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 Minify all CSS files. Check the rendered layout across breakpoints and interaction states before proposing a fix.
.claude/skills/thedaviddias-css-minification/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-02 | ✗→✓ | ▲ Improved | -5% | 0% |
| case-03 | ✓→✓ | ＝ Same ✓ | -21% | 0% |
| case-04 | ✓→✓ | ＝ Same ✓ | 8% | 0% |
| case-05 | ✓→✓ | ＝ Same ✓ | -5% | 0% |
| case-06 | ✓→✓ | ＝ Same ✓ | -16% | 0% |
Unminified CSS wastes bandwidth with whitespace, comments, and verbose syntax—minification delivers the same styles in significantly fewer bytes.
Verify that all CSS files are minified in production to reduce file size, eliminate whitespace, and improve page load performance.
Set up CSS minification in your build process using tools like cssnano, CleanCSS, or built-in framework minification to compress CSS files.
Explain how CSS minification improves website performance by reducing file sizes, decreasing bandwidth usage, and speeding up page load times.
Review stylesheets, component styles, and responsive states related to Minify all CSS files. 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-minification
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-02 | fail→pass | 11,621 | 8,855 | -24% | 1 | 1 | 0% | 2,138 | 2,039 | -5% | 0 | 0 | — |
case-01 | fail→fail | 8,470 | 8,404 | -1% | 1 | 1 | 0% | 1,478 | 2,166 | +47% | 0 | 0 | — |
case-03 | pass→pass | 10,270 | 6,398 | -38% | 1 | 1 | 0% | 1,893 | 1,497 | -21% | 0 | 0 | — |
case-04 | pass→pass | 11,112 | 10,364 | -7% | 1 | 1 | 0% | 1,998 | 2,150 | +8% | 0 | 0 | — |
case-05 | pass→pass | 9,727 | 7,423 | -24% | 1 | 1 | 0% | 1,897 | 1,793 | -5% | 0 | 0 | — |
case-06 | pass→pass | 14,136 | 9,919 | -30% | 1 | 1 | 0% | 2,667 | 2,235 | -16% | 0 | 0 | — |
case-07 | pass→pass | 7,441 | 5,879 | -21% | 1 | 1 | 0% | 1,336 | 1,351 | +1% | 0 | 0 | — |
case-08 | pass→pass | 14,944 | 10,945 | -27% | 1 | 1 | 0% | 3,218 | 2,370 | -26% | 0 | 0 | — |
case-09 | pass→pass | 15,332 | 11,855 | -23% | 1 | 1 | 0% | 2,695 | 2,392 | -11% | 0 | 0 | — |
case-10 | pass→pass | 7,454 | 7,434 | -0% | 1 | 1 | 0% | 1,583 | 1,637 | +3% | 0 | 0 | — |
case-11 | pass→pass | 15,556 | 11,603 | -25% | 1 | 1 | 0% | 2,871 | 2,612 | -9% | 0 | 0 | — |
case-12 | pass→pass | 14,608 | 12,008 | -18% | 1 | 1 | 0% | 2,765 | 2,534 | -8% | 0 | 0 | — |
case-13 | pass→pass | 12,694 | 5,314 | -58% | 1 | 1 | 0% | 2,526 | 1,572 | -38% | 0 | 0 | — |
case-14 | pass→pass | 7,047 | 6,036 | -14% | 1 | 1 | 0% | 1,495 | 1,303 | -13% | 0 | 0 | — |
case-15 | pass→pass | 12,026 | 9,672 | -20% | 1 | 1 | 0% | 2,041 | 2,064 | +1% | 0 | 0 | — |
case-16 | pass→pass | 13,952 | 12,746 | -9% | 1 | 1 | 0% | 2,450 | 2,667 | +9% | 0 | 0 | — |
case-17 | pass→pass | 10,677 | 8,197 | -23% | 1 | 1 | 0% | 1,552 | 1,644 | +6% | 0 | 0 | — |
case-18 | pass→pass | 14,472 | 13,267 | -8% | 1 | 1 | 0% | 2,422 | 2,548 | +5% | 0 | 0 | — |
case-19 | pass→pass | 3,473 | 3,366 | -3% | 1 | 1 | 0% | 573 | 861 | +50% | 0 | 0 | — |
case-20 | pass→pass | 14,617 | 15,791 | +8% | 1 | 1 | 0% | 2,780 | 3,456 | +24% | 0 | 0 | — |
case-21 | pass→pass | 11,052 | 8,312 | -25% | 1 | 1 | 0% | 1,995 | 1,835 | -8% | 0 | 0 | — |
case-22 | pass→pass | 10,869 | 7,841 | -28% | 1 | 1 | 0% | 1,989 | 1,807 | -9% | 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.