Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Use when reviewing templates, rendered HTML, or shared components related to Validate HTML against W3C standards. Validate the final browser-facing markup, not just the source framework abstraction.
.claude/skills/thedaviddias-w3c-compliant/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-08 | ✗→✓ | ▲ Improved | 40% | 0% |
| case-02 | ✓→✓ | ＝ Same ✓ | 7% | 0% |
| case-03 | ✓→✓ | ＝ Same ✓ | -23% | 0% |
| case-04 | ✓→✓ | ＝ Same ✓ | -26% | 0% |
| case-05 | ✓→✓ | ＝ Same ✓ | -9% | 0% |
Invalid HTML causes unpredictable rendering across browsers, breaks accessibility tools, and makes debugging significantly harder.
Validate HTML markup using the W3C HTML Validator to identify and fix markup errors that could cause cross-browser compatibility issues.
Run HTML through W3C validator and fix reported errors including unclosed tags, missing attributes, and invalid nesting.
Explain why W3C HTML validation is important for cross-browser compatibility, accessibility, and maintainable code.
Review templates, server-rendered HTML, and shared components that output markup related to Validate HTML against W3C standards. Flag exact elements, attributes, and routes where the rendered HTML violates the rule.
For full implementation details, code examples, and framework-specific guidance, see references/rule.md.
Rule page: https://frontendchecklist.io/en/rules/html/w3c-compliant
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 11,438 | 9,796 | -14% | 1 | 1 | 0% | 1,825 | 1,892 | +4% | 0 | 0 | — |
case-02 | pass→pass | 5,560 | 4,353 | -22% | 1 | 1 | 0% | 1,008 | 1,075 | +7% | 0 | 0 | — |
case-03 | pass→pass | 12,121 | 8,159 | -33% | 1 | 1 | 0% | 2,212 | 1,701 | -23% | 0 | 0 | — |
case-04 | pass→pass | 8,843 | 5,079 | -43% | 1 | 1 | 0% | 1,584 | 1,171 | -26% | 0 | 0 | — |
case-05 | pass→pass | 8,492 | 6,913 | -19% | 1 | 1 | 0% | 1,534 | 1,391 | -9% | 0 | 0 | — |
case-06 | pass→pass | 9,778 | 5,278 | -46% | 1 | 1 | 0% | 1,806 | 1,298 | -28% | 0 | 0 | — |
case-07 | pass→pass | 2,429 | 2,391 | -2% | 1 | 1 | 0% | 420 | 700 | +67% | 0 | 0 | — |
case-08 | fail→pass | 7,243 | 8,890 | +23% | 1 | 1 | 0% | 1,325 | 1,857 | +40% | 0 | 0 | — |
case-09 | pass→pass | 5,321 | 5,022 | -6% | 1 | 1 | 0% | 1,019 | 1,225 | +20% | 0 | 0 | — |
case-10 | pass→pass | 6,104 | 4,172 | -32% | 1 | 1 | 0% | 1,018 | 1,034 | +2% | 0 | 0 | — |
case-11 | pass→pass | 12,201 | 8,424 | -31% | 1 | 1 | 0% | 1,991 | 1,641 | -18% | 0 | 0 | — |
case-12 | pass→pass | 8,145 | 5,242 | -36% | 1 | 1 | 0% | 1,514 | 1,193 | -21% | 0 | 0 | — |
case-13 | pass→pass | 5,223 | 7,344 | +41% | 1 | 1 | 0% | 1,014 | 1,471 | +45% | 0 | 0 | — |
case-14 | pass→pass | 6,572 | 5,958 | -9% | 1 | 1 | 0% | 1,104 | 1,334 | +21% | 0 | 0 | — |
case-15 | pass→pass | 6,737 | 5,181 | -23% | 1 | 1 | 0% | 1,142 | 1,071 | -6% | 0 | 0 | — |
case-16 | pass→pass | 4,107 | 4,239 | +3% | 1 | 1 | 0% | 728 | 1,095 | +50% | 0 | 0 | — |
case-17 | pass→pass | 6,460 | 5,643 | -13% | 1 | 1 | 0% | 1,128 | 1,278 | +13% | 0 | 0 | — |
case-18 | pass→pass | 9,685 | 8,035 | -17% | 1 | 1 | 0% | 1,577 | 1,499 | -5% | 0 | 0 | — |
case-19 | pass→pass | 6,034 | 4,128 | -32% | 1 | 1 | 0% | 1,033 | 994 | -4% | 0 | 0 | — |
case-20 | pass→pass | 6,657 | 7,566 | +14% | 1 | 1 | 0% | 1,301 | 1,578 | +21% | 0 | 0 | — |
case-21 | pass→pass | 6,654 | 4,930 | -26% | 1 | 1 | 0% | 1,470 | 1,362 | -7% | 0 | 0 | — |
case-22 | pass→pass | 5,638 | 3,729 | -34% | 1 | 1 | 0% | 1,104 | 1,088 | -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 +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.