Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Apply the Von Restorff Effect to make the most important element distinctly different from its surroundings.
.claude/skills/owl-listener-von-restorff-effect/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-04 | ✗→✓ | ▲ Improved | 17% | 0% |
| case-17 | ✗→✓ | ▲ Improved | 10% | 0% |
| case-01 | ✓→✓ | ＝ Same ✓ | 10% | 0% |
| case-02 | ✓→✓ | ＝ Same ✓ | -6% | 0% |
| case-03 | ✓→✓ | ＝ Same ✓ | 2% | 0% |
You are an expert in visual differentiation and its effect on memory and attention.
You apply the Von Restorff Effect (also called the Isolation Effect) to ensure the one element that most needs attention is visually distinct — and that distinctiveness is earned, not scattered.
An item that differs from its surroundings is more likely to be noticed and remembered. Visual homogeneity is the baseline; deviation draws the eye. This is why:
The effect depends on contrast with context. If everything is differentiated, nothing is. The principle only works when:
| Context | How to Apply | |---|---| | Call to action | One filled/primary button; all others ghost or text | | Pricing | Highlight one recommended tier; reduce visual weight of others | | Navigation | Active state distinctly different from inactive | | Data tables | Use row highlight or bold type for the key record | | Notifications | Badge or accent color reserved for actionable items only | | Onboarding | One step or card at a time, visually isolated from upcoming steps |
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | pass→pass | 12,924 | 11,916 | -8% | 1 | 1 | 0% | 2,175 | 2,397 | +10% | 0 | 0 | — |
case-02 | pass→pass | 14,283 | 10,091 | -29% | 1 | 1 | 0% | 2,369 | 2,235 | -6% | 0 | 0 | — |
case-03 | pass→pass | 11,445 | 8,614 | -25% | 1 | 1 | 0% | 1,840 | 1,877 | +2% | 0 | 0 | — |
case-04 | fail→pass | 10,817 | 12,185 | +13% | 1 | 1 | 0% | 2,034 | 2,376 | +17% | 0 | 0 | — |
case-05 | pass→pass | 13,315 | 8,887 | -33% | 1 | 1 | 0% | 2,147 | 1,772 | -17% | 0 | 0 | — |
case-06 | pass→pass | 12,470 | 12,086 | -3% | 1 | 1 | 0% | 2,010 | 2,295 | +14% | 0 | 0 | — |
case-07 | pass→pass | 11,697 | 8,956 | -23% | 1 | 1 | 0% | 1,926 | 1,937 | +1% | 0 | 0 | — |
case-08 | pass→pass | 12,015 | 9,671 | -20% | 1 | 1 | 0% | 1,899 | 2,240 | +18% | 0 | 0 | — |
case-09 | pass→pass | 11,227 | 8,029 | -28% | 1 | 1 | 0% | 2,093 | 1,833 | -12% | 0 | 0 | — |
case-10 | pass→pass | 11,346 | 9,467 | -17% | 1 | 1 | 0% | 2,113 | 2,142 | +1% | 0 | 0 | — |
case-11 | pass→pass | 11,717 | 9,809 | -16% | 1 | 1 | 0% | 2,019 | 2,212 | +10% | 0 | 0 | — |
case-12 | pass→pass | 12,254 | 10,714 | -13% | 1 | 1 | 0% | 1,836 | 2,114 | +15% | 0 | 0 | — |
case-13 | pass→pass | 12,012 | 8,595 | -28% | 1 | 1 | 0% | 2,067 | 1,790 | -13% | 0 | 0 | — |
case-14 | pass→pass | 11,194 | 12,779 | +14% | 1 | 1 | 0% | 1,820 | 2,609 | +43% | 0 | 0 | — |
case-15 | pass→pass | 11,483 | 10,081 | -12% | 1 | 1 | 0% | 2,077 | 2,007 | -3% | 0 | 0 | — |
case-16 | pass→pass | 12,246 | 8,655 | -29% | 1 | 1 | 0% | 2,013 | 1,726 | -14% | 0 | 0 | — |
case-17 | fail→pass | 13,415 | 11,991 | -11% | 1 | 1 | 0% | 2,346 | 2,584 | +10% | 0 | 0 | — |
case-18 | pass→pass | 9,281 | 9,122 | -2% | 1 | 1 | 0% | 1,557 | 1,732 | +11% | 0 | 0 | — |
case-19 | pass→pass | 10,663 | 10,078 | -5% | 1 | 1 | 0% | 1,769 | 1,954 | +10% | 0 | 0 | — |
case-20 | pass→pass | 12,250 | 11,193 | -9% | 1 | 1 | 0% | 1,968 | 2,244 | +14% | 0 | 0 | — |
case-21 | pass→pass | 11,823 | 9,464 | -20% | 1 | 1 | 0% | 1,906 | 1,874 | -2% | 0 | 0 | — |
case-22 | pass→pass | 12,479 | 9,406 | -25% | 1 | 1 | 0% | 1,921 | 1,773 | -8% | 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 +9 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.