tao-train-optical-inspection: worth installing? Reported results

Reported results for tao-train-optical-inspection (NVIDIA/skills) from NVIDIA, one row per receipt, in plain words.

Skill skills/tao-train-optical-inspection in NVIDIA/skills.

One row per receipt. No figure on this page combines two rows.

Reported by: Reported by NVIDIA in its benchmark card
Model: Claude Opus 4.8 (aws/anthropic/bedrock-claude-opus-4-8, as the source names it)
Harness: claude-code, version not stated
Date: 2026-09-21
Tasks and runs: 1 task, as printed
Result: NVIDIA reported a score of 99.4% overall and no uplift figure (3 attempts per task). Overall: 99.4% — baseline ran, but no comparable score was available; uplift unavailable Security: 100.0% → 100.0% (±0.0 points) Correctness: 13.3% → 100.0% (+86.7 points) Discoverability: 100.0% — baseline ran, but no comparable score was available; uplift unavailable Effectiveness: 10.0% → 100.0% (+90.0 points) Efficiency: 96.9% — baseline ran, but no comparable score was available; uplift unavailable
Tokens: 71,041 tokens with the skill and 844,751 without, over all cases, as printed; the card gives no cost
Freshness: Reported on Opus 4.8. Registered opus models released after this run, as of the build date: Opus 5.5.
Skill commit: 14a98ae278cec7e867695189d0f4003c08e4f765
Receipt: tao-train-optical-inspection-aws-anthropic-bedrock-claude-opus-4-8-imported-nvidia-cards-claude-code-2026-09-21-a3e62e62298d.json

Reported by: Reported by NVIDIA in its benchmark card
Model: GPT-5.5 (openai/openai/gpt-5.5, as the source names it)
Harness: codex, version not stated
Date: 2026-09-21
Tasks and runs: 1 task, as printed
Result: NVIDIA reported a score of 76.9% overall and no uplift figure (3 attempts per task). Overall: 76.9% — baseline ran, but no comparable score was available; uplift unavailable Security: 100.0% → 100.0% (±0.0 points) Correctness: 100.0% → 70.0% (-30.0 points) Discoverability: 47.5% — baseline ran, but no comparable score was available; uplift unavailable Effectiveness: 31.7% → 68.3% (+36.6 points) Efficiency: 98.7% — baseline ran, but no comparable score was available; uplift unavailable
Tokens: 45,044 tokens with the skill and 14,044 without, over all cases, as printed; the card gives no cost
Freshness: Reported on GPT-5.5. Freshness unknown: the registry has no release date for it.
Skill commit: 14a98ae278cec7e867695189d0f4003c08e4f765
Receipt: tao-train-optical-inspection-openai-openai-gpt-5-5-imported-nvidia-cards-codex-2026-09-21-a3e62e62298d.json

Built on 2026-10-15 from committed receipts.