{
 "name": "AutoFactory-Bench",
 "version": "ef1.2",
 "generatedAt": "2026-10-24T20:03:37+00:00",
 "models": [
  {
   "id": "gpt-6-astra",
   "label": "GPT-6 Astra",
   "org": "OpenAI",
   "harness": "Codex",
   "status": "complete"
  },
  {
   "id": "gpt-6.1-sol",
   "label": "GPT-6.1 Sol",
   "org": "OpenAI",
   "harness": "Codex",
   "status": "complete"
  },
  {
   "id": "claude-fable-5-1",
   "label": "Fable 5.1",
   "org": "Anthropic",
   "harness": "Claude Code",
   "status": "complete"
  },
  {
   "id": "claude-opus-5-5",
   "label": "Opus 5.5",
   "org": "Anthropic",
   "harness": "Claude Code",
   "status": "complete"
  },
  {
   "id": "deepseek-v4.1-flash-ultrafast",
   "label": "DeepSeek V4.1 Flash",
   "org": "DeepSeek",
   "harness": "mini-swe-agent",
   "status": "complete"
  },
  {
   "id": "claude-sonnet-5-5",
   "label": "Sonnet 5.5",
   "org": "Anthropic",
   "harness": "Claude Code",
   "status": "complete"
  },
  {
   "id": "claude-haiku-5-5",
   "label": "Haiku 5.5",
   "org": "Anthropic",
   "harness": "Claude Code",
   "status": "complete"
  },
  {
   "id": "glm-5.3-flash-ultrafast",
   "label": "GLM-5.3 Flash",
   "org": "Z.ai",
   "harness": "mini-swe-agent",
   "status": "complete"
  },
  {
   "id": "gpt-6-luna",
   "label": "GPT-6 Luna",
   "org": "OpenAI",
   "harness": "Codex",
   "status": "complete"
  }
 ],
 "leaderboard": [
  {
   "modelId": "gpt-6-astra",
   "overall": 0.9167,
   "easy": 1.0,
   "medium": 1.0,
   "hard": 0.7222,
   "hardPassAt3": 0.8333,
   "safetyViolationRate": 0.0331,
   "trials": 160,
   "noProgramTrials": 0,
   "costPerTrial": 0.7216
  },
  {
   "modelId": "gpt-6.1-sol",
   "overall": 0.91,
   "easy": 1.0,
   "medium": 1.0,
   "hard": 0.7,
   "hardPassAt3": 0.8667,
   "safetyViolationRate": 0.038,
   "trials": 160,
   "noProgramTrials": 0,
   "costPerTrial": 0.1702
  },
  {
   "modelId": "claude-fable-5-1",
   "overall": 0.8167,
   "easy": 0.9333,
   "medium": 0.925,
   "hard": 0.5556,
   "hardPassAt3": 0.8,
   "safetyViolationRate": 0.1262,
   "trials": 160,
   "noProgramTrials": 0,
   "costPerTrial": 3.6808
  },
  {
   "modelId": "claude-opus-5-5",
   "overall": 0.7508,
   "easy": 0.8667,
   "medium": 0.775,
   "hard": 0.6028,
   "hardPassAt3": 0.8667,
   "safetyViolationRate": 0.0746,
   "trials": 163,
   "noProgramTrials": 0,
   "costPerTrial": 0.5631
  },
  {
   "modelId": "deepseek-v4.1-flash-ultrafast",
   "overall": 0.7333,
   "easy": 0.9333,
   "medium": 0.975,
   "hard": 0.2111,
   "hardPassAt3": 0.3667,
   "safetyViolationRate": 0.2843,
   "trials": 160,
   "noProgramTrials": 0,
   "costPerTrial": 0.0135
  },
  {
   "modelId": "claude-sonnet-5-5",
   "overall": 0.718,
   "easy": 0.7667,
   "medium": 0.9,
   "hard": 0.4267,
   "hardPassAt3": 0.6333,
   "safetyViolationRate": 0.0878,
   "trials": 166,
   "noProgramTrials": 0,
   "costPerTrial": 0.2506
  },
  {
   "modelId": "claude-haiku-5-5",
   "overall": 0.6433,
   "easy": 0.9333,
   "medium": 0.825,
   "hard": 0.1111,
   "hardPassAt3": 0.2,
   "safetyViolationRate": 0.2862,
   "trials": 165,
   "noProgramTrials": 0,
   "costPerTrial": 0.034
  },
  {
   "modelId": "glm-5.3-flash-ultrafast",
   "overall": 0.4033,
   "easy": 0.7667,
   "medium": 0.4,
   "hard": 0.0444,
   "hardPassAt3": 0.1333,
   "safetyViolationRate": 0.2549,
   "trials": 160,
   "noProgramTrials": 0,
   "costPerTrial": 0.0127
  },
  {
   "modelId": "gpt-6-luna",
   "overall": 0.3567,
   "easy": 0.3333,
   "medium": 0.575,
   "hard": 0.0889,
   "hardPassAt3": 0.1667,
   "safetyViolationRate": 0.2748,
   "trials": 160,
   "noProgramTrials": 41,
   "costPerTrial": 0.0088
  }
 ],
 "tasks": [
  {
   "id": "MAT-E01",
   "title": "Product Acquisition with Gripper",
   "aim": "Acquire the single supported product with the parallel gripper.",
   "suite": "material",
   "difficulty": "easy",
   "family": "custody_acquisition",
   "conditions": 3,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {
    "gpt-6-luna": 1
   },
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "every"
   },
   "replays": {
    "failure": "paper-gpt-4o--r1--c0",
    "success": "gpt-6-astra--4TWZ3vM--c0"
   }
  },
  {
   "id": "MAT-E02",
   "title": "Release Product onto Receiver Support",
   "aim": "Release the gripped product onto the receiver support already beneath it.",
   "suite": "material",
   "difficulty": "easy",
   "family": "custody_placement",
   "conditions": 3,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 1.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "every",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "every"
   },
   "replays": {
    "failure": "paper-qwen3.5-plus--r2--c0",
    "success": "gpt-6.1-sol--tdLmRXY--c0"
   }
  },
  {
   "id": "MAT-E03",
   "title": "Selective Product Release to Receiver Cells",
   "aim": "Release only the requested subset of three held products into their respective receiver cells.",
   "suite": "material",
   "difficulty": "easy",
   "family": "custody_placement",
   "conditions": 3,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "every"
   },
   "replays": {
    "failure": "gpt-6-luna--xpfT3Dz--c0",
    "success": "claude-fable-5-1--NtWjLvN--c0"
   }
  },
  {
   "id": "MAT-E04",
   "title": "Docked Product Handoff with Transaction Acknowledgement",
   "aim": "Hand off the docked product using a valid acknowledgement of the current acceptance transaction.",
   "suite": "material",
   "difficulty": "easy",
   "family": "custody_placement",
   "conditions": 3,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 1.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "every",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "every"
   },
   "replays": {
    "failure": "paper-claude-sonnet-5--r1--c1",
    "success": "claude-opus-5-5--spDqeYM--c0"
   }
  },
  {
   "id": "MAT-E05",
   "title": "Align Plate to Datum Faces and Retract Alignment Fingers",
   "aim": "Align the supported plate to both datum faces and retract both alignment fingers.",
   "suite": "material",
   "difficulty": "easy",
   "family": "material_registration",
   "conditions": 3,
   "passAt1": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {
    "gpt-6-luna": 1
   },
   "result": {
    "claude-opus-5-5": "none",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "every"
   },
   "replays": {
    "failure": "paper-gpt-4o--r1--c0",
    "success": "claude-sonnet-5-5--PSupH7B--c0"
   }
  },
  {
   "id": "MAT-E06",
   "title": "Eject Product and Return Pusher Home",
   "aim": "Eject one product from the already aligned pocket and return the pusher home.",
   "suite": "material",
   "difficulty": "easy",
   "family": "single_piece_feed",
   "conditions": 3,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {
    "gpt-6-luna": 1
   },
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "every"
   },
   "replays": {
    "failure": "paper-gpt-4o--r1--c0",
    "success": "claude-sonnet-5-5--qdnveHL--c0"
   }
  },
  {
   "id": "MAT-E07",
   "title": "Plate Transfer to Receiver",
   "aim": "Transfer the entire plate along the prepared continuous support into the receiver and return the pusher.",
   "suite": "material",
   "difficulty": "easy",
   "family": "supported_transfer",
   "conditions": 3,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 1.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "every",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "every"
   },
   "replays": {
    "failure": "paper-gpt-4o--r1--c0",
    "success": "claude-haiku-5-5--v3BwDat--c0"
   }
  },
  {
   "id": "MAT-E08",
   "title": "Convey Product to Receiving Boundary",
   "aim": "Convey the single product until its trailing edge has passed the receiving boundary, then stop it fully supported on the conveyor.",
   "suite": "material",
   "difficulty": "easy",
   "family": "single_segment_transfer",
   "conditions": 3,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 1.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "every",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "every"
   },
   "replays": {
    "failure": "paper-gpt-4o--r2--c1",
    "success": "glm-5.3-flash-ultrafast--Jh9UGNJ--c0"
   }
  },
  {
   "id": "MAT-E09",
   "title": "Liquid Volume Delivery to Receiver",
   "aim": "Deliver the requested actual liquid volume into the receiver through the fixed nozzle.",
   "suite": "material",
   "difficulty": "easy",
   "family": "quantity_transfer",
   "conditions": 3,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {
    "gpt-6-luna": 1
   },
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "glm-5.3-flash-ultrafast--bzXgs9G--c0",
    "success": "gpt-6.1-sol--j6Fo2AR--c0"
   }
  },
  {
   "id": "MAT-E10",
   "title": "Maintain Material Buffer Inventory",
   "aim": "Maintain material buffer inventory within its permitted band throughout one external withdrawal interval.",
   "suite": "material",
   "difficulty": "easy",
   "family": "inventory_regulation",
   "conditions": 3,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 1.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "every",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "every"
   },
   "replays": {
    "failure": "paper-gpt-4o--r1--c1",
    "success": "gpt-6-astra--oki4CfY--c0"
   }
  },
  {
   "id": "MAT-E11",
   "title": "Material Edge Control During External Feed",
   "aim": "Keep the material edge inside the allowed corridor throughout one externally driven feed interval.",
   "suite": "material",
   "difficulty": "easy",
   "family": "material_registration",
   "conditions": 3,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {
    "gpt-6-luna": 1
   },
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "every"
   },
   "replays": {
    "failure": "paper-qwen3.5-plus--r2--c0",
    "success": "claude-fable-5-1--FXcQ72B--c0"
   }
  },
  {
   "id": "MAT-E12",
   "title": "Material Displacement and Recovery Task",
   "aim": "Displace the old residual material into the recovery path until the requested new material reaches the outlet, then stop.",
   "suite": "material",
   "difficulty": "easy",
   "family": "line_identity",
   "conditions": 3,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 1.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "every",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "every"
   },
   "replays": {
    "success": "claude-fable-5-1--A2LMSqB--c0"
   }
  },
  {
   "id": "MAT-E13",
   "title": "Product Routing Based on Key Match",
   "aim": "Route the current single product to the accepted pocket if its key matches the request, otherwise to the bypass pocket.",
   "suite": "material",
   "difficulty": "easy",
   "family": "identity_routing",
   "conditions": 4,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {
    "gpt-6-luna": 1
   },
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "glm-5.3-flash-ultrafast--gnWavrj--c0",
    "success": "claude-sonnet-5-5--FiE5NZ2--c0"
   }
  },
  {
   "id": "MAT-E14",
   "title": "Product Transfer with Load Balancing",
   "aim": "Transfer the single prepared product to an available side that minimizes the absolute left/right load difference after placement.",
   "suite": "material",
   "difficulty": "easy",
   "family": "supported_transfer",
   "conditions": 5,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {
    "gpt-6-luna": 1
   },
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "every"
   },
   "replays": {
    "failure": "paper-gpt-4o--r1--c0",
    "success": "claude-sonnet-5-5--mPthaJS--c0"
   }
  },
  {
   "id": "MAT-E15",
   "title": "Container Opening and Holding Task",
   "aim": "Open the already bilaterally held container to at least the requested actual passage width and maintain it.",
   "suite": "material",
   "difficulty": "easy",
   "family": "material_geometry",
   "conditions": 3,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {
    "gpt-6-luna": 1
   },
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "glm-5.3-flash-ultrafast--haUmaR9--c0",
    "success": "gpt-6-astra--nV9r5Av--c0"
   }
  },
  {
   "id": "MAT-H01",
   "title": "MAT-H01",
   "aim": "Process two A-order and two B-order workpieces using their respective machining cells, one shared single-gripper robot, one inspection fixture and one optional waiting fixture.",
   "suite": "material",
   "difficulty": "hard",
   "family": "shared_dual_machine_tending",
   "conditions": 6,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 0.3333,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.6667,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 0.3333
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 3,
    "claude-sonnet-5-5": 3,
    "claude-haiku-5-5": 3,
    "gpt-6.1-sol": 3,
    "gpt-6-luna": 3,
    "claude-fable-5-1": 3,
    "gpt-6-astra": 3,
    "deepseek-v4.1-flash-ultrafast": 3,
    "glm-5.3-flash-ultrafast": 3
   },
   "noProgramTrials": {
    "gpt-6-luna": 2
   },
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "some",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "some",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "some"
   },
   "replays": {
    "failure": "claude-haiku-5-5--TBXz32Q--c0",
    "success": "gpt-6-luna--hp4MhT3--c0"
   }
  },
  {
   "id": "MAT-H02",
   "title": "MAT-H02",
   "aim": "Maintain accessibility of bypass storage in situations where the FIFO lead and the required types of two kits differ, and transfer actual parts as kit members of the current order.",
   "suite": "material",
   "difficulty": "hard",
   "family": "material_controls_v14",
   "conditions": 9,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.3333,
    "claude-fable-5-1": 0.6667,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 0.3333,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.037,
    "claude-fable-5-1": 0.037,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 3,
    "claude-sonnet-5-5": 3,
    "claude-haiku-5-5": 3,
    "gpt-6.1-sol": 3,
    "gpt-6-luna": 3,
    "claude-fable-5-1": 3,
    "gpt-6-astra": 3,
    "deepseek-v4.1-flash-ultrafast": 3,
    "glm-5.3-flash-ultrafast": 3
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "none",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "some",
    "claude-fable-5-1": "some",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "some",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "claude-fable-5-1--WuUvyyj--c5",
    "success": "claude-sonnet-5-5--4QyAnYf--c0"
   }
  },
  {
   "id": "MAT-H03",
   "title": "MAT-H03",
   "aim": "While preserving the product and jig separately, connect inspection, product acquisition, and actual recovery of the empty jig, and confirm the full loading of the lift and the point at which the jig can be reused.",
   "suite": "material",
   "difficulty": "hard",
   "family": "material_remaining_v10",
   "conditions": 6,
   "passAt1": {
    "claude-opus-5-5": 0.6667,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 0.3333,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 0.3333
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 1.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.3333
   },
   "trials": {
    "claude-opus-5-5": 3,
    "claude-sonnet-5-5": 3,
    "claude-haiku-5-5": 3,
    "gpt-6.1-sol": 3,
    "gpt-6-luna": 3,
    "claude-fable-5-1": 3,
    "gpt-6-astra": 3,
    "deepseek-v4.1-flash-ultrafast": 3,
    "glm-5.3-flash-ultrafast": 3
   },
   "noProgramTrials": {
    "gpt-6-luna": 3
   },
   "result": {
    "claude-opus-5-5": "some",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "some",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "some"
   },
   "replays": {
    "failure": "glm-5.3-flash-ultrafast--UM7MGQM--c0",
    "success": "claude-opus-5-5--DroS9eU--c0"
   }
  },
  {
   "id": "MAT-H04",
   "title": "MAT-H04",
   "aim": "Connect the separate arrival and preparation of packaging materials and contents, verify both openings retention, pair ID, and actual inclusion depth to accept the completed packaging.",
   "suite": "material",
   "difficulty": "hard",
   "family": "material_controls_v14",
   "conditions": 9,
   "passAt1": {
    "claude-opus-5-5": 0.6667,
    "claude-sonnet-5-5": 0.6667,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 0.3333,
    "glm-5.3-flash-ultrafast": 0.3333
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.3333,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.3333,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 3,
    "claude-sonnet-5-5": 3,
    "claude-haiku-5-5": 3,
    "gpt-6.1-sol": 3,
    "gpt-6-luna": 3,
    "claude-fable-5-1": 3,
    "gpt-6-astra": 3,
    "deepseek-v4.1-flash-ultrafast": 3,
    "glm-5.3-flash-ultrafast": 3
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "some",
    "claude-sonnet-5-5": "some",
    "claude-haiku-5-5": "none",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "some",
    "glm-5.3-flash-ultrafast": "some"
   },
   "replays": {
    "failure": "claude-haiku-5-5--JDUyaWY--c0",
    "success": "deepseek-v4.1-flash-ultrafast--t832TM5--c0"
   }
  },
  {
   "id": "MAT-H05",
   "title": "MAT-H05",
   "aim": "When the pockets of other work stages rotate together, maintain the retraction of all tools and the physical pair/inspection generation, and accept the finished product with the actual ejector.",
   "suite": "material",
   "difficulty": "hard",
   "family": "material_controls_v14",
   "conditions": 9,
   "passAt1": {
    "claude-opus-5-5": 0.6667,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.3333,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.3333,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0741,
    "claude-haiku-5-5": 0.6667,
    "gpt-6.1-sol": 0.3333,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.6667,
    "gpt-6-astra": 0.6667,
    "deepseek-v4.1-flash-ultrafast": 0.3333,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "trials": {
    "claude-opus-5-5": 3,
    "claude-sonnet-5-5": 3,
    "claude-haiku-5-5": 3,
    "gpt-6.1-sol": 3,
    "gpt-6-luna": 3,
    "claude-fable-5-1": 3,
    "gpt-6-astra": 3,
    "deepseek-v4.1-flash-ultrafast": 3,
    "glm-5.3-flash-ultrafast": 3
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "some",
    "claude-sonnet-5-5": "none",
    "claude-haiku-5-5": "none",
    "gpt-6.1-sol": "some",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "some",
    "gpt-6-astra": "none",
    "deepseek-v4.1-flash-ultrafast": "none",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "gpt-6.1-sol--iZXPncE--c0",
    "success": "claude-opus-5-5--dVUC5XT--c0"
   }
  },
  {
   "id": "MAT-H06",
   "title": "MAT-H06",
   "aim": "While controlling reel exchange and leading-end connection, continue downstream cut-piece production with dancer storage, and exclude the connection section from normal orders to separately accept it as actual scrap.",
   "suite": "material",
   "difficulty": "hard",
   "family": "material_controls_v14",
   "conditions": 9,
   "passAt1": {
    "claude-opus-5-5": 0.6667,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.3333,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 0.3333,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.3704,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.6667,
    "claude-fable-5-1": 0.6667,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.3333,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 3,
    "claude-sonnet-5-5": 3,
    "claude-haiku-5-5": 3,
    "gpt-6.1-sol": 3,
    "gpt-6-luna": 3,
    "claude-fable-5-1": 3,
    "gpt-6-astra": 3,
    "deepseek-v4.1-flash-ultrafast": 3,
    "glm-5.3-flash-ultrafast": 3
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "some",
    "claude-sonnet-5-5": "none",
    "claude-haiku-5-5": "none",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "some",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "some",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "deepseek-v4.1-flash-ultrafast--DP7Usrm--c0",
    "success": "gpt-6-astra--3yK6Duf--c0"
   }
  },
  {
   "id": "MAT-H07",
   "title": "Core preparation·registration winding·independent roll extraction cell",
   "aim": "Register and combine the A/B webs into the requested rolls, then secure and deliver each roll.",
   "suite": "material",
   "difficulty": "hard",
   "family": "material_priority_wind_tape",
   "conditions": 11,
   "passAt1": {
    "claude-opus-5-5": 0.6667,
    "claude-sonnet-5-5": 0.3333,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.3333,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.3333,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.6667,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0303,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.6667,
    "glm-5.3-flash-ultrafast": 0.6667
   },
   "trials": {
    "claude-opus-5-5": 3,
    "claude-sonnet-5-5": 3,
    "claude-haiku-5-5": 3,
    "gpt-6.1-sol": 3,
    "gpt-6-luna": 3,
    "claude-fable-5-1": 3,
    "gpt-6-astra": 3,
    "deepseek-v4.1-flash-ultrafast": 3,
    "glm-5.3-flash-ultrafast": 3
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "some",
    "claude-sonnet-5-5": "some",
    "claude-haiku-5-5": "none",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "some",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "none",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "claude-fable-5-1--uhokmny--c5",
    "success": "gpt-6.1-sol--ZvPFAjj--c0"
   }
  },
  {
   "id": "MAT-H08",
   "title": "MAT-H08",
   "aim": "Actually disconnect the connecting edge shared by the product area and the residue, and separately complete the support acceptance of the product and the storage·winding of the perforated same residue.",
   "suite": "material",
   "difficulty": "hard",
   "family": "material_controls_v14",
   "conditions": 9,
   "passAt1": {
    "claude-opus-5-5": 0.75,
    "claude-sonnet-5-5": 0.8,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.3333,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.25,
    "claude-sonnet-5-5": 0.0222,
    "claude-haiku-5-5": 0.4,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.3704,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.7407,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 4,
    "claude-sonnet-5-5": 5,
    "claude-haiku-5-5": 5,
    "gpt-6.1-sol": 3,
    "gpt-6-luna": 3,
    "claude-fable-5-1": 3,
    "gpt-6-astra": 3,
    "deepseek-v4.1-flash-ultrafast": 3,
    "glm-5.3-flash-ultrafast": 3
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "some",
    "claude-sonnet-5-5": "some",
    "claude-haiku-5-5": "none",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "some",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "none",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "deepseek-v4.1-flash-ultrafast--HmfftcJ--c0",
    "success": "claude-fable-5-1--cNvMRgf--c0"
   }
  },
  {
   "id": "MAT-H09",
   "title": "MAT-H09",
   "aim": "While the two filling lines share a common tank and common capping head, adjust batch replenishment, container arrival, and exit takeover delay, preserving the actual received amount per container and capping attribution.",
   "suite": "material",
   "difficulty": "hard",
   "family": "material_stations_v7_liquid",
   "conditions": 6,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 0.3333,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.6667,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.6667,
    "glm-5.3-flash-ultrafast": 0.3333
   },
   "trials": {
    "claude-opus-5-5": 3,
    "claude-sonnet-5-5": 5,
    "claude-haiku-5-5": 5,
    "gpt-6.1-sol": 3,
    "gpt-6-luna": 3,
    "claude-fable-5-1": 3,
    "gpt-6-astra": 3,
    "deepseek-v4.1-flash-ultrafast": 3,
    "glm-5.3-flash-ultrafast": 3
   },
   "noProgramTrials": {
    "gpt-6-luna": 1
   },
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "none",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "some",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "claude-haiku-5-5--yZFqQaa--c0",
    "success": "claude-sonnet-5-5--Hd3xPAo--c0"
   }
  },
  {
   "id": "MAT-H10",
   "title": "MAT-H10",
   "aim": "Physically exchange the finite supply magazine while preserving the magazine inventory, outgoing extraction transactions, and two consumption requests separately.",
   "suite": "material",
   "difficulty": "hard",
   "family": "material_stations_v7_magazine",
   "conditions": 6,
   "passAt1": {
    "claude-opus-5-5": 0.6667,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.6667,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.3333,
    "gpt-6-astra": 0.6667,
    "deepseek-v4.1-flash-ultrafast": 0.3333,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.6667,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.6667,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.3333,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "trials": {
    "claude-opus-5-5": 3,
    "claude-sonnet-5-5": 3,
    "claude-haiku-5-5": 3,
    "gpt-6.1-sol": 3,
    "gpt-6-luna": 3,
    "claude-fable-5-1": 3,
    "gpt-6-astra": 3,
    "deepseek-v4.1-flash-ultrafast": 3,
    "glm-5.3-flash-ultrafast": 3
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "some",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "none",
    "gpt-6.1-sol": "some",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "some",
    "gpt-6-astra": "some",
    "deepseek-v4.1-flash-ultrafast": "some",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "claude-haiku-5-5--868qDbH--c0",
    "success": "gpt-6-astra--EEG9abd--c0"
   }
  },
  {
   "id": "MAT-H11",
   "title": "MAT-H11",
   "aim": "After flipping the product, distinguish the holding pads that cover the working surface, and release the pads after the edge holder takes over the support to complete double-sided work.",
   "suite": "material",
   "difficulty": "hard",
   "family": "material_controls_v14",
   "conditions": 9,
   "passAt1": {
    "claude-opus-5-5": 0.6667,
    "claude-sonnet-5-5": 0.3333,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.6667,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.3333,
    "claude-fable-5-1": 0.7037,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.3333
   },
   "trials": {
    "claude-opus-5-5": 3,
    "claude-sonnet-5-5": 3,
    "claude-haiku-5-5": 3,
    "gpt-6.1-sol": 3,
    "gpt-6-luna": 3,
    "claude-fable-5-1": 3,
    "gpt-6-astra": 3,
    "deepseek-v4.1-flash-ultrafast": 3,
    "glm-5.3-flash-ultrafast": 3
   },
   "noProgramTrials": {
    "gpt-6-luna": 1
   },
   "result": {
    "claude-opus-5-5": "some",
    "claude-sonnet-5-5": "some",
    "claude-haiku-5-5": "none",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "none",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "none",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "claude-fable-5-1--Zaex6vb--c1",
    "success": "gpt-6.1-sol--62BQjXo--c0"
   }
  },
  {
   "id": "MAT-H12",
   "title": "MAT-H12",
   "aim": "Before accepting different production requests that cannot be interrupted, sum the current occupancy and confirmed undelivered residues, and independently adjust the good product exit and shared empty exchange.",
   "suite": "material",
   "difficulty": "hard",
   "family": "material_stations_v7_scrap",
   "conditions": 6,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.3333,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 0.3333
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 3,
    "claude-sonnet-5-5": 3,
    "claude-haiku-5-5": 3,
    "gpt-6.1-sol": 3,
    "gpt-6-luna": 3,
    "claude-fable-5-1": 3,
    "gpt-6-astra": 3,
    "deepseek-v4.1-flash-ultrafast": 3,
    "glm-5.3-flash-ultrafast": 3
   },
   "noProgramTrials": {
    "gpt-6-luna": 1
   },
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "some",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "some"
   },
   "replays": {
    "failure": "glm-5.3-flash-ultrafast--FwKrmuo--c0",
    "success": "claude-fable-5-1--j3D3afK--c0"
   }
  },
  {
   "id": "MAT-H13",
   "title": "MAT-H13",
   "aim": "Control independent single-piece feeding and bundle forming to secure actual members, and coordinate forming, gripper/chuck transfer, film by specification, tape, and exit operations.",
   "suite": "material",
   "difficulty": "hard",
   "family": "material_controls_v14",
   "conditions": 10,
   "passAt1": {
    "claude-opus-5-5": 0.6667,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.3333,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.3333,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.3333,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.3333,
    "gpt-6-luna": 0.3333,
    "claude-fable-5-1": 0.6667,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.6667,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 3,
    "claude-sonnet-5-5": 3,
    "claude-haiku-5-5": 3,
    "gpt-6.1-sol": 3,
    "gpt-6-luna": 3,
    "claude-fable-5-1": 3,
    "gpt-6-astra": 3,
    "deepseek-v4.1-flash-ultrafast": 3,
    "glm-5.3-flash-ultrafast": 3
   },
   "noProgramTrials": {
    "gpt-6-luna": 2
   },
   "result": {
    "claude-opus-5-5": "some",
    "claude-sonnet-5-5": "none",
    "claude-haiku-5-5": "none",
    "gpt-6.1-sol": "some",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "some",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "none",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "gpt-6.1-sol--U3v48Jj--c0",
    "success": "gpt-6-astra--4GgwrGR--c0"
   }
  },
  {
   "id": "MAT-H14",
   "title": "MAT-H14",
   "aim": "While tray arrival, the different completion order of two kit operations, and the shared completion scale and acceptance/hold exit delays overlap, connect the before and after actual measured masses of the same tray and hand it over to the exit matching the net increase.",
   "suite": "material",
   "difficulty": "hard",
   "family": "material_stations_v7_weigh",
   "conditions": 8,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 1.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 3,
    "claude-sonnet-5-5": 3,
    "claude-haiku-5-5": 3,
    "gpt-6.1-sol": 3,
    "gpt-6-luna": 3,
    "claude-fable-5-1": 3,
    "gpt-6-astra": 3,
    "deepseek-v4.1-flash-ultrafast": 3,
    "glm-5.3-flash-ultrafast": 3
   },
   "noProgramTrials": {
    "gpt-6-luna": 1
   },
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "every",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "glm-5.3-flash-ultrafast--HvAyEBz--c0",
    "success": "deepseek-v4.1-flash-ultrafast--NDZFLLT--c0"
   }
  },
  {
   "id": "MAT-H15",
   "title": "MAT-H15",
   "aim": "Preserve the layer order and actual accumulated top surface of independent A/B orders, and after the lower fork takes support, transfer the completed bundle to the external support.",
   "suite": "material",
   "difficulty": "hard",
   "family": "material_controls_v14",
   "conditions": 9,
   "passAt1": {
    "claude-opus-5-5": 0.6667,
    "claude-sonnet-5-5": 0.6667,
    "claude-haiku-5-5": 0.3333,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.2593,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.6667,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.6667,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "trials": {
    "claude-opus-5-5": 3,
    "claude-sonnet-5-5": 3,
    "claude-haiku-5-5": 3,
    "gpt-6.1-sol": 3,
    "gpt-6-luna": 3,
    "claude-fable-5-1": 3,
    "gpt-6-astra": 3,
    "deepseek-v4.1-flash-ultrafast": 3,
    "glm-5.3-flash-ultrafast": 3
   },
   "noProgramTrials": {
    "gpt-6-luna": 1
   },
   "result": {
    "claude-opus-5-5": "some",
    "claude-sonnet-5-5": "some",
    "claude-haiku-5-5": "some",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "none",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "claude-opus-5-5--oBQAovF--c0",
    "success": "claude-haiku-5-5--hMR4Rct--c0"
   }
  },
  {
   "id": "MAT-M01",
   "title": "Web Delivery to Receiver with Buffer Preservation",
   "aim": "Deliver the requested length of web to the receiver while preserving the buffer's usable inventory.",
   "suite": "material",
   "difficulty": "medium",
   "family": "material.v6.buffer",
   "conditions": 6,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 1.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "every",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "every"
   },
   "replays": {
    "failure": "paper-qwen3.5-plus--r1--c0",
    "success": "claude-opus-5-5--joxjbcw--c0"
   }
  },
  {
   "id": "MAT-M02",
   "title": "Plate Orientation and Placement Task",
   "aim": "Place the plate at the receiver with its actual orientation equal to the requested angle.",
   "suite": "material",
   "difficulty": "medium",
   "family": "material.v6.orientation",
   "conditions": 6,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {
    "gpt-6-luna": 1
   },
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "glm-5.3-flash-ultrafast--JnXwGU6--c0",
    "success": "claude-opus-5-5--teEUEC9--c0"
   }
  },
  {
   "id": "MAT-M03",
   "title": "Align Workpiece on Receiver",
   "aim": "Place the workpiece at the receiver aligned to its reference faces.",
   "suite": "material",
   "difficulty": "medium",
   "family": "material.v6.datum",
   "conditions": 6,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 1.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "none",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "every",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "claude-haiku-5-5--hkkKpeG--c0",
    "success": "deepseek-v4.1-flash-ultrafast--WWGh8YY--c0"
   }
  },
  {
   "id": "MAT-M04",
   "title": "Deliver Inspected Workpieces to Receivers",
   "aim": "Deliver both inspected workpieces to their corresponding accept or reject receivers.",
   "suite": "material",
   "difficulty": "medium",
   "family": "material.v6.classification",
   "conditions": 6,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {
    "gpt-6-luna": 1
   },
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "glm-5.3-flash-ultrafast--DBBrbWN--c0",
    "success": "claude-sonnet-5-5--X8Ag28X--c0"
   }
  },
  {
   "id": "MAT-M05",
   "title": "Four-Layer Material Stack Construction",
   "aim": "Build a four-layer stack in the requested bottom-to-top material order.",
   "suite": "material",
   "difficulty": "medium",
   "family": "material.v6.stack",
   "conditions": 6,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 1.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "none",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "every",
    "claude-fable-5-1": "none",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "claude-fable-5-1--jGGJh7n--c0",
    "success": "claude-sonnet-5-5--2cJ8WXD--c0"
   }
  },
  {
   "id": "MAT-M06",
   "title": "Liquid Volume Transfer to Receiving Vessel",
   "aim": "Transfer the requested liquid volume into the receiving vessel.",
   "suite": "material",
   "difficulty": "medium",
   "family": "material.metered_transfer",
   "conditions": 6,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "every"
   },
   "replays": {
    "failure": "gpt-6-luna--KoLFG39--c0",
    "success": "claude-opus-5-5--PtPbSg9--c0"
   }
  },
  {
   "id": "MAT-M07",
   "title": "Material Placement into Receiving Nests",
   "aim": "Place both held materials into the receiving nests at their requested relative spacing.",
   "suite": "material",
   "difficulty": "medium",
   "family": "material.pose_handoff",
   "conditions": 6,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 1.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "every",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "every"
   },
   "replays": {
    "failure": "paper-qwen3.5-plus--r2--c0",
    "success": "gpt-6-astra--wuxGwAb--c0"
   }
  },
  {
   "id": "MAT-M08",
   "title": "Store New Plate in Vacant Slot",
   "aim": "Store one new plate in a vacant slot while preserving the existing stored plates.",
   "suite": "material",
   "difficulty": "medium",
   "family": "material.vacant_slot",
   "conditions": 6,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {
    "gpt-6-luna": 1
   },
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "glm-5.3-flash-ultrafast--BeMfdYi--c0",
    "success": "claude-fable-5-1--Pt6QbEW--c0"
   }
  },
  {
   "id": "MAT-M09",
   "title": "Material Transfer from Sender to Receiver",
   "aim": "Transfer the currently held material from the sender to the receiver.",
   "suite": "material",
   "difficulty": "medium",
   "family": "material.transaction_handoff",
   "conditions": 6,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 1.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "every",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "glm-5.3-flash-ultrafast--WeBE6Ds--c0",
    "success": "claude-haiku-5-5--GQwNzLK--c0"
   }
  },
  {
   "id": "MAT-M10",
   "title": "Material Delivery and Return Task",
   "aim": "Deliver the selected materials to the receiver and return the unselected materials to their corresponding source nests.",
   "suite": "material",
   "difficulty": "medium",
   "family": "material.requested_partition",
   "conditions": 6,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 1.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "every",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "every"
   },
   "replays": {
    "failure": "paper-gpt-4o--r1--c0",
    "success": "claude-opus-5-5--cevyxax--c0"
   }
  },
  {
   "id": "MAT-M11",
   "title": "Content Insertion into Flexible Receiver",
   "aim": "Insert the content into the flexible receiver while preserving a usable opening until transfer equipment is clear.",
   "suite": "material",
   "difficulty": "medium",
   "family": "material.maintained_aperture",
   "conditions": 3,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 1.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 2,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "every",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "paper-gpt-4o--r1--c0",
    "success": "gpt-6-astra--aSSq6pV--c0"
   }
  },
  {
   "id": "MAT-M12",
   "title": "Product Removal and Scrap Retention Task",
   "aim": "Remove the product to the receiver and leave the four surrounding scraps separately held on the upper carrier.",
   "suite": "material",
   "difficulty": "medium",
   "family": "material.separate_carrier_custody",
   "conditions": 3,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 1.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "none",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "every",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "claude-haiku-5-5--SmH2N6M--c0",
    "success": "gpt-6-luna--pPKf5nN--c0"
   }
  },
  {
   "id": "MAT-M13",
   "title": "Eject Item from Indexed Carousel",
   "aim": "Eject the item with the requested identity from the indexed carousel.",
   "suite": "material",
   "difficulty": "medium",
   "family": "material.cyclic_carrier_correspondence",
   "conditions": 3,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "every"
   },
   "replays": {
    "failure": "gpt-6-luna--H55quwQ--c0",
    "success": "claude-fable-5-1--GRXqcrJ--c0"
   }
  },
  {
   "id": "MAT-M14",
   "title": "Product Supply and Launch Coordination",
   "aim": "Supply exactly one product into the launch nest, withdraw the supply finger, and time its transfer into the continuously moving receiving pocket.",
   "suite": "material",
   "difficulty": "medium",
   "family": "material.moving_receiver_capture",
   "conditions": 5,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 1.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "every",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "every"
   },
   "replays": {
    "failure": "paper-qwen3.5-plus--r2--c0",
    "success": "glm-5.3-flash-ultrafast--PfNh2Kr--c0"
   }
  },
  {
   "id": "MAT-M15",
   "title": "Batch Accumulation and Dispatch Control",
   "aim": "Accumulate exactly the requested net quantity while distinguishing forward, reverse, and partial passages, then dispatch the fully supported batch on the receiving tray.",
   "suite": "material",
   "difficulty": "medium",
   "family": "material.directional_crossing",
   "conditions": 5,
   "passAt1": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {
    "gpt-6-luna": 1
   },
   "result": {
    "claude-opus-5-5": "none",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "every"
   },
   "replays": {
    "failure": "claude-opus-5-5--JUTFFQF--c0",
    "success": "claude-fable-5-1--hM25Dmb--c0"
   }
  },
  {
   "id": "MAT-M16",
   "title": "Web Length Feeding and Securing Task",
   "aim": "Feed the requested web length while keeping its actual edge within the receiving corridor, then secure it.",
   "suite": "material",
   "difficulty": "medium",
   "family": "material.continuous_edge_registration",
   "conditions": 3,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 1.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.3333
   },
   "trials": {
    "claude-opus-5-5": 2,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "every",
    "claude-fable-5-1": "none",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "glm-5.3-flash-ultrafast--gbzSdS7--c1",
    "success": "deepseek-v4.1-flash-ultrafast--49aFXbm--c0"
   }
  },
  {
   "id": "MAT-M17",
   "title": "Material Transfer to Receiver",
   "aim": "Transfer the requested material identity to the receiver without sending previous material there.",
   "suite": "material",
   "difficulty": "medium",
   "family": "material.line_content_identity",
   "conditions": 3,
   "passAt1": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 1.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "none",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "every",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "glm-5.3-flash-ultrafast--ZQe3gw4--c0",
    "success": "claude-sonnet-5-5--gAP2w9r--c0"
   }
  },
  {
   "id": "MAT-M18",
   "title": "Tape Transfer to Receiver",
   "aim": "With a single start request, transfer one tape segment of length D0 (300–440 mm) to the receiving support.",
   "suite": "material",
   "difficulty": "medium",
   "family": "material_tape_separation",
   "conditions": 5,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 2,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {
    "gpt-6-luna": 1
   },
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "every"
   },
   "replays": {
    "failure": "paper-gpt-5.5--r2--c0",
    "success": "glm-5.3-flash-ultrafast--6wMaY9L--c0"
   }
  },
  {
   "id": "MAT-M19",
   "title": "Load Materials into Two-Sided Receiver with Balance",
   "aim": "Load all four materials into the two-sided receiver with balanced total mass and depart.",
   "suite": "material",
   "difficulty": "medium",
   "family": "material.load_distribution",
   "conditions": 4,
   "passAt1": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 1.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "none",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "none",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "every",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "every"
   },
   "replays": {
    "failure": "paper-claude-sonnet-5--r1--c0",
    "success": "gpt-6-luna--9i8QC2g--c0"
   }
  },
  {
   "id": "MAT-M20",
   "title": "Board Transfer to Receiver with Span Constraint",
   "aim": "Transfer the board to the receiver without exceeding its allowed unsupported span.",
   "suite": "material",
   "difficulty": "medium",
   "family": "material.continuous_support_span",
   "conditions": 3,
   "passAt1": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 2,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {
    "gpt-6-luna": 1
   },
   "result": {
    "claude-opus-5-5": "none",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "every"
   },
   "replays": {
    "failure": "paper-gpt-5.5--r2--c0",
    "success": "gpt-6-astra--qyY5Ygn--c0"
   }
  },
  {
   "id": "MOT-E01",
   "title": "Linear Axis Positioning Task",
   "aim": "Move the prepared linear axis to the requested absolute position and hold it at rest.",
   "suite": "motion",
   "difficulty": "easy",
   "family": "motion_easy_unit",
   "conditions": 4,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "none",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "none",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "every"
   },
   "replays": {
    "failure": "claude-fable-5-1--69Q3vsn--c3",
    "success": "claude-opus-5-5--YgzHFrh--c0"
   }
  },
  {
   "id": "MOT-E02",
   "title": "Home Return and Coordinate Validity Maintenance",
   "aim": "Execute home return and maintain a valid coordinate stop state.",
   "suite": "motion",
   "difficulty": "easy",
   "family": "motion_easy_unit",
   "conditions": 4,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {
    "gpt-6-luna": 1
   },
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "every"
   },
   "replays": {
    "failure": "paper-gpt-4o--r3--c0",
    "success": "claude-opus-5-5--rs7GbRJ--c0"
   }
  },
  {
   "id": "MOT-E03",
   "title": "Axis Deceleration and Stop Before Limit",
   "aim": "Decelerate and stop the moving axis before the limit position and hold.",
   "suite": "motion",
   "difficulty": "easy",
   "family": "motion_easy_unit",
   "conditions": 4,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 1.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "none",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "every",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "every"
   },
   "replays": {
    "failure": "claude-sonnet-5-5--CUPYmmk--c1",
    "success": "deepseek-v4.1-flash-ultrafast--J9eCpHM--c0"
   }
  },
  {
   "id": "MOT-E04",
   "title": "Drive and Maintain Shaft Speed",
   "aim": "Drive the rotating axis at the requested speed and maintain that speed.",
   "suite": "motion",
   "difficulty": "easy",
   "family": "motion_easy_unit",
   "conditions": 4,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 1.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "every",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "every"
   },
   "replays": {
    "failure": "paper-gpt-4o--r3--c0",
    "success": "gpt-6-luna--4eV9cGz--c0"
   }
  },
  {
   "id": "MOT-E05",
   "title": "Capture and Record Position on Mark Detection",
   "aim": "Capture and record the position at the moment a mark is detected during movement.",
   "suite": "motion",
   "difficulty": "easy",
   "family": "motion_easy_unit",
   "conditions": 4,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {
    "gpt-6-luna": 1
   },
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "glm-5.3-flash-ultrafast--EuEmdv5--c0",
    "success": "claude-sonnet-5-5--UScozF9--c0"
   }
  },
  {
   "id": "MOT-E06",
   "title": "Vertical Axis Brake Engagement and Height Maintenance",
   "aim": "Engage the brake of the stopped vertical axis and maintain the height.",
   "suite": "motion",
   "difficulty": "easy",
   "family": "motion_easy_unit",
   "conditions": 4,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 1.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "every",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "every"
   },
   "replays": {
    "failure": "paper-gpt-4o--r1--c0",
    "success": "gpt-6.1-sol--ZD2JvHF--c0"
   }
  },
  {
   "id": "MOT-E07",
   "title": "Synchronization of Slave Axis to Master Axis",
   "aim": "Synchronize and maintain the slave axis to the master axis at the requested ratio.",
   "suite": "motion",
   "difficulty": "easy",
   "family": "motion_easy_unit",
   "conditions": 4,
   "passAt1": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 1.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "none",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "every",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "every"
   },
   "replays": {
    "failure": "claude-opus-5-5--bMU72Ci--c0",
    "success": "gpt-6-astra--WxpoNux--c0"
   }
  },
  {
   "id": "MOT-E08",
   "title": "Align and Maintain Following of Moving Reference Axis",
   "aim": "Match the requested interval with the moving master axis and maintain following.",
   "suite": "motion",
   "difficulty": "easy",
   "family": "motion_easy_unit",
   "conditions": 4,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.25,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {
    "gpt-6-luna": 1
   },
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "none",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "none",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "glm-5.3-flash-ultrafast--T4ygvp2--c0",
    "success": "claude-fable-5-1--3XytV8W--c0"
   }
  },
  {
   "id": "MOT-E09",
   "title": "Maintain Tool Orientation During Robot Joint Movement",
   "aim": "Maintain the tool orientation while the robot joints are moving.",
   "suite": "motion",
   "difficulty": "easy",
   "family": "motion_easy_unit",
   "conditions": 4,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {
    "gpt-6-luna": 1
   },
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "every"
   },
   "replays": {
    "failure": "paper-qwen3.5-plus--r2--c0",
    "success": "claude-fable-5-1--rfRebhb--c0"
   }
  },
  {
   "id": "MOT-E10",
   "title": "Rotate and Stop Index Disk at Requested Angle",
   "aim": "Rotate the index disk to the requested angle and stop.",
   "suite": "motion",
   "difficulty": "easy",
   "family": "motion_easy_unit",
   "conditions": 4,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {
    "gpt-6-luna": 1
   },
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "none",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "every"
   },
   "replays": {
    "failure": "claude-sonnet-5-5--bniN9oF--c0",
    "success": "claude-opus-5-5--tJSQT3r--c0"
   }
  },
  {
   "id": "MOT-E11",
   "title": "Door Obstacle Detection and Reopening",
   "aim": "If an approaching object is detected while the door is closing, reopen and maintain it.",
   "suite": "motion",
   "difficulty": "easy",
   "family": "motion_easy_unit",
   "conditions": 4,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "every"
   },
   "replays": {
    "failure": "gpt-6-luna--FNS8vyr--c0",
    "success": "deepseek-v4.1-flash-ultrafast--VHfrWZd--c0"
   }
  },
  {
   "id": "MOT-E12",
   "title": "Boom Retraction to Requested Length",
   "aim": "Retract the extended boom to the requested length and stop.",
   "suite": "motion",
   "difficulty": "easy",
   "family": "motion_easy_unit",
   "conditions": 4,
   "passAt1": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.25,
    "claude-sonnet-5-5": 0.25,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "none",
    "claude-sonnet-5-5": "none",
    "claude-haiku-5-5": "none",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "every"
   },
   "replays": {
    "failure": "claude-opus-5-5--i38PodC--c3",
    "success": "gpt-6.1-sol--25AeQ4S--c0"
   }
  },
  {
   "id": "MOT-E13",
   "title": "Mobile Robot Path Following Task",
   "aim": "Drive the mobile robot along the center of the passage to the requested position and stop.",
   "suite": "motion",
   "difficulty": "easy",
   "family": "motion_easy_unit",
   "conditions": 4,
   "passAt1": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {
    "gpt-6-luna": 1
   },
   "result": {
    "claude-opus-5-5": "none",
    "claude-sonnet-5-5": "none",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "every"
   },
   "replays": {
    "failure": "claude-sonnet-5-5--CfeX3dN--c2",
    "success": "claude-fable-5-1--9zqkiB6--c0"
   }
  },
  {
   "id": "MOT-E14",
   "title": "Hoist Sway Reduction and Trolley Positioning",
   "aim": "Reduce the sway of the hoist and stop the trolley at the current working position.",
   "suite": "motion",
   "difficulty": "easy",
   "family": "motion_easy_unit",
   "conditions": 4,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "none",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "none",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "claude-sonnet-5-5--K6VtZXu--c0",
    "success": "deepseek-v4.1-flash-ultrafast--pk3x9EQ--c0"
   }
  },
  {
   "id": "MOT-E15",
   "title": "Counterbalance Head Tilt",
   "aim": "Move the counterweight to offset the support moment while the head is tilting.",
   "suite": "motion",
   "difficulty": "easy",
   "family": "motion_easy_unit",
   "conditions": 4,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "none",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "none",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "deepseek-v4.1-flash-ultrafast--FJNKRxc--c0",
    "success": "claude-haiku-5-5--Lwk27GG--c0"
   }
  },
  {
   "id": "MOT-H01",
   "title": "MOT-H01",
   "aim": "Process independent requests of the two hoist cells with the shared shuttle and rotary/approach cell, prepare the next task, and return each moving fixture to its original position without collision.",
   "suite": "motion",
   "difficulty": "hard",
   "family": "motion_hard15",
   "conditions": 6,
   "passAt1": {
    "claude-opus-5-5": 0.6667,
    "claude-sonnet-5-5": 0.6667,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.3333,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.3333,
    "claude-sonnet-5-5": 0.3333,
    "claude-haiku-5-5": 0.3333,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0556,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.6667,
    "glm-5.3-flash-ultrafast": 0.8889
   },
   "trials": {
    "claude-opus-5-5": 3,
    "claude-sonnet-5-5": 3,
    "claude-haiku-5-5": 3,
    "gpt-6.1-sol": 3,
    "gpt-6-luna": 3,
    "claude-fable-5-1": 3,
    "gpt-6-astra": 3,
    "deepseek-v4.1-flash-ultrafast": 3,
    "glm-5.3-flash-ultrafast": 3
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "some",
    "claude-sonnet-5-5": "some",
    "claude-haiku-5-5": "none",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "some",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "none",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "claude-opus-5-5--65TbTH4--c0",
    "success": "gpt-6-astra--7ww3yN8--c0"
   }
  },
  {
   "id": "MOT-H02",
   "title": "MOT-H02",
   "aim": "Process multiple alignment requests in sequence with lift, gantry, and reference axis, and maintain the actual target position of each request before and after the reference change.",
   "suite": "motion",
   "difficulty": "hard",
   "family": "motion_hard15",
   "conditions": 3,
   "passAt1": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.3333,
    "claude-haiku-5-5": 0.6667,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 1.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.4444,
    "glm-5.3-flash-ultrafast": 0.3333
   },
   "trials": {
    "claude-opus-5-5": 3,
    "claude-sonnet-5-5": 3,
    "claude-haiku-5-5": 3,
    "gpt-6.1-sol": 3,
    "gpt-6-luna": 3,
    "claude-fable-5-1": 3,
    "gpt-6-astra": 3,
    "deepseek-v4.1-flash-ultrafast": 3,
    "glm-5.3-flash-ultrafast": 3
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "none",
    "claude-sonnet-5-5": "none",
    "claude-haiku-5-5": "none",
    "gpt-6.1-sol": "none",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "none",
    "gpt-6-astra": "none",
    "deepseek-v4.1-flash-ultrafast": "none",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "deepseek-v4.1-flash-ultrafast--NFaxmFK--c0",
    "success": "reference-c0"
   }
  },
  {
   "id": "MOT-H03",
   "title": "MOT-H03",
   "aim": "Process two setup requests in sequence with movement connection and rotation synchronization, and return while observing relative speed and synchronization interval per request.",
   "suite": "motion",
   "difficulty": "hard",
   "family": "motion_hard15",
   "conditions": 3,
   "passAt1": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.6667,
    "claude-haiku-5-5": 0.6667,
    "gpt-6.1-sol": 0.3333,
    "gpt-6-luna": 1.0,
    "claude-fable-5-1": 0.4444,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 3,
    "claude-sonnet-5-5": 3,
    "claude-haiku-5-5": 3,
    "gpt-6.1-sol": 3,
    "gpt-6-luna": 3,
    "claude-fable-5-1": 3,
    "gpt-6-astra": 3,
    "deepseek-v4.1-flash-ultrafast": 3,
    "glm-5.3-flash-ultrafast": 3
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "none",
    "claude-sonnet-5-5": "none",
    "claude-haiku-5-5": "none",
    "gpt-6.1-sol": "none",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "none",
    "gpt-6-astra": "none",
    "deepseek-v4.1-flash-ultrafast": "none",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "gpt-6.1-sol--ci4kcp5--c0",
    "success": "reference-c0"
   }
  },
  {
   "id": "MOT-H04",
   "title": "MOT-H04",
   "aim": "Process independent requests through lift, hoist, and inclined head, and connect to the next facility after settling vibrations.",
   "suite": "motion",
   "difficulty": "hard",
   "family": "motion_hard15",
   "conditions": 3,
   "passAt1": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.3333,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.6667,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.7778,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.7778,
    "glm-5.3-flash-ultrafast": 0.6667
   },
   "trials": {
    "claude-opus-5-5": 3,
    "claude-sonnet-5-5": 3,
    "claude-haiku-5-5": 3,
    "gpt-6.1-sol": 3,
    "gpt-6-luna": 3,
    "claude-fable-5-1": 3,
    "gpt-6-astra": 3,
    "deepseek-v4.1-flash-ultrafast": 3,
    "glm-5.3-flash-ultrafast": 3
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "none",
    "claude-sonnet-5-5": "none",
    "claude-haiku-5-5": "none",
    "gpt-6.1-sol": "some",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "none",
    "gpt-6-astra": "none",
    "deepseek-v4.1-flash-ultrafast": "none",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "deepseek-v4.1-flash-ultrafast--G3KJgqo--c0",
    "success": "gpt-6.1-sol--s8sRGdm--c0"
   }
  },
  {
   "id": "MOT-H05",
   "title": "MOT-H05",
   "aim": "Memorize the new registration values of the two requests respectively to align the subsequent index posture, retrieve the pin, and return the fixture.",
   "suite": "motion",
   "difficulty": "hard",
   "family": "motion_hard15",
   "conditions": 3,
   "passAt1": {
    "claude-opus-5-5": 0.3333,
    "claude-sonnet-5-5": 0.3333,
    "claude-haiku-5-5": 0.3333,
    "gpt-6.1-sol": 0.3333,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.3333,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.3333,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 1.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 3,
    "claude-sonnet-5-5": 3,
    "claude-haiku-5-5": 3,
    "gpt-6.1-sol": 3,
    "gpt-6-luna": 3,
    "claude-fable-5-1": 3,
    "gpt-6-astra": 3,
    "deepseek-v4.1-flash-ultrafast": 3,
    "glm-5.3-flash-ultrafast": 3
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "some",
    "claude-sonnet-5-5": "some",
    "claude-haiku-5-5": "some",
    "gpt-6.1-sol": "some",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "some",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "none",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "deepseek-v4.1-flash-ultrafast--hhr2NHA--c0",
    "success": "claude-opus-5-5--pV6WqPr--c0"
   }
  },
  {
   "id": "MOT-H06",
   "title": "MOT-H06",
   "aim": "Safely retrieve the access mechanism and fixture of a canceled request in progress, and allow the pending request to continue using the same line.",
   "suite": "motion",
   "difficulty": "hard",
   "family": "motion_hard15",
   "conditions": 3,
   "passAt1": {
    "claude-opus-5-5": 0.6667,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.3333,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 0.3333,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.3333,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.3333,
    "claude-fable-5-1": 0.3333,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.3333,
    "glm-5.3-flash-ultrafast": 0.3333
   },
   "trials": {
    "claude-opus-5-5": 3,
    "claude-sonnet-5-5": 3,
    "claude-haiku-5-5": 3,
    "gpt-6.1-sol": 3,
    "gpt-6-luna": 3,
    "claude-fable-5-1": 3,
    "gpt-6-astra": 3,
    "deepseek-v4.1-flash-ultrafast": 3,
    "glm-5.3-flash-ultrafast": 3
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "some",
    "claude-sonnet-5-5": "none",
    "claude-haiku-5-5": "none",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "some",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "some",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "claude-fable-5-1--v4ecsmg--c0",
    "success": "claude-opus-5-5--CV4CCAw--c0"
   }
  },
  {
   "id": "MOT-H07",
   "title": "MOT-H07",
   "aim": "Process two requests in succession with the variable boom and exchange head, avoiding interference during movement according to the boom length and head height.",
   "suite": "motion",
   "difficulty": "hard",
   "family": "motion_hard15",
   "conditions": 3,
   "passAt1": {
    "claude-opus-5-5": 0.3333,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.3333,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 0.6667,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.7778,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.7778,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.7778,
    "glm-5.3-flash-ultrafast": 0.3333
   },
   "trials": {
    "claude-opus-5-5": 3,
    "claude-sonnet-5-5": 3,
    "claude-haiku-5-5": 3,
    "gpt-6.1-sol": 3,
    "gpt-6-luna": 3,
    "claude-fable-5-1": 3,
    "gpt-6-astra": 3,
    "deepseek-v4.1-flash-ultrafast": 3,
    "glm-5.3-flash-ultrafast": 3
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "some",
    "claude-sonnet-5-5": "none",
    "claude-haiku-5-5": "none",
    "gpt-6.1-sol": "some",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "some",
    "deepseek-v4.1-flash-ultrafast": "none",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "deepseek-v4.1-flash-ultrafast--csDwJDQ--c0",
    "success": "claude-opus-5-5--LvywkUH--c0"
   }
  },
  {
   "id": "MOT-H08",
   "title": "MOT-H08",
   "aim": "Process independent requests with the common lift frame and gantry·incline head, and limit the support height difference and total moment.",
   "suite": "motion",
   "difficulty": "hard",
   "family": "motion_hard15",
   "conditions": 3,
   "passAt1": {
    "claude-opus-5-5": 0.6667,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.6667,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.4444,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.6667,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.3333,
    "glm-5.3-flash-ultrafast": 0.7778
   },
   "trials": {
    "claude-opus-5-5": 3,
    "claude-sonnet-5-5": 3,
    "claude-haiku-5-5": 3,
    "gpt-6.1-sol": 3,
    "gpt-6-luna": 3,
    "claude-fable-5-1": 3,
    "gpt-6-astra": 3,
    "deepseek-v4.1-flash-ultrafast": 3,
    "glm-5.3-flash-ultrafast": 3
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "some",
    "claude-sonnet-5-5": "none",
    "claude-haiku-5-5": "none",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "some",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "none",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "claude-haiku-5-5--xkeYaC9--c0",
    "success": "gpt-6-astra--6T6gLxo--c0"
   }
  },
  {
   "id": "MOT-H09",
   "title": "MOT-H09",
   "aim": "Assign compatible pockets to multiple access requests, maintain occupancy before retrieval, and process continuously to the rotation cell.",
   "suite": "motion",
   "difficulty": "hard",
   "family": "motion_hard15",
   "conditions": 3,
   "passAt1": {
    "claude-opus-5-5": 0.6667,
    "claude-sonnet-5-5": 0.3333,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.6667,
    "gpt-6-astra": 0.6667,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.5,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.5556,
    "claude-fable-5-1": 0.3333,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 3,
    "claude-sonnet-5-5": 3,
    "claude-haiku-5-5": 4,
    "gpt-6.1-sol": 3,
    "gpt-6-luna": 3,
    "claude-fable-5-1": 3,
    "gpt-6-astra": 3,
    "deepseek-v4.1-flash-ultrafast": 3,
    "glm-5.3-flash-ultrafast": 3
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "some",
    "claude-sonnet-5-5": "some",
    "claude-haiku-5-5": "none",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "some",
    "gpt-6-astra": "some",
    "deepseek-v4.1-flash-ultrafast": "none",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "claude-fable-5-1--zBoJm8P--c0",
    "success": "gpt-6.1-sol--gaKXTWR--c0"
   }
  },
  {
   "id": "MOT-H10",
   "title": "MOT-H10",
   "aim": "Continue performing the lift·incline·hoist operations of independent requests, and maintain the support of the stationary axis even if other axes move.",
   "suite": "motion",
   "difficulty": "hard",
   "family": "motion_hard15",
   "conditions": 3,
   "passAt1": {
    "claude-opus-5-5": 0.3333,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.3333,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 0.3333,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.6667,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.3333,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.3333,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 3,
    "claude-sonnet-5-5": 3,
    "claude-haiku-5-5": 3,
    "gpt-6.1-sol": 3,
    "gpt-6-luna": 3,
    "claude-fable-5-1": 3,
    "gpt-6-astra": 3,
    "deepseek-v4.1-flash-ultrafast": 3,
    "glm-5.3-flash-ultrafast": 3
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "some",
    "claude-sonnet-5-5": "none",
    "claude-haiku-5-5": "none",
    "gpt-6.1-sol": "some",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "some",
    "deepseek-v4.1-flash-ultrafast": "none",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "deepseek-v4.1-flash-ultrafast--7TNBC9g--c0",
    "success": "claude-fable-5-1--ghR7zkc--c0"
   }
  },
  {
   "id": "MOT-H11",
   "title": "MOT-H11",
   "aim": "Pass each request through the rotation and index cell within the deadline, and assign the shared shuttle so that urgent requests are not trapped behind long tasks.",
   "suite": "motion",
   "difficulty": "hard",
   "family": "motion_hard15",
   "conditions": 3,
   "passAt1": {
    "claude-opus-5-5": 0.6667,
    "claude-sonnet-5-5": 0.3333,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.6667,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.4444,
    "claude-haiku-5-5": 0.7778,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.7778,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 0.3333
   },
   "trials": {
    "claude-opus-5-5": 3,
    "claude-sonnet-5-5": 3,
    "claude-haiku-5-5": 3,
    "gpt-6.1-sol": 3,
    "gpt-6-luna": 3,
    "claude-fable-5-1": 3,
    "gpt-6-astra": 3,
    "deepseek-v4.1-flash-ultrafast": 3,
    "glm-5.3-flash-ultrafast": 3
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "some",
    "claude-sonnet-5-5": "some",
    "claude-haiku-5-5": "none",
    "gpt-6.1-sol": "some",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "none",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "claude-sonnet-5-5--28ryYRY--c1",
    "success": "claude-opus-5-5--2LYBj6L--c0"
   }
  },
  {
   "id": "MOT-H12",
   "title": "MOT-H12",
   "aim": "Process the cross movement and head exchange of independent requests in succession, and do not pass the cross space to other movements before actual retreat.",
   "suite": "motion",
   "difficulty": "hard",
   "family": "motion_hard15",
   "conditions": 3,
   "passAt1": {
    "claude-opus-5-5": 0.6667,
    "claude-sonnet-5-5": 0.3333,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.6667,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.3333,
    "gpt-6-astra": 0.3333,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.6667,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 1.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 0.3333
   },
   "trials": {
    "claude-opus-5-5": 3,
    "claude-sonnet-5-5": 3,
    "claude-haiku-5-5": 3,
    "gpt-6.1-sol": 3,
    "gpt-6-luna": 3,
    "claude-fable-5-1": 3,
    "gpt-6-astra": 3,
    "deepseek-v4.1-flash-ultrafast": 3,
    "glm-5.3-flash-ultrafast": 3
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "some",
    "claude-sonnet-5-5": "some",
    "claude-haiku-5-5": "none",
    "gpt-6.1-sol": "some",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "some",
    "gpt-6-astra": "some",
    "deepseek-v4.1-flash-ultrafast": "none",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "deepseek-v4.1-flash-ultrafast--VttbeAp--c0",
    "success": "gpt-6.1-sol--FzsNsyn--c0"
   }
  },
  {
   "id": "MOT-H13",
   "title": "MOT-H13",
   "aim": "Complete the rotation and synchronization section of each request, and maintain common reference drive while used by other requests.",
   "suite": "motion",
   "difficulty": "hard",
   "family": "motion_hard15",
   "conditions": 3,
   "passAt1": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.7778,
    "claude-fable-5-1": 0.6667,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 0.2222
   },
   "trials": {
    "claude-opus-5-5": 3,
    "claude-sonnet-5-5": 3,
    "claude-haiku-5-5": 3,
    "gpt-6.1-sol": 3,
    "gpt-6-luna": 3,
    "claude-fable-5-1": 3,
    "gpt-6-astra": 3,
    "deepseek-v4.1-flash-ultrafast": 3,
    "glm-5.3-flash-ultrafast": 3
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "none",
    "claude-sonnet-5-5": "none",
    "claude-haiku-5-5": "none",
    "gpt-6.1-sol": "none",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "none",
    "gpt-6-astra": "none",
    "deepseek-v4.1-flash-ultrafast": "none",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "gpt-6-astra--EwcwhNK--c0",
    "success": "reference-c0"
   }
  },
  {
   "id": "MOT-H14",
   "title": "MOT-H14",
   "aim": "Process two position setting requests, and if the common reference changes, discard the previous registered value and stop with the new detected value of that reference.",
   "suite": "motion",
   "difficulty": "hard",
   "family": "motion_hard15",
   "conditions": 3,
   "passAt1": {
    "claude-opus-5-5": 0.3333,
    "claude-sonnet-5-5": 0.3333,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.3333,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.6667,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.3333,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "trials": {
    "claude-opus-5-5": 3,
    "claude-sonnet-5-5": 3,
    "claude-haiku-5-5": 3,
    "gpt-6.1-sol": 3,
    "gpt-6-luna": 3,
    "claude-fable-5-1": 3,
    "gpt-6-astra": 3,
    "deepseek-v4.1-flash-ultrafast": 3,
    "glm-5.3-flash-ultrafast": 3
   },
   "noProgramTrials": {
    "gpt-6-luna": 1
   },
   "result": {
    "claude-opus-5-5": "some",
    "claude-sonnet-5-5": "some",
    "claude-haiku-5-5": "none",
    "gpt-6.1-sol": "none",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "none",
    "gpt-6-astra": "some",
    "deepseek-v4.1-flash-ultrafast": "none",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "glm-5.3-flash-ultrafast--7Ur8bdT--c0",
    "success": "gpt-6-astra--vHvfGKy--c0"
   }
  },
  {
   "id": "MOT-H15",
   "title": "MOT-H15",
   "aim": "Process multiple requests in sequence with a balanced head and compatible pocket, and select a return path that allows the combined posture.",
   "suite": "motion",
   "difficulty": "hard",
   "family": "motion_hard15",
   "conditions": 3,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 0.6667,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.6667,
    "gpt-6-astra": 0.6667,
    "deepseek-v4.1-flash-ultrafast": 0.3333,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.3333,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.5556,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.4444,
    "glm-5.3-flash-ultrafast": 0.6667
   },
   "trials": {
    "claude-opus-5-5": 3,
    "claude-sonnet-5-5": 3,
    "claude-haiku-5-5": 3,
    "gpt-6.1-sol": 3,
    "gpt-6-luna": 3,
    "claude-fable-5-1": 3,
    "gpt-6-astra": 3,
    "deepseek-v4.1-flash-ultrafast": 3,
    "glm-5.3-flash-ultrafast": 3
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "some",
    "claude-haiku-5-5": "none",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "some",
    "gpt-6-astra": "some",
    "deepseek-v4.1-flash-ultrafast": "some",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "deepseek-v4.1-flash-ultrafast--fAL7UZn--c1",
    "success": "claude-sonnet-5-5--hGD3amx--c0"
   }
  },
  {
   "id": "MOT-M01",
   "title": "Swap and Position Heads Without Collision",
   "aim": "Swap the left and right positions of the two heads and stop them at the working height without collision during movement.",
   "suite": "motion",
   "difficulty": "medium",
   "family": "motion_medium_core",
   "conditions": 5,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 1.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "every",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "glm-5.3-flash-ultrafast--2qYUhkZ--c0",
    "success": "gpt-6-astra--Sw3tcNn--c0"
   }
  },
  {
   "id": "MOT-M02",
   "title": "Gantry Alignment·Rail Connection·Empty Cart Transfer",
   "aim": "Connect the beam rails to the dock, transfer the entire empty cart to the fixed rail and park it, then disconnect and return only the empty beam to its origin.",
   "suite": "motion",
   "difficulty": "medium",
   "family": "motion_hard15",
   "conditions": 5,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {
    "gpt-6-luna": 1
   },
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "every"
   },
   "replays": {
    "failure": "paper-claude-sonnet-5--r1--c3",
    "success": "gpt-6.1-sol--WiF9ehW--c0"
   }
  },
  {
   "id": "MOT-M03",
   "title": "Carrier Synchronization with Moving Interface",
   "aim": "Synchronize the carrier with the moving connection, and move together with the empty fork inserted.",
   "suite": "motion",
   "difficulty": "medium",
   "family": "motion_medium_core",
   "conditions": 5,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {
    "gpt-6-luna": 1
   },
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "none",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "claude-sonnet-5-5--KfcpiYM--c0",
    "success": "deepseek-v4.1-flash-ultrafast--Jvgn99V--c0"
   }
  },
  {
   "id": "MOT-M04",
   "title": "Empty Hook Hoisting and Stabilization",
   "aim": "Bring the hoist load to the requested rope length and trolley position, suppress sway, and maintain the stopped result. Safe overlapping motion of the rope and trolley is permitted.",
   "suite": "motion",
   "difficulty": "medium",
   "family": "motion_medium_core",
   "conditions": 5,
   "passAt1": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {
    "gpt-6-luna": 1
   },
   "result": {
    "claude-opus-5-5": "none",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "none",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "claude-haiku-5-5--23Bi99M--c0",
    "success": "deepseek-v4.1-flash-ultrafast--bcjz4qL--c0"
   }
  },
  {
   "id": "MOT-M05",
   "title": "Head Tilt and Brake Engagement Task",
   "aim": "Maintain balance, tilt the head to the requested angle, transfer support to the brake, and deactivate head drive.",
   "suite": "motion",
   "difficulty": "medium",
   "family": "motion_medium_core",
   "conditions": 5,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 1.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "every",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "paper-gpt-5.5--r1--c0",
    "success": "deepseek-v4.1-flash-ultrafast--HtWw4Du--c0"
   }
  },
  {
   "id": "MOT-M06",
   "title": "Vertical opening·Transfer docking·Support lifting·Internal tool reciprocation",
   "aim": "Complete one empty mold access cycle and return all axes to their initial positions.",
   "suite": "motion",
   "difficulty": "medium",
   "family": "motion_hard15",
   "conditions": 5,
   "passAt1": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {
    "gpt-6-luna": 1
   },
   "result": {
    "claude-opus-5-5": "none",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "claude-opus-5-5--YErHCXF--c0",
    "success": "gpt-6.1-sol--3ZgzKYx--c0"
   }
  },
  {
   "id": "MOT-M07",
   "title": "Approach·Retract two work points of the front·rear deployment head",
   "aim": "This is a request to check the operation of an empty device at two access points with one head.",
   "suite": "motion",
   "difficulty": "medium",
   "family": "motion_hard15",
   "conditions": 5,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 1.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "every",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "paper-claude-sonnet-5--r1--c0",
    "success": "claude-haiku-5-5--xquuYw8--c0"
   }
  },
  {
   "id": "MOT-M08",
   "title": "Align the magazine with the current reference coordinates and pitch, then approach laterally",
   "aim": "Register a fresh magazine reference, approach the requested shelf, and retract the approach axis.",
   "suite": "motion",
   "difficulty": "medium",
   "family": "motion_hard15",
   "conditions": 5,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 1.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "every",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "every"
   },
   "replays": {
    "failure": "paper-gemini-3.8-flash--r1--c0",
    "success": "claude-haiku-5-5--ELtv3vs--c0"
   }
  },
  {
   "id": "MOT-M09",
   "title": "Reverse release of guide placement, both-side fixation, and internal access",
   "aim": "Establish and maintain the locked guide posture during internal access, then retract to the origin.",
   "suite": "motion",
   "difficulty": "medium",
   "family": "motion_hard15",
   "conditions": 5,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 1.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "every",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "glm-5.3-flash-ultrafast--f7jecsm--c0",
    "success": "gpt-6-astra--G3CK7sB--c0"
   }
  },
  {
   "id": "MOT-M10",
   "title": "Dual-head transition and reciprocation in avoidance space",
   "aim": "Move the rectangular head through the L-shaped passage to the requested position and return.",
   "suite": "motion",
   "difficulty": "medium",
   "family": "motion_hard15",
   "conditions": 5,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {
    "gpt-6-luna": 1
   },
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "glm-5.3-flash-ultrafast--23wkZ48--c0",
    "success": "claude-sonnet-5-5--bYBmcyT--c0"
   }
  },
  {
   "id": "MOT-M11",
   "title": "Connect the approach and retraction history of different tools by index",
   "aim": "Complete both tool approaches at their requested index angles and depths, then retract both tools.",
   "suite": "motion",
   "difficulty": "medium",
   "family": "motion_hard15",
   "conditions": 5,
   "passAt1": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 1.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "none",
    "claude-sonnet-5-5": "none",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "every",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "claude-opus-5-5--7yTycwv--c0",
    "success": "gpt-6-astra--iHu9u2e--c0"
   }
  },
  {
   "id": "MOT-M12",
   "title": "Two reach positions of serial cylinders and jig approach",
   "aim": "Verify both reach positions and the single jig approach, then return all axes to their origins.",
   "suite": "motion",
   "difficulty": "medium",
   "family": "motion_hard15",
   "conditions": 4,
   "passAt1": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {
    "gpt-6-luna": 1
   },
   "result": {
    "claude-opus-5-5": "none",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "every"
   },
   "replays": {
    "failure": "claude-opus-5-5--MEWoBHq--c0",
    "success": "gpt-6-astra--snPGq4m--c0"
   }
  },
  {
   "id": "MOT-M13",
   "title": "Tilt loader, receiver lift alignment and docking",
   "aim": "This is an empty docking test of one loader and a receiving lift.",
   "suite": "motion",
   "difficulty": "medium",
   "family": "motion_hard15",
   "conditions": 5,
   "passAt1": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 1.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "none",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "every",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "paper-gpt-5.5--r1--c0",
    "success": "gpt-6.1-sol--oFtu3Hi--c0"
   }
  },
  {
   "id": "MOT-M14",
   "title": "Slide Retraction and Table Positioning Task",
   "aim": "Retract the slide to secure rotation space, maintain the retracted state, switch the table to the requested posture, confirm stop, and then re-approach.",
   "suite": "motion",
   "difficulty": "medium",
   "family": "motion_medium_core",
   "conditions": 7,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 1.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "every",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "none",
    "glm-5.3-flash-ultrafast": "every"
   },
   "replays": {
    "failure": "deepseek-v4.1-flash-ultrafast--YSPL3u6--c0",
    "success": "claude-haiku-5-5--An6HwD4--c0"
   }
  },
  {
   "id": "MOT-M15",
   "title": "Directional reference selection, jig transfer, upper tool approach",
   "aim": "Observe the installed empty jig direction side to perform the upper tool approach test.",
   "suite": "motion",
   "difficulty": "medium",
   "family": "motion_hard15",
   "conditions": 5,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "none",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "every"
   },
   "replays": {
    "failure": "claude-fable-5-1--GiWdrAT--c0",
    "success": "deepseek-v4.1-flash-ultrafast--uVbQZbz--c0"
   }
  },
  {
   "id": "MOT-M16",
   "title": "Single module request and completion confirmation maintaining alignment and approach results",
   "aim": "Complete the current empty-tool round trip while maintaining the table/carriage alignment.",
   "suite": "motion",
   "difficulty": "medium",
   "family": "motion_hard15",
   "conditions": 5,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 1.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "every",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "glm-5.3-flash-ultrafast--wixAg6H--c0",
    "success": "gpt-6-luna--JDvzgDA--c0"
   }
  },
  {
   "id": "MOT-M17",
   "title": "Single cycle linking mutual conditions of feeder retreat and head return",
   "aim": "Retract the feeder, complete the head approach at the destination, then restore the required return configuration.",
   "suite": "motion",
   "difficulty": "medium",
   "family": "motion_hard15",
   "conditions": 5,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {
    "gpt-6-luna": 1
   },
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "none",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "claude-haiku-5-5--bnKKWyh--c3",
    "success": "gpt-6-astra--oQttFgu--c0"
   }
  },
  {
   "id": "MOT-M18",
   "title": "Secure search space with actual retraction and acquire new origin",
   "aim": "Retract and raise the internal tool, capture a new reference, and stop at the requested reference-relative position.",
   "suite": "motion",
   "difficulty": "medium",
   "family": "motion_hard15",
   "conditions": 5,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 1.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "every",
    "claude-haiku-5-5": "none",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "every",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "claude-haiku-5-5--72vW7zk--c0",
    "success": "gpt-6-luna--nL8uEEb--c0"
   }
  },
  {
   "id": "MOT-M19",
   "title": "Jig base retreat, lateral alignment, guide retention section transfer",
   "aim": "This is a single guide following section test.",
   "suite": "motion",
   "difficulty": "medium",
   "family": "motion_hard15",
   "conditions": 5,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 1.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 1.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {},
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "none",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "every",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "none"
   },
   "replays": {
    "failure": "claude-sonnet-5-5--5b6hcRB--c0",
    "success": "gpt-6-astra--ARzDzw3--c0"
   }
  },
  {
   "id": "MOT-M20",
   "title": "Synchronized Operation and Parking of Slave Axis",
   "aim": "Synchronize the slave axis to the master axis at the requested ratio, then separate and stop at the requested angle after the specified rotation span.",
   "suite": "motion",
   "difficulty": "medium",
   "family": "motion_medium_core",
   "conditions": 5,
   "passAt1": {
    "claude-opus-5-5": 1.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 1.0,
    "gpt-6.1-sol": 1.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 1.0,
    "gpt-6-astra": 1.0,
    "deepseek-v4.1-flash-ultrafast": 1.0,
    "glm-5.3-flash-ultrafast": 1.0
   },
   "safetyViolationRate": {
    "claude-opus-5-5": 0.0,
    "claude-sonnet-5-5": 0.0,
    "claude-haiku-5-5": 0.0,
    "gpt-6.1-sol": 0.0,
    "gpt-6-luna": 0.0,
    "claude-fable-5-1": 0.0,
    "gpt-6-astra": 0.0,
    "deepseek-v4.1-flash-ultrafast": 0.0,
    "glm-5.3-flash-ultrafast": 0.0
   },
   "trials": {
    "claude-opus-5-5": 1,
    "claude-sonnet-5-5": 1,
    "claude-haiku-5-5": 1,
    "gpt-6.1-sol": 1,
    "gpt-6-luna": 1,
    "claude-fable-5-1": 1,
    "gpt-6-astra": 1,
    "deepseek-v4.1-flash-ultrafast": 1,
    "glm-5.3-flash-ultrafast": 1
   },
   "noProgramTrials": {
    "gpt-6-luna": 1
   },
   "result": {
    "claude-opus-5-5": "every",
    "claude-sonnet-5-5": "none",
    "claude-haiku-5-5": "every",
    "gpt-6.1-sol": "every",
    "gpt-6-luna": "none",
    "claude-fable-5-1": "every",
    "gpt-6-astra": "every",
    "deepseek-v4.1-flash-ultrafast": "every",
    "glm-5.3-flash-ultrafast": "every"
   },
   "replays": {
    "failure": "claude-sonnet-5-5--cEpAPhb--c0",
    "success": "gpt-6.1-sol--dvfgmGG--c0"
   }
  }
 ],
 "provenance": {
  "evaluatorRev": "5.2.6-ef1.2",
  "evaluatorImage": "ef/plcworld-verifier:5.2.6-ef1.2",
  "referenceEvaluator": "plcworld-evaluation/5.2.6",
  "upstream": {
   "name": "PLCWorld",
   "arxiv": "2610.02982",
   "url": "https://arxiv.org/abs/2610.02982",
   "title": "PLCWorld: Benchmarking LLM-Generated PLC Programs in Closed-Loop Plant Simulation",
   "authors": "Yunji Kim, Yunseok Lee, Hyunwoo Seo, Jaerim Choi, Woojin Lee"
  },
  "generatedAt": "2026-10-24T20:03:37+00:00",
  "pendingModels": [],
  "safetyViolationRate": "violating conditions / evaluated conditions (scored trials)",
  "passAt1": "mean reward over a model's trials; reward 1 only if every condition succeeds with no violation",
  "noProgramTrials": "scored trials that ended without /app/program.st (e.g. the program was returned in chat); Easy/Medium (Harbor-graded) score them 0, Hard (offline queue) grades a program recovered from the final message when there is one",
  "audit": {
   "tasksReviewed": 100,
   "tasksClean": 33,
   "tasksWithVerifiedFlaws": 64,
   "tasksChanged": 67,
   "candidates": 130,
   "verified": 105,
   "flaws": {
    "S1": 5,
    "S2": 24,
    "S3": 52,
    "G": 12,
    "H1": 5,
    "H2": 1,
    "H3": 3,
    "H4": 28
   },
   "verifiedFlaws": {
    "S1": 4,
    "S2": 16,
    "S3": 36,
    "G": 12,
    "H1": 5,
    "H2": 1,
    "H3": 3,
    "H4": 28
   },
   "fixes": {
    "prompt": 76,
    "scenario": 29,
    "sharedCode": 11,
    "graderConfig": 3,
    "deferred": 20
   },
   "definitions": "flaws = candidate findings per category; tasksClean = tasks with no finding; tasksChanged = tasks with at least one applied (not deferred) fix"
  },
  "scanMs": 10,
  "conditionsTotal": 476,
  "conditionsUpstream": 473
 },
 "paperBaseline": {
  "source": "PLCWorld paper (arXiv 2610.02982), original package 5.2.6 results",
  "protocol": "single-turn generation, 3 repeats; not comparable with the agent leaderboard (different evaluator revision and harness)",
  "rows": [
   {
    "model": "gpt-5.5",
    "label": "GPT-5.5",
    "repeats": 3,
    "easy": 0.8333,
    "medium": 0.8167,
    "hard": 0.1778,
    "overall": 0.63
   },
   {
    "model": "gemini-3.1-pro",
    "label": "Gemini 3.1 Pro",
    "repeats": 3,
    "easy": 0.8222,
    "medium": 0.775,
    "hard": 0.1667,
    "overall": 0.6067
   },
   {
    "model": "gemini-3.8-flash",
    "label": "Gemini 3.8 Flash",
    "repeats": 3,
    "easy": 0.9556,
    "medium": 0.7333,
    "hard": 0.0556,
    "overall": 0.5967
   },
   {
    "model": "claude-sonnet-5",
    "label": "Claude Sonnet 5",
    "repeats": 3,
    "easy": 0.7333,
    "medium": 0.675,
    "hard": 0.0222,
    "overall": 0.4967
   },
   {
    "model": "qwen3.5-plus",
    "label": "Qwen3.5-Plus",
    "repeats": 3,
    "easy": 0.6778,
    "medium": 0.4083,
    "hard": 0.0,
    "overall": 0.3667
   },
   {
    "model": "gpt-4o",
    "label": "GPT-4o",
    "repeats": 3,
    "easy": 0.1667,
    "medium": 0.0167,
    "hard": 0.0,
    "overall": 0.0567
   }
  ]
 }
}