Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Establish clear visual hierarchy through size, weight, color, spacing, and positioning.
.claude/skills/owl-listener-visual-hierarchy/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -1% | 0% |
| case-04 | ✗→✓ | ▲ Improved | 8% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 18% | 0% |
| case-21 | ✓→✗ | ▼ Worse | 59% | 0% |
| case-20 | ✓→✗ | ▼ Worse | -13% | 0% |
You are an expert in creating clear visual hierarchy that guides users through interfaces.
You establish visual hierarchy ensuring users see the most important content first and can scan efficiently.
Larger elements draw attention first. Use size differences of at least 1.5x for clear distinction.
Bold text, thicker strokes, and filled icons carry more visual weight than light variants.
High contrast attracts attention. Use color strategically for CTAs, status, and emphasis.
More whitespace around an element increases its perceived importance.
Top-left (in LTR layouts) gets seen first. Above the fold matters. F-pattern and Z-pattern scanning.
Isolated elements stand out. Grouped elements are scanned as a unit.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 14,720 | 12,058 | -18% | 1 | 1 | 0% | 2,735 | 2,696 | -1% | 0 | 0 | — |
case-02 | pass→pass | 13,664 | 12,594 | -8% | 1 | 1 | 0% | 2,282 | 2,367 | +4% | 0 | 0 | — |
case-03 | pass→pass | 12,943 | 11,797 | -9% | 1 | 1 | 0% | 2,181 | 2,345 | +8% | 0 | 0 | — |
case-04 | fail→pass | 11,731 | 12,435 | +6% | 1 | 1 | 0% | 2,220 | 2,403 | +8% | 0 | 0 | — |
case-21 | pass→fail | 6,454 | 8,282 | +28% | 1 | 1 | 0% | 1,193 | 1,897 | +59% | 0 | 0 | — |
case-05 | fail→pass | 15,941 | 17,609 | +10% | 1 | 1 | 0% | 3,082 | 3,652 | +18% | 0 | 0 | — |
case-06 | pass→pass | 10,313 | 8,758 | -15% | 1 | 1 | 0% | 1,684 | 1,577 | -6% | 0 | 0 | — |
case-07 | pass→pass | 8,845 | 9,084 | +3% | 1 | 1 | 0% | 1,692 | 2,019 | +19% | 0 | 0 | — |
case-08 | pass→pass | 15,476 | 12,565 | -19% | 1 | 1 | 0% | 2,782 | 2,410 | -13% | 0 | 0 | — |
case-09 | pass→pass | 16,941 | 15,441 | -9% | 1 | 1 | 0% | 2,955 | 2,803 | -5% | 0 | 0 | — |
case-10 | pass→pass | 13,042 | 13,031 | -0% | 1 | 1 | 0% | 2,379 | 2,553 | +7% | 0 | 0 | — |
case-11 | pass→pass | 7,555 | 8,247 | +9% | 1 | 1 | 0% | 1,193 | 1,776 | +49% | 0 | 0 | — |
case-12 | pass→pass | 11,690 | 11,069 | -5% | 1 | 1 | 0% | 2,180 | 2,434 | +12% | 0 | 0 | — |
case-13 | pass→pass | 16,531 | 15,205 | -8% | 1 | 1 | 0% | 2,655 | 2,703 | +2% | 0 | 0 | — |
case-14 | pass→pass | 14,365 | 15,817 | +10% | 1 | 1 | 0% | 2,567 | 3,075 | +20% | 0 | 0 | — |
case-15 | pass→pass | 12,459 | 12,110 | -3% | 1 | 1 | 0% | 2,083 | 2,495 | +20% | 0 | 0 | — |
case-16 | fail→fail | 15,418 | 12,308 | -20% | 1 | 1 | 0% | 2,570 | 2,642 | +3% | 0 | 0 | — |
case-17 | pass→pass | 9,605 | 5,844 | -39% | 1 | 1 | 0% | 1,686 | 1,315 | -22% | 0 | 0 | — |
case-18 | pass→pass | 17,227 | 15,852 | -8% | 1 | 1 | 0% | 2,862 | 3,038 | +6% | 0 | 0 | — |
case-19 | pass→pass | 13,037 | 9,873 | -24% | 1 | 1 | 0% | 2,060 | 1,990 | -3% | 0 | 0 | — |
case-20 | pass→fail | 17,310 | 13,562 | -22% | 1 | 1 | 0% | 3,840 | 3,349 | -13% | 0 | 0 | — |
case-22 | pass→fail | 8,888 | 7,010 | -21% | 1 | 1 | 0% | 1,916 | 1,756 | -8% | 0 | 0 | — |
case-23 | pass→pass | 16,719 | 19,684 | +18% | 1 | 1 | 0% | 3,359 | 4,101 | +22% | 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. 23 cases were attempted. The headline lift of 0 percentage points is the difference between those two pass rates over the 23 comparable cases. 3 cases got worse with the skill loaded, and they are 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.
Other measured skills in the registry, with their headline benchmark lift.