Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Comprehensive web quality audit covering performance, accessibility, SEO, and best practices. Use when asked to "audit my site", "review web quality", "run lighthouse audit", "check page quality", or "optimize my website".
.claude/skills/mkurman-web-quality-audit/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -45% | 0% |
| case-02 | ✗→✓ | ▲ Improved | -23% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -39% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -40% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -32% | 0% |
-|-------------|--------| | Critical | Security vulnerabilities, complete failures | Fix immediately | | High | Core Web Vitals failures, major a11y barriers | Fix before launch | | Medium | Performance opportunities, SEO improvements | Fix within sprint | | Low | Minor optimizations, code quality | Fix when convenient |
When performing an audit, structure findings as:
markdown## Audit results ### Critical issues (X found) - **[Category]** Issue description. File: `path/to/file.js:123` - **Impact:** Why this matters - **Fix:** Specific code change or recommendation ### High priority (X found) ... ### Summary - Performance: X issues (Y critical) - Accessibility: X issues (Y critical) - SEO: X issues - Best Practices: X issues ### Recommended priority 1. First fix this because... 2. Then address... 3. Finally optimize...
For detailed guidelines on specific areas:
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 34,510 | 23,774 | -31% | 1 | 1 | 0% | 6,145 | 3,372 | -45% | 0 | 0 | — |
case-02 | fail→pass | 91,843 | 34,945 | -62% | 1 | 1 | 0% | 6,210 | 4,770 | -23% | 0 | 0 | — |
case-03 | fail→pass | 30,634 | 19,606 | -36% | 1 | 1 | 0% | 6,043 | 3,683 | -39% | 0 | 0 | — |
case-04 | fail→pass | 10,962 | 4,101 | -63% | 1 | 1 | 0% | 1,787 | 1,074 | -40% | 0 | 0 | — |
case-05 | fail→pass | 11,994 | 5,370 | -55% | 1 | 1 | 0% | 1,918 | 1,303 | -32% | 0 | 0 | — |
case-06 | fail→pass | 11,169 | 5,032 | -55% | 1 | 1 | 0% | 1,614 | 1,218 | -25% | 0 | 0 | — |
case-07 | pass→pass | 11,767 | 4,879 | -59% | 1 | 1 | 0% | 1,757 | 1,129 | -36% | 0 | 0 | — |
case-08 | fail→fail | 17,953 | 20,296 | +13% | 1 | 1 | 0% | 2,789 | 2,125 | -24% | 0 | 0 | — |
case-09 | pass→pass | 17,848 | 11,314 | -37% | 1 | 1 | 0% | 2,474 | 2,073 | -16% | 0 | 0 | — |
case-10 | pass→pass | 13,285 | 3,682 | -72% | 1 | 1 | 0% | 2,016 | 927 | -54% | 0 | 0 | — |
case-11 | fail→pass | 13,494 | 4,620 | -66% | 1 | 1 | 0% | 1,826 | 783 | -57% | 0 | 0 | — |
case-12 | fail→pass | 19,488 | 11,274 | -42% | 1 | 1 | 0% | 2,535 | 2,279 | -10% | 0 | 0 | — |
case-13 | fail→pass | 10,628 | 3,753 | -65% | 1 | 1 | 0% | 1,824 | 1,044 | -43% | 0 | 0 | — |
case-14 | fail→pass | 10,939 | 6,852 | -37% | 1 | 1 | 0% | 1,872 | 1,278 | -32% | 0 | 0 | — |
case-15 | fail→pass | 10,614 | 4,134 | -61% | 1 | 1 | 0% | 1,722 | 997 | -42% | 0 | 0 | — |
case-16 | fail→pass | 17,447 | 22,085 | +27% | 1 | 1 | 0% | 2,624 | 1,853 | -29% | 0 | 0 | — |
case-17 | fail→pass | 11,932 | 5,355 | -55% | 1 | 1 | 0% | 1,883 | 1,279 | -32% | 0 | 0 | — |
case-18 | pass→pass | 16,054 | 12,974 | -19% | 1 | 1 | 0% | 2,524 | 2,566 | +2% | 0 | 0 | — |
case-19 | fail→pass | 9,688 | 5,148 | -47% | 1 | 1 | 0% | 1,503 | 1,248 | -17% | 0 | 0 | — |
case-20 | pass→pass | 8,972 | 12,003 | +34% | 1 | 1 | 0% | 1,752 | 2,437 | +39% | 0 | 0 | — |
case-21 | pass→pass | 4,885 | 3,845 | -21% | 1 | 1 | 0% | 976 | 1,187 | +22% | 0 | 0 | — |
case-22 | pass→pass | 6,040 | 5,866 | -3% | 1 | 1 | 0% | 1,128 | 1,557 | +38% | 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 +64 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.