Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Use when reviewing stylesheets, component styles, and responsive behavior related to Inline critical CSS for faster rendering. Check the rendered layout across breakpoints and interaction states before proposing a fix.
.claude/skills/thedaviddias-css-critical/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 81% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 13% | 0% |
| case-23 | ✓→✗ | ▼ Worse | 0% | 0% |
| case-16 | ✓→✓ | ＝ Same ✓ | -8% | 0% |
| case-21 | ✓→✓ | ＝ Same ✓ | -16% | 0% |
Critical CSS eliminates render-blocking resources, enabling browsers to paint content immediately—directly improving LCP and FCP scores.
Analyze this page's CSS loading strategy to identify critical above-the-fold styles that should be inlined and non-critical styles that can be loaded asynchronously.
Extract and inline critical CSS for above-the-fold content, then load remaining CSS asynchronously to improve initial page render performance.
Explain how critical CSS improves Core Web Vitals by reducing render-blocking resources and enabling faster First Contentful Paint and Largest Contentful Paint.
Review stylesheets, component styles, and responsive states related to Inline critical CSS for faster rendering. Flag exact selectors, declarations, or breakpoints that violate the rule in the rendered UI.
For full implementation details, code examples, and framework-specific guidance, see references/rule.md.
Rule page: https://frontendchecklist.io/en/rules/css/css-critical
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-16 | pass→pass | 13,577 | 11,098 | -18% | 1 | 1 | 0% | 2,503 | 2,307 | -8% | 0 | 0 | — |
case-21 | pass→pass | 15,208 | 10,099 | -34% | 1 | 1 | 0% | 2,675 | 2,243 | -16% | 0 | 0 | — |
case-01 | fail→pass | 9,913 | 13,499 | +36% | 1 | 1 | 0% | 1,757 | 3,185 | +81% | 0 | 0 | — |
case-02 | fail→pass | 11,789 | 13,544 | +15% | 1 | 1 | 0% | 2,127 | 2,404 | +13% | 0 | 0 | — |
case-22 | pass→pass | 11,863 | 7,657 | -35% | 1 | 1 | 0% | 2,330 | 1,739 | -25% | 0 | 0 | — |
case-03 | pass→pass | 13,506 | 8,800 | -35% | 1 | 1 | 0% | 2,362 | 1,711 | -28% | 0 | 0 | — |
case-04 | pass→pass | 9,364 | 7,626 | -19% | 1 | 1 | 0% | 1,963 | 1,730 | -12% | 0 | 0 | — |
case-05 | pass→pass | 11,668 | 10,326 | -12% | 1 | 1 | 0% | 2,341 | 2,173 | -7% | 0 | 0 | — |
case-06 | pass→pass | 13,699 | 10,437 | -24% | 1 | 1 | 0% | 2,524 | 2,298 | -9% | 0 | 0 | — |
case-07 | pass→pass | 8,895 | 6,440 | -28% | 1 | 1 | 0% | 1,608 | 1,410 | -12% | 0 | 0 | — |
case-08 | pass→pass | 11,479 | 6,486 | -43% | 1 | 1 | 0% | 2,009 | 1,505 | -25% | 0 | 0 | — |
case-09 | pass→pass | 13,439 | 10,385 | -23% | 1 | 1 | 0% | 2,700 | 2,365 | -12% | 0 | 0 | — |
case-10 | pass→pass | 7,662 | 6,925 | -10% | 1 | 1 | 0% | 1,695 | 1,727 | +2% | 0 | 0 | — |
case-23 | pass→fail | 16,411 | 13,552 | -17% | 1 | 1 | 0% | 2,505 | 2,499 | -0% | 0 | 0 | — |
case-11 | pass→pass | 15,152 | 13,716 | -9% | 1 | 1 | 0% | 2,774 | 2,802 | +1% | 0 | 0 | — |
case-12 | pass→pass | 7,397 | 4,688 | -37% | 1 | 1 | 0% | 1,552 | 1,337 | -14% | 0 | 0 | — |
case-13 | pass→pass | 4,208 | 5,010 | +19% | 1 | 1 | 0% | 776 | 1,162 | +50% | 0 | 0 | — |
case-14 | pass→pass | 8,298 | 5,339 | -36% | 1 | 1 | 0% | 1,673 | 1,249 | -25% | 0 | 0 | — |
case-15 | pass→pass | 9,242 | 10,691 | +16% | 1 | 1 | 0% | 1,702 | 2,250 | +32% | 0 | 0 | — |
case-17 | pass→pass | 14,666 | 14,465 | -1% | 1 | 1 | 0% | 2,511 | 2,765 | +10% | 0 | 0 | — |
case-18 | pass→pass | 11,045 | 7,611 | -31% | 1 | 1 | 0% | 2,014 | 1,827 | -9% | 0 | 0 | — |
case-19 | pass→pass | 9,362 | 8,210 | -12% | 1 | 1 | 0% | 1,787 | 1,958 | +10% | 0 | 0 | — |
case-20 | pass→pass | 12,753 | 9,227 | -28% | 1 | 1 | 0% | 2,404 | 2,027 | -16% | 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. 23 cases were attempted. The headline lift of +4 percentage points is the difference between those two pass rates over the 23 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.