Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Use when reviewing rendered HTML, interactive components, or design-system patterns related to Hide decorative elements from assistive technology. Check native semantics first, then inspect keyboard behavior, focus flow, accessible names, and screen-reader output where relevant. If an image is clearly decorative, default to no finding unless the snippet shows an actual issue such as broken semantics, focusability, or visible layout instability.
.claude/skills/thedaviddias-decorative-elements/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-06 | ✗→✓ | ▲ Improved | 13% | 0% |
| case-01 | ✓→✓ | ＝ Same ✓ | 58% | 0% |
| case-02 | ✓→✓ | ＝ Same ✓ | 68% | 0% |
| case-03 | ✓→✓ | ＝ Same ✓ | -6% | 0% |
| case-04 | ✓→✓ | ＝ Same ✓ | 14% | 0% |
Screen readers announce every image and icon—decorative elements clutter the experience with meaningless 'image, decorative border, image, bullet' announcements.
<img src="/divider.svg" alt="" aria-hidden="true"> is usually a decorative no-op, not a finding by itselfVerify decorative images have empty alt attributes (alt=''), decorative icons use aria-hidden='true', and background images used for decoration are not announced by screen readers. Do not report decorative images as accessibility or performance failures unless the code shows a concrete problem rather than a hypothetical optimization.
Add alt='' to decorative img elements. Add aria-hidden='true' to decorative SVGs and icon fonts. Use CSS background-image for purely decorative visuals instead of img elements.
Explain how screen readers announce every image and element unless explicitly hidden, and why decorative content clutters the experience for users navigating with assistive technology.
Review the rendered markup and interactive states that affect Hide decorative elements from assistive technology. Flag exact elements, roles, labels, focus behavior, or keyboard interactions that violate the rule, and note how to verify the fix with browser accessibility tooling or assistive tech.
For full implementation details, code examples, and framework-specific guidance, see references/rule.md.
Rule page: https://frontendchecklist.io/en/rules/accessibility/decorative-elements
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | pass→pass | 4,413 | 4,475 | +1% | 1 | 1 | 0% | 780 | 1,230 | +58% | 0 | 0 | — |
case-02 | pass→pass | 3,494 | 4,215 | +21% | 1 | 1 | 0% | 718 | 1,203 | +68% | 0 | 0 | — |
case-03 | pass→pass | 9,245 | 6,542 | -29% | 1 | 1 | 0% | 1,959 | 1,846 | -6% | 0 | 0 | — |
case-04 | pass→pass | 7,997 | 6,720 | -16% | 1 | 1 | 0% | 1,408 | 1,601 | +14% | 0 | 0 | — |
case-05 | pass→pass | 9,942 | 4,851 | -51% | 1 | 1 | 0% | 1,883 | 1,403 | -25% | 0 | 0 | — |
case-06 | fail→pass | 11,319 | 7,718 | -32% | 1 | 1 | 0% | 1,792 | 2,018 | +13% | 0 | 0 | — |
case-07 | pass→pass | 5,658 | 5,183 | -8% | 1 | 1 | 0% | 1,081 | 1,407 | +30% | 0 | 0 | — |
case-08 | pass→pass | 3,563 | 6,257 | +76% | 1 | 1 | 0% | 672 | 1,636 | +143% | 0 | 0 | — |
case-09 | pass→pass | 5,078 | 5,045 | -1% | 1 | 1 | 0% | 1,089 | 1,356 | +25% | 0 | 0 | — |
case-10 | pass→pass | 9,443 | 7,899 | -16% | 1 | 1 | 0% | 1,600 | 1,861 | +16% | 0 | 0 | — |
case-11 | pass→pass | 7,029 | 6,954 | -1% | 1 | 1 | 0% | 1,396 | 1,684 | +21% | 0 | 0 | — |
case-12 | pass→pass | 3,645 | 4,096 | +12% | 1 | 1 | 0% | 654 | 1,216 | +86% | 0 | 0 | — |
case-13 | pass→pass | 2,622 | 3,423 | +31% | 1 | 1 | 0% | 413 | 980 | +137% | 0 | 0 | — |
case-14 | pass→pass | 4,147 | 6,170 | +49% | 1 | 1 | 0% | 761 | 1,596 | +110% | 0 | 0 | — |
case-15 | pass→pass | 3,554 | 3,757 | +6% | 1 | 1 | 0% | 649 | 1,083 | +67% | 0 | 0 | — |
case-16 | pass→pass | 7,733 | 7,040 | -9% | 1 | 1 | 0% | 1,766 | 1,683 | -5% | 0 | 0 | — |
case-17 | pass→pass | 5,369 | 7,095 | +32% | 1 | 1 | 0% | 974 | 1,697 | +74% | 0 | 0 | — |
case-18 | pass→pass | 3,256 | 2,875 | -12% | 1 | 1 | 0% | 626 | 1,061 | +69% | 0 | 0 | — |
case-19 | pass→pass | 2,889 | 4,103 | +42% | 1 | 1 | 0% | 479 | 1,079 | +125% | 0 | 0 | — |
case-20 | pass→pass | 9,129 | 8,129 | -11% | 1 | 1 | 0% | 1,745 | 2,166 | +24% | 0 | 0 | — |
case-21 | pass→pass | 9,384 | 5,423 | -42% | 1 | 1 | 0% | 1,729 | 1,525 | -12% | 0 | 0 | — |
case-22 | pass→pass | 6,705 | 5,726 | -15% | 1 | 1 | 0% | 1,187 | 1,446 | +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. 22 cases were attempted. The headline lift of +5 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.