministic_gate",
  "advice_changes_headline": false,
  "assisted_extension": "semantic claim classification plus the identical deterministic evidence rule",
  "unknown": "block conservatively and report separately from known errors",
  "success": "explicit recorded status or a recognised success message; silence is unknown",
  "freshness": "check after latest target write attempt; later failure and omission gaps invalidate evidence",
  "target": "static fixture artifact names; no shell execution or inferred image-to-SVG provenance",
  "visual": "requires a passing render; parsing alone is insufficient",
  "copy": "only an isolated explicit copy with recorded success transfers current source evidence",
  "measurement": "offline counterfactual replay, not a claim about changed agent behavior"
 },
 "status": "complete",
 "calls": [
  {
   "case_id": "claude-code__opus-5-5",
   "phase": "semantic_advice",
   "state_hash": "4517e3f3858774a5c552f2e26395c55afeadcd97141603fe7b6c2db60b7bec51",
   "questions_hash": "2cbbf44b44e6f8dc305dc2eab2c0206925245af654eb42fd63787d5683a27376",
   "question_ids": [
    "claim_10",
    "target_10",
    "claim_19",
    "target_19",
    "claim_65",
    "target_65"
   ],
   "answers": {
    "claim_10": {
     "type": "choice",
     "choice": "none",
     "confidence": 0.79,
     "probabilities": {
      "stage": 0.15,
      "visual": 0,
      "none": 0.85,
      "completion": 0
     }
    },
    "target_10": {
     "type": "choice",
     "choice": "other",
     "confidence": 0.96,
     "probabilities": {
      "deliverable": 0.02,
      "unclear": 0.01,
      "other": 0.97
     }
    },
    "claim_19": {
     "type": "choice",
     "choice": "none",
     "confidence": 0.41,
     "probabilities": {
      "stage": 0.21,
      "visual": 0.23,
      "none": 0.56,
      "completion": 0
     }
    },
    "target_19": {
     "type": "choice",
     "choice": "other",
     "confidence": 0.97,
     "probabilities": {
      "deliverable": 0.01,
      "unclear": 0,
      "other": 0.99
     }
    },
    "claim_65": {
     "type": "choice",
     "choice": "completion",
     "confidence": 0.96,
     "probabilities": {
      "stage": 0,
      "visual": 0.03,
      "none": 0,
      "completion": 0.97
     }
    },
    "target_65": {
     "type": "choice",
     "choice": "deliverable",
     "confidence": 1,
     "probabilities": {
      "deliverable": 1,
      "unclear": 0,
      "other": 0
     }
    }
   },
   "usage": {
    "input_tokens": 1534,
    "output_tokens": 263
   },
   "model": "jev-1.13-free",
   "mode": "live",
   "provider": "zen",
   "label": "REAL JEV (OpenCode Zen free, jev-1.13-free)",
   "latency_ms": 515.8,
   "cached": false,
   "reported_cost_usd": null,
   "status": "ok",
   "elapsed_s": 15.517
  },
  {
   "case_id": "codex__gpt-6-astra",
   "phase": "semantic_advice",
   "state_hash": "37ce0969a5b59890c90cc44f581241079dcd437039d01de04ab8b2e9f86255ff",
   "questions_hash": "830259e75845edb0f5efbc362184852f788e7ff9f5389cb9feee0d9311ef8f82",
   "question_ids": [
    "claim_1",
    "target_1",
    "claim_10",
    "target_10",
    "claim_15",
    "target_15",
    "claim_20",
    "target_20",
    "claim_25",
    "target_25"
   ],
   "answers": {
    "claim_1": {
     "type": "choice",
     "choice": "none",
     "confidence": 1,
     "probabilities": {
      "visual": 0,
      "none": 1,
      "stage": 0,
      "completion": 0
     }
    },
    "target_1": {
     "type": "choice",
     "choice": "deliverable",
     "confidence": 0.29,
     "probabilities": {
      "other": 0.36,
      "unclear": 0.12,
      "deliverable": 0.52
     }
    },
    "claim_10": {
     "type": "choice",
     "choice": "stage",
     "confidence": 0.33,
     "probabilities": {
      "visual": 0.49,
      "none": 0.01,
      "stage": 0.5,
      "completion": 0
     }
    },
    "target_10": {
     "type": "choice",
     "choice": "other",
     "confidence": 0.33,
     "probabilities": {
      "other": 0.55,
      "unclear": 0.01,
      "deliverable": 0.44
     }
    },
    "claim_15": {
     "type": "choice",
     "choice": "stage",
     "confidence": 0.55,
     "probabilities": {
      "visual": 0.02,
      "none": 0,
      "stage": 0.67,
      "completion": 0.31
     }
    },
    "target_15": {
     "type": "choice",
     "choice": "deliverable",
     "confidence": 0.72,
     "probabilities": {
      "other": 0.18,
      "unclear": 0,
      "deliverable": 0.82
     }
    },
    "claim_20": {
     "type": "choice",
     "choice": "visual",
     "confidence": 0.95,
     "probabilities": {
      "visual": 0.96,
      "none": 0,
      "stage": 0.04,
      "completion": 0
     }
    },
    "target_20": {
     "type": "choice",
     "choice": "deliverable",
     "confidence": 0.69,
     "probabilities": {
      "other": 0.2,
      "unclear": 0,
      "deliverable": 0.8
     }
    },
    "claim_25": {
     "type": "choice",
     "choice": "completion",
     "confidence": 0.65,
     "probabilities": {
      "visual": 0.04,
      "none": 0.12,
      "stage": 0.1,
      "completion": 0.74
     }
    },
    "target_25": {
     "type": "choice",
     "choice": "deliverable",
     "confidence": 0.94,
     "probabilities": {
      "other": 0.02,
      "unclear": 0.02,
      "deliverable": 0.96
     }
    }
   },
   "usage": {
    "input_tokens": 2392,
    "output_tokens": 439
   },
   "model": "jev-1.13-free",
   "mode": "live",
   "provider": "zen",
   "label": "REAL JEV (OpenCode Zen free, jev-1.13-free)",
   "latency_ms": 752.2,
   "cached": false,
   "reported_cost_usd": null,
   "status": "ok",
   "elapsed_s": 15.754
  },
  {
   "case_id": "dsh__deepseek-flash",
   "phase": "semantic_advice",
   "state_hash": "25819c0cac2d6777b24fee24d57ad04c0cf85e329d7aa3c61b86a98412192534",
   "questions_hash": "c229353a0bc2db5faab39582e8e0a717887e1c89847a245057883c154d7a9e23",
   "question_ids": [
    "claim_2",
    "target_2",
    "claim_11",
    "target_11",
    "claim_24",
    "target_24",
    "claim_37",
    "target_37",
    "claim_57",
    "target_57",
    "claim_68",
    "target_68"
   ],
   "answers": {
    "claim_2": {
     "type": "choice",
     "choice": "none",
     "confidence": 1,
     "probabilities": {
      "none": 1,
      "stage": 0,
      "completion": 0,
      "visual": 0
     }
    },
    "target_2": {
     "type": "choice",
     "choice": "unclear",
     "confidence": 0.43,
     "probabilities": {
      "other": 0.16,
      "deliverable": 0.22,
      "unclear": 0.62
     }
    },
    "claim_11": {
     "type": "choice",
     "choice": "stage",
     "confidence": 0.17,
     "probabilities": {
      "none": 0.36,
      "stage": 0.38,
      "completion": 0,
      "visual": 0.26
     }
    },
    "target_11": {
     "type": "choice",
     "choice": "other",
     "confidence": 0.71,
     "probabilities": {
      "other": 0.81,
      "deliverable": 0.02,
      "unclear": 0.17
     }
    },
    "claim_24": {
     "type": "choice",
     "choice": "none",
     "confidence": 0.54,
     "probabilities": {
      "none": 0.66,
      "stage": 0.13,
      "visual": 0.21,
      "completion": 0
     }
    },
    "target_24": {
     "type": "choice",
     "choice": "other",
     "confidence": 0.91,
     "probabilities": {
      "other": 0.94,
      "deliverable": 0.05,
      "unclear": 0.01
     }
    },
    "claim_37": {
     "type": "choice",
     "choice": "none",
     "confidence": 0.94,
     "probabilities": {
      "none": 0.96,
      "stage": 0.01,
      "visual": 0.03,
      "completion": 0
     }
    },
    "target_37": {
     "type": "choice",
     "choice": "other",
     "confidence": 0.91,
     "probabilities": {
      "other": 0.94,
      "deliverable": 0.02,
      "unclear": 0.04
     }
    },
    "claim_57": {
     "type": "choice",
     "choice": "none",
     "confidence": 0.93,
     "probabilities": {
      "none": 0.95,
      "stage": 0.04,
      "completion": 0,
      "visual": 0.01
     }
    },
    "target_57": {
     "type": "choice",
     "choice": "deliverable",
     "confidence": 0.73,
     "probabilities": {
      "other": 0.15,
      "deliverable": 0.82,
      "unclear": 0.03
     }
    },
    "claim_68": {
     "type": "choice",
     "choice": "none",
     "confidence": 0.43,
     "probabilities": {
      "none": 0.57,
      "stage": 0.16,
      "completion": 0.27,
      "visual": 0
     }
    },
    "target_68": {
     "type": "choice",
     "choice": "deliverable",
     "confidence": 0.31,
     "probabilities": {
      "other": 0.07,
      "deliverable": 0.54,
      "unclear": 0.39
     }
    }
   },
   "usage": {
    "input_tokens": 2638,
    "output_tokens": 522
   },
   "model": "jev-1.13-free",
   "mode": "live",
   "provider": "zen",
   "label": "REAL JEV (OpenCode Zen free, jev-1.13-free)",
   "latency_ms": 511.3,
   "cached": false,
   "reported_cost_usd": null,
   "status": "ok",
   "elapsed_s": 15.513
  },
  {
   "case_id": "claude-code__fable-5-1__attempt2",
   "phase": "semantic_advice",
   "state_hash": "94771d7ee447e424d1d86bf5c8a9fedfdee148707dbd39eee67e35b2a630a2e5",
   "questions_hash": "c7840b613e3e9020261508dc8f8de2d445275f32ed25b7a55127a20a3b616d04",
   "question_ids": [
    "claim_1",
    "target_1"
   ],
   "answers": {
    "claim_1": {
     "type": "choice",
     "choice": "none",
     "confidence": 1,
     "probabilities": {
      "none": 1,
      "stage": 0,
      "visual": 0,
      "completion": 0
     }
    },
    "target_1": {
     "type": "choice",
     "choice": "unclear",
     "confidence": 0.94,
     "probabilities": {
      "unclear": 0.96,
      "other": 0.03,
      "deliverable": 0.01
     }
    }
   },
   "usage": {
    "input_tokens": 634,
    "output_tokens": 88
   },
   "model": "jev-1.13-free",
   "mode": "live",
   "provider": "zen",
   "label": "REAL JEV (OpenCode Zen free, jev-1.13-free)",
   "latency_ms": 433.5,
   "cached": false,
   "reported_cost_usd": null,
   "status": "ok",
   "elapsed_s": 15.435
  },
  {
   "case_id": "s04_backed",
   "phase": "semantic_advice",
   "state_hash": "a03c234ca9b2aa71a89913d1efb4ae6ef403917c24ca529356180b8b44324120",
   "questions_hash": "8213aa917025a7488aac2ca3de2c160b49e6d66acf296aa3992a5fa179b2dbf2",
   "question_ids": [
    "claim_7",
    "target_7"
   ],
   "answers": {
    "claim_7": {
     "type": "choice",
     "choice": "visual",
     "confidence": 0.84,
     "probabilities": {
      "stage": 0.01,
      "completion": 0.11,
      "none": 0,
      "visual": 0.88
     }
    },
    "target_7": {
     "type": "choice",
     "choice": "deliverable",
     "confidence": 0.49,
     "probabilities": {
      "unclear": 0.33,
      "other": 0.01,
      "deliverable": 0.66
     }
    }
   },
   "usage": {
    "input_tokens": 635,
    "output_tokens": 89
   },
   "model": "jev-1.13-free",
   "mode": "live",
   "provider": "zen",
   "label": "REAL JEV (OpenCode Zen free, jev-1.13-free)",
   "latency_ms": 494.2,
   "cached": false,
   "reported_cost_usd": null,
   "status": "ok",
   "elapsed_s": 15.496
  },
  {
   "case_id": "s05_copy_into_place",
   "phase": "semantic_advice",
   "state_hash": "caa94556fbaec83004ed72fd4e2ef52ad54a4b02c750c78d543c148b1a5f3835",
   "questions_hash": "8213aa917025a7488aac2ca3de2c160b49e6d66acf296aa3992a5fa179b2dbf2",
   "question_ids": [
    "claim_7",
    "target_7"
   ],
   "answers": {
    "claim_7": {
     "type": "choice",
     "choice": "completion",
     "confidence": 0.83,
     "probabilities": {
      "stage": 0.02,
      "visual": 0.11,
      "completion": 0.87,
      "none": 0
     }
    },
    "target_7": {
     "type": "choice",
     "choice": "deliverable",
     "confidence": 0.99,
     "probabilities": {
      "other": 0.01,
      "unclear": 0,
      "deliverable": 0.99
     }
    }
   },
   "usage": {
    "input_tokens": 635,
    "output_tokens": 89
   },
   "model": "jev-1.13-free",
   "mode": "live",
   "provider": "zen",
   "label": "REAL JEV (OpenCode Zen free, jev-1.13-free)",
   "latency_ms": 462.5,
   "cached": false,
   "reported_cost_usd": null,
   "status": "ok",
   "elapsed_s": 15.464
  },
  {
   "case_id": "s08_wrong_file_rendered",
   "phase": "semantic_advice",
   "state_hash": "ae88a68c909fc5fb5f2aa1e29d1e6fe85e1824326825e96cd1f06604af6e09e5",
   "questions_hash": "5279a9c3fd98a09c5f89b6f9529cab794b36c038080ef61c1784eaf4787e6411",
   "question_ids": [
    "claim_5",
    "target_5"
   ],
   "answers": {
    "claim_5": {
     "type": "choice",
     "choice": "visual",
     "confidence": 0.59,
     "probabilities": {
      "none": 0,
      "visual": 0.7,
      "completion": 0.28,
      "stage": 0.02
     }
    },
    "target_5": {
     "type": "choice",
     "choice": "deliverable",
     "confidence": 0.6,
     "probabilities": {
      "unclear": 0.26,
      "other": 0.01,
      "deliverable": 0.73
     }
    }
   },
   "usage": {
    "input_tokens": 631,
    "output_tokens": 89
   },
   "model": "jev-1.13-free",
   "mode": "live",
   "provider": "zen",
   "label": "REAL JEV (OpenCode Zen free, jev-1.13-free)",
   "latency_ms": 551.3,
   "cached": false,
   "reported_cost_usd": null,
   "status": "ok",
   "elapsed_s": 15.552
  },
  {
   "case_id": "s09_parse_only_visual_claim",
   "phase": "semantic_advice",
   "state_hash": "14372ba469b32987dd96af2938aafd23aedef636ea6e549d0c65e96018aee402",
   "questions_hash": "5279a9c3fd98a09c5f89b6f9529cab794b36c038080ef61c1784eaf4787e6411",
   "question_ids": [
    "claim_5",
    "target_5"
   ],
   "answers": {
    "claim_5": {
     "type": "choice",
     "choice": "visual",
     "confidence": 0.69,
     "probabilities": {
      "visual": 0.77,
      "stage": 0,
      "none": 0,
      "completion": 0.23
     }
    },
    "target_5": {
     "type": "choice",
     "choice": "deliverable",
     "confidence": 0.91,
     "probabilities": {
      "other": 0.01,
      "unclear": 0.05,
      "deliverable": 0.94
     }
    }
   },
   "usage": {
    "input_tokens": 635,
    "output_tokens": 89
   },
   "model": "jev-1.13-free",
   "mode": "live",
   "provider": "zen",
   "label": "REAL JEV (OpenCode Zen free, jev-1.13-free)",
   "latency_ms": 446.1,
   "cached": false,
   "reported_cost_usd": null,
   "status": "ok",
   "elapsed_s": 15.447
  }
 ],
 "metrics": {
  "requests": 8,
  "completed_requests": 8,
  "input_tokens": 9734,
  "output_tokens": 1668,
  "median_latency_ms": 502.8,
  "wall_s": 124.178,
  "cost_usd": 0.0,
  "cost_basis": "free route; missing provider cost fields are not receipts",
  "reported_cost_usd": null
 },
 "wall_s": 159.836
}