Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Audit design token usage across a product for consistency and coverage.
.claude/skills/owl-listener-design-token-audit/SKILL.md| Model | Eval pass | Runs |
|---|---|---|
| gemini-3.6-flash | 98% | 47 |
| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-06 | ✗→✓ | ▲ Improved | 40% | 0% |
| case-14 | ✗→✓ | ▲ Improved | 33% | 0% |
| case-17 | ✗→✓ | ▲ Improved | 60% | 0% |
| case-19 | ✗→✓ | ▲ Improved | 29% | 0% |
| case-21 | ✓→✓ | ＝ Same ✓ | 17% | 0% |
You are an expert in auditing design token adoption and consistency across products.
You audit how design tokens are used (or not used) in a product, identifying inconsistencies, gaps, and hard-coded values.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-21 | pass→pass | 12,781 | 13,086 | +2% | 1 | 1 | 0% | 2,748 | 3,202 | +17% | 0 | 0 | — |
case-01 | pass→pass | 9,786 | 9,105 | -7% | 1 | 1 | 0% | 1,697 | 2,113 | +25% | 0 | 0 | — |
case-02 | fail→fail | 5,770 | 5,104 | -12% | 1 | 1 | 0% | 920 | 1,218 | +32% | 0 | 0 | — |
case-03 | pass→pass | 6,376 | 7,521 | +18% | 1 | 1 | 0% | 1,184 | 1,671 | +41% | 0 | 0 | — |
case-04 | pass→pass | 9,901 | 11,472 | +16% | 1 | 1 | 0% | 1,628 | 2,160 | +33% | 0 | 0 | — |
case-05 | pass→pass | 11,957 | 13,496 | +13% | 1 | 1 | 0% | 1,891 | 2,767 | +46% | 0 | 0 | — |
case-06 | fail→pass | 10,139 | 11,974 | +18% | 1 | 1 | 0% | 1,715 | 2,403 | +40% | 0 | 0 | — |
case-07 | pass→pass | 13,747 | 12,585 | -8% | 1 | 1 | 0% | 2,507 | 2,342 | -7% | 0 | 0 | — |
case-08 | pass→pass | 11,930 | 8,957 | -25% | 1 | 1 | 0% | 1,920 | 1,885 | -2% | 0 | 0 | — |
case-09 | pass→pass | 9,970 | 11,318 | +14% | 1 | 1 | 0% | 1,798 | 2,407 | +34% | 0 | 0 | — |
case-10 | pass→pass | 13,007 | 13,731 | +6% | 1 | 1 | 0% | 1,955 | 2,552 | +31% | 0 | 0 | — |
case-11 | pass→pass | 8,457 | 8,276 | -2% | 1 | 1 | 0% | 1,457 | 1,774 | +22% | 0 | 0 | — |
case-12 | pass→pass | 10,460 | 9,938 | -5% | 1 | 1 | 0% | 1,746 | 2,171 | +24% | 0 | 0 | — |
case-13 | pass→pass | 12,869 | 9,974 | -22% | 1 | 1 | 0% | 2,009 | 1,996 | -1% | 0 | 0 | — |
case-14 | fail→pass | 13,007 | 13,946 | +7% | 1 | 1 | 0% | 1,993 | 2,660 | +33% | 0 | 0 | — |
case-15 | pass→pass | 13,951 | 14,011 | +0% | 1 | 1 | 0% | 2,583 | 2,715 | +5% | 0 | 0 | — |
case-16 | pass→pass | 5,388 | 7,400 | +37% | 1 | 1 | 0% | 1,012 | 1,744 | +72% | 0 | 0 | — |
case-17 | fail→pass | 5,702 | 5,825 | +2% | 1 | 1 | 0% | 965 | 1,541 | +60% | 0 | 0 | — |
case-18 | pass→pass | 8,623 | 6,109 | -29% | 1 | 1 | 0% | 1,412 | 1,516 | +7% | 0 | 0 | — |
case-19 | fail→pass | 8,203 | 8,012 | -2% | 1 | 1 | 0% | 1,331 | 1,716 | +29% | 0 | 0 | — |
case-20 | pass→pass | 17,829 | 13,616 | -24% | 1 | 1 | 0% | 4,043 | 3,321 | -18% | 0 | 0 | — |
case-22 | pass→pass | 15,507 | 16,499 | +6% | 1 | 1 | 0% | 3,244 | 3,603 | +11% | 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. The headline lift of +18 percentage points is the difference between those two pass rates over the 22 comparable cases.
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.
Other measured skills in the registry, with their headline benchmark lift.