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 Use logical heading hierarchy. Check native semantics first, then inspect keyboard behavior, focus flow, accessible names, and screen-reader output where relevant.
.claude/skills/thedaviddias-heading-hierarchy/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-05 | ✗→✓ | ▲ Improved | -2% | 0% |
| case-20 | ✗→✓ | ▲ Improved | -13% | 0% |
| case-01 | ✓→✓ | = Same ✓ | 6% | 0% |
| case-02 | ✓→✓ | = Same ✓ | 1% | 0% |
| case-03 | ✓→✓ | = Same ✓ | -21% | 0% |
Screen reader users navigate by headings like a table of contents—skipping levels or poor hierarchy makes content impossible to understand and navigate.
Verify pages have exactly one h1, headings follow sequential order without skipping levels (h1 → h2 → h3), and heading text clearly describes the content that follows. Use accessibility tools to generate a heading outline.
Structure headings to reflect content hierarchy. Start with h1 for the page title, use h2 for major sections, h3 for subsections. Never skip heading levels or use headings purely for visual styling.
Explain how screen reader users navigate by headings like a table of contents, and why skipped levels or missing structure forces users to work harder to understand page organization.
Review the rendered markup and interactive states that affect Use logical heading hierarchy. 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/heading-hierarchy
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | pass→pass | 9,262 | 8,952 | -3% | 1 | 1 | 0% | 1,836 | 1,945 | +6% | 0 | 0 | — |
case-02 | pass→pass | 8,402 | 5,848 | -30% | 1 | 1 | 0% | 1,516 | 1,536 | +1% | 0 | 0 | — |
case-03 | pass→pass | 6,451 | 3,357 | -48% | 1 | 1 | 0% | 1,342 | 1,060 | -21% | 0 | 0 | — |
case-04 | pass→pass | 11,214 | 6,368 | -43% | 1 | 1 | 0% | 2,021 | 1,578 | -22% | 0 | 0 | — |
case-05 | fail→pass | 6,519 | 4,749 | -27% | 1 | 1 | 0% | 1,298 | 1,278 | -2% | 0 | 0 | — |
case-06 | pass→pass | 6,094 | 5,925 | -3% | 1 | 1 | 0% | 1,219 | 1,364 | +12% | 0 | 0 | — |
case-07 | pass→pass | 5,818 | 4,524 | -22% | 1 | 1 | 0% | 1,277 | 1,173 | -8% | 0 | 0 | — |
case-08 | fail→fail | 11,313 | 7,134 | -37% | 1 | 1 | 0% | 2,190 | 1,753 | -20% | 0 | 0 | — |
case-09 | pass→pass | 10,333 | 12,736 | +23% | 1 | 1 | 0% | 1,987 | 2,664 | +34% | 0 | 0 | — |
case-10 | pass→pass | 11,311 | 10,464 | -7% | 1 | 1 | 0% | 2,417 | 2,291 | -5% | 0 | 0 | — |
case-11 | pass→pass | 6,356 | 4,333 | -32% | 1 | 1 | 0% | 1,080 | 1,188 | +10% | 0 | 0 | — |
case-12 | pass→pass | 9,281 | 4,622 | -50% | 1 | 1 | 0% | 1,813 | 1,160 | -36% | 0 | 0 | — |
case-13 | pass→pass | 5,975 | 6,051 | +1% | 1 | 1 | 0% | 1,268 | 1,385 | +9% | 0 | 0 | — |
case-14 | pass→pass | 9,906 | 7,802 | -21% | 1 | 1 | 0% | 2,162 | 1,852 | -14% | 0 | 0 | — |
case-15 | pass→pass | 7,397 | 4,533 | -39% | 1 | 1 | 0% | 1,295 | 1,211 | -6% | 0 | 0 | — |
case-16 | pass→pass | 11,755 | 12,983 | +10% | 1 | 1 | 0% | 2,543 | 2,818 | +11% | 0 | 0 | — |
case-17 | pass→pass | 5,736 | 4,218 | -26% | 1 | 1 | 0% | 1,221 | 1,068 | -13% | 0 | 0 | — |
case-18 | pass→pass | 6,406 | 4,893 | -24% | 1 | 1 | 0% | 1,330 | 1,233 | -7% | 0 | 0 | — |
case-19 | pass→pass | 9,526 | 8,711 | -9% | 1 | 1 | 0% | 1,696 | 2,028 | +20% | 0 | 0 | — |
case-20 | fail→pass | 11,055 | 8,348 | -24% | 1 | 1 | 0% | 2,090 | 1,815 | -13% | 0 | 0 | — |
case-21 | pass→pass | 6,686 | 5,660 | -15% | 1 | 1 | 0% | 1,401 | 1,328 | -5% | 0 | 0 | — |
case-22 | pass→pass | 9,354 | 5,905 | -37% | 1 | 1 | 0% | 1,516 | 1,361 | -10% | 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.