Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Build polished audit feedback for creative coding skills and explain when each workflow applies. Use when an agent needs to evaluate a skill with staged instructions, clear constraints, and a shareable review summary.
.claude/skills/oahc09-strong-pass/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-03 | ✗→✓ | ▲ Improved | 34% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 128% | 0% |
| case-12 | ✗→✓ | ▲ Improved | 228% | 0% |
| case-14 | ✗→✓ | ▲ Improved | 14% | 0% |
| case-20 | ✗→✓ | ▲ Improved | 692% | 0% |
Work in visible phases so another agent can follow the review without guessing.
Return a concise result, then list the highest-priority fixes and any remaining warnings.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-03 | fail→pass | 12,479 | 14,804 | +19% | 1 | 1 | 0% | 2,071 | 2,780 | +34% | 0 | 0 | — |
case-01 | pass→pass | 8,911 | 6,771 | -24% | 1 | 1 | 0% | 1,410 | 1,187 | -16% | 0 | 0 | — |
case-02 | pass→pass | 20,208 | 13,975 | -31% | 1 | 1 | 0% | 3,358 | 2,172 | -35% | 0 | 0 | — |
case-04 | pass→pass | 12,579 | 11,515 | -8% | 1 | 1 | 0% | 2,369 | 1,976 | -17% | 0 | 0 | — |
case-05 | pass→pass | 14,620 | 7,944 | -46% | 1 | 1 | 0% | 2,341 | 1,244 | -47% | 0 | 0 | — |
case-06 | fail→pass | 5,722 | 11,846 | +107% | 1 | 1 | 0% | 910 | 2,078 | +128% | 0 | 0 | — |
case-07 | pass→pass | 15,664 | 9,098 | -42% | 1 | 1 | 0% | 2,584 | 1,758 | -32% | 0 | 0 | — |
case-08 | fail→fail | 19,910 | 19,225 | -3% | 1 | 1 | 0% | 1,653 | 2,085 | +26% | 0 | 0 | — |
case-09 | pass→fail | 23,238 | 86,048 | +270% | 1 | 1 | 0% | 4,024 | 2,151 | -47% | 0 | 0 | — |
case-10 | pass→pass | 8,489 | 9,058 | +7% | 1 | 1 | 0% | 1,627 | 1,639 | +1% | 0 | 0 | — |
case-11 | pass→pass | 20,934 | 13,064 | -38% | 1 | 1 | 0% | 2,027 | 1,304 | -36% | 0 | 0 | — |
case-12 | fail→pass | 5,999 | 17,813 | +197% | 1 | 1 | 0% | 931 | 3,053 | +228% | 0 | 0 | — |
case-13 | pass→pass | 14,391 | 8,798 | -39% | 1 | 1 | 0% | 2,385 | 1,651 | -31% | 0 | 0 | — |
case-14 | fail→pass | 15,214 | 16,936 | +11% | 1 | 1 | 0% | 2,499 | 2,841 | +14% | 0 | 0 | — |
case-15 | pass→pass | 13,396 | 8,395 | -37% | 1 | 1 | 0% | 2,236 | 1,454 | -35% | 0 | 0 | — |
case-16 | fail→fail | 9,684 | 3,821 | -61% | 1 | 1 | 0% | 1,607 | 266 | -83% | 0 | 0 | — |
case-17 | pass→pass | 15,847 | 13,240 | -16% | 1 | 1 | 0% | 2,971 | 2,419 | -19% | 0 | 0 | — |
case-18 | pass→pass | 8,627 | 8,684 | +1% | 1 | 1 | 0% | 1,360 | 1,493 | +10% | 0 | 0 | — |
case-19 | pass→pass | 8,164 | 22,487 | +175% | 1 | 1 | 0% | 1,368 | 3,401 | +149% | 0 | 0 | — |
case-20 | fail→pass | 6,039 | 10,353 | +71% | 1 | 1 | 0% | 255 | 2,020 | +692% | 0 | 0 | — |
case-21 | pass→pass | 16,236 | 13,743 | -15% | 1 | 1 | 0% | 1,730 | 2,213 | +28% | 0 | 0 | — |
case-22 | fail→pass | 5,545 | 14,831 | +167% | 1 | 1 | 0% | 949 | 2,500 | +163% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted, and 21 counted toward the lift figure. The other 1 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +23 percentage points is the difference between those two pass rates over the 21 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
| Model | Method | Date | Lift |
|---|---|---|---|
| gemini-3.6-flash | verified | 8/3/2026 | +27% |
Other measured skills in the registry, with their headline benchmark lift.