Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Use when reviewing image assets, markup, and CDN or build transforms related to Support high-DPI retina displays. Check encoded size, rendered size, loading strategy, and above-the-fold impact together.
.claude/skills/thedaviddias-retina-display/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-03 | ✗→✓ | ▲ Improved | -2% | 0% |
| case-09 | ✓→✗ | ▼ Worse | -22% | 0% |
| case-01 | ✓→✓ | ＝ Same ✓ | -23% | 0% |
| case-18 | ✓→✓ | ＝ Same ✓ | -16% | 0% |
| case-19 | ✓→✓ | ＝ Same ✓ | -35% | 0% |
Standard images look blurry on retina displays (2x, 3x pixel density)—high-resolution assets ensure sharp visuals on modern devices.
Check if high-resolution images are provided for retina displays (2x, 3x).
Add retina-ready images using srcset or CSS image-set for high-DPI displays.
Explain how retina images ensure sharp visuals on high-density displays.
Review image assets, markup, and delivery configuration related to Support high-DPI retina displays. Flag exact files or components where format choice, sizing, or loading behavior violates the rule, and describe how to confirm the fix in DevTools.
For full implementation details, code examples, and framework-specific guidance, see references/rule.md.
Rule page: https://frontendchecklist.io/en/rules/images/retina-display
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | pass→pass | 8,725 | 5,943 | -32% | 1 | 1 | 0% | 1,868 | 1,430 | -23% | 0 | 0 | — |
case-18 | pass→pass | 10,597 | 7,567 | -29% | 1 | 1 | 0% | 1,934 | 1,621 | -16% | 0 | 0 | — |
case-19 | pass→pass | 12,822 | 8,009 | -38% | 1 | 1 | 0% | 2,730 | 1,780 | -35% | 0 | 0 | — |
case-02 | pass→pass | 13,114 | 10,033 | -23% | 1 | 1 | 0% | 2,903 | 2,441 | -16% | 0 | 0 | — |
case-03 | fail→pass | 7,845 | 6,500 | -17% | 1 | 1 | 0% | 1,679 | 1,646 | -2% | 0 | 0 | — |
case-04 | pass→pass | 8,696 | 9,090 | +5% | 1 | 1 | 0% | 1,839 | 2,168 | +18% | 0 | 0 | — |
case-05 | pass→pass | 10,523 | 8,853 | -16% | 1 | 1 | 0% | 2,197 | 2,040 | -7% | 0 | 0 | — |
case-06 | pass→pass | 14,779 | 10,158 | -31% | 1 | 1 | 0% | 2,590 | 2,171 | -16% | 0 | 0 | — |
case-17 | pass→pass | 8,632 | 7,162 | -17% | 1 | 1 | 0% | 1,867 | 1,651 | -12% | 0 | 0 | — |
case-07 | pass→pass | 6,722 | 5,587 | -17% | 1 | 1 | 0% | 1,494 | 1,418 | -5% | 0 | 0 | — |
case-08 | pass→pass | 13,642 | 10,676 | -22% | 1 | 1 | 0% | 2,571 | 2,466 | -4% | 0 | 0 | — |
case-09 | pass→fail | 16,409 | 9,595 | -42% | 1 | 1 | 0% | 2,893 | 2,260 | -22% | 0 | 0 | — |
case-10 | pass→pass | 9,741 | 7,251 | -26% | 1 | 1 | 0% | 1,819 | 1,704 | -6% | 0 | 0 | — |
case-11 | pass→pass | 9,722 | 5,875 | -40% | 1 | 1 | 0% | 1,868 | 1,527 | -18% | 0 | 0 | — |
case-12 | pass→pass | 11,368 | 7,508 | -34% | 1 | 1 | 0% | 2,315 | 1,777 | -23% | 0 | 0 | — |
case-13 | pass→pass | 14,726 | 10,452 | -29% | 1 | 1 | 0% | 2,878 | 2,445 | -15% | 0 | 0 | — |
case-14 | pass→pass | 15,080 | 10,909 | -28% | 1 | 1 | 0% | 2,809 | 2,479 | -12% | 0 | 0 | — |
case-15 | pass→pass | 7,711 | 6,098 | -21% | 1 | 1 | 0% | 1,576 | 1,361 | -14% | 0 | 0 | — |
case-16 | pass→pass | 8,579 | 7,565 | -12% | 1 | 1 | 0% | 1,793 | 1,796 | +0% | 0 | 0 | — |
case-20 | pass→pass | 9,376 | 7,996 | -15% | 1 | 1 | 0% | 1,783 | 1,846 | +4% | 0 | 0 | — |
case-21 | pass→pass | 9,263 | 7,429 | -20% | 1 | 1 | 0% | 1,813 | 1,656 | -9% | 0 | 0 | — |
case-22 | pass→pass | 11,446 | 9,334 | -18% | 1 | 1 | 0% | 1,948 | 1,938 | -1% | 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 0 percentage points is the difference between those two pass rates over the 22 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.
Other measured skills in the registry, with their headline benchmark lift.