Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Use when auditing slow page loads, heavy assets, or rendering delays related to Optimize web font loading. Verify the actual bottleneck in DevTools, Lighthouse, or field data before recommending changes.
.claude/skills/thedaviddias-font-loading/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-08 | ✗→✓ | ▲ Improved | 6% | 0% |
| case-01 | ✓→✓ | ＝ Same ✓ | -6% | 0% |
| case-02 | ✓→✓ | ＝ Same ✓ | 9% | 0% |
| case-03 | ✓→✓ | ＝ Same ✓ | -18% | 0% |
| case-04 | ✓→✓ | ＝ Same ✓ | 13% | 0% |
Slow font loading can cause 'Flash of Invisible Text' (FOIT) or significant layout shifts, negatively impacting both user experience and Core Web Vitals.
font-display: swap to prevent invisible text during loadCheck font loading strategies and verify that font-display is used and critical fonts are preloaded.
Add font-display: swap to @font-face declarations and use <link rel='preload'> for the most important fonts.
Explain how font loading affects perceived performance and layout stability.
Review the routes, assets, and loading behavior that affect Optimize web font loading. Flag exact files, requests, or rendering steps that add unnecessary network, CPU, or layout cost, and describe the measurement method used to confirm the issue.
For full implementation details, code examples, and framework-specific guidance, see references/rule.md.
Rule page: https://frontendchecklist.io/en/rules/performance/font-loading
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | pass→pass | 11,665 | 9,265 | -21% | 1 | 1 | 0% | 2,180 | 2,053 | -6% | 0 | 0 | — |
case-02 | pass→pass | 13,757 | 12,590 | -8% | 1 | 1 | 0% | 2,456 | 2,685 | +9% | 0 | 0 | — |
case-03 | pass→pass | 11,770 | 8,792 | -25% | 1 | 1 | 0% | 2,239 | 1,847 | -18% | 0 | 0 | — |
case-04 | pass→pass | 4,006 | 3,105 | -22% | 1 | 1 | 0% | 780 | 881 | +13% | 0 | 0 | — |
case-05 | pass→pass | 6,484 | 6,824 | +5% | 1 | 1 | 0% | 1,252 | 1,402 | +12% | 0 | 0 | — |
case-15 | pass→pass | 9,732 | 7,079 | -27% | 1 | 1 | 0% | 1,657 | 1,733 | +5% | 0 | 0 | — |
case-06 | pass→pass | 11,272 | 6,955 | -38% | 1 | 1 | 0% | 2,158 | 1,586 | -27% | 0 | 0 | — |
case-07 | pass→pass | 10,208 | 6,386 | -37% | 1 | 1 | 0% | 1,804 | 1,456 | -19% | 0 | 0 | — |
case-08 | fail→pass | 15,414 | 13,825 | -10% | 1 | 1 | 0% | 2,519 | 2,663 | +6% | 0 | 0 | — |
case-09 | pass→pass | 15,661 | 15,290 | -2% | 1 | 1 | 0% | 2,597 | 2,862 | +10% | 0 | 0 | — |
case-16 | pass→pass | 2,484 | 3,129 | +26% | 1 | 1 | 0% | 420 | 788 | +88% | 0 | 0 | — |
case-10 | pass→pass | 4,023 | 3,985 | -1% | 1 | 1 | 0% | 705 | 973 | +38% | 0 | 0 | — |
case-11 | pass→pass | 18,042 | 10,907 | -40% | 1 | 1 | 0% | 2,222 | 2,001 | -10% | 0 | 0 | — |
case-12 | pass→pass | 12,643 | 12,076 | -4% | 1 | 1 | 0% | 2,171 | 2,502 | +15% | 0 | 0 | — |
case-13 | pass→pass | 2,827 | 2,914 | +3% | 1 | 1 | 0% | 478 | 714 | +49% | 0 | 0 | — |
case-14 | pass→pass | 9,666 | 8,573 | -11% | 1 | 1 | 0% | 1,719 | 1,909 | +11% | 0 | 0 | — |
case-17 | pass→pass | 16,741 | 11,820 | -29% | 1 | 1 | 0% | 2,726 | 2,233 | -18% | 0 | 0 | — |
case-18 | pass→pass | 11,290 | 10,152 | -10% | 1 | 1 | 0% | 1,982 | 2,014 | +2% | 0 | 0 | — |
case-19 | pass→pass | 3,194 | 2,887 | -10% | 1 | 1 | 0% | 574 | 950 | +66% | 0 | 0 | — |
case-20 | pass→pass | 11,999 | 11,281 | -6% | 1 | 1 | 0% | 2,218 | 2,262 | +2% | 0 | 0 | — |
case-21 | pass→pass | 3,378 | 2,194 | -35% | 1 | 1 | 0% | 603 | 649 | +8% | 0 | 0 | — |
case-22 | pass→pass | 7,179 | 7,780 | +8% | 1 | 1 | 0% | 1,272 | 1,719 | +35% | 0 | 0 | — |
case-23 | pass→pass | 5,192 | 4,843 | -7% | 1 | 1 | 0% | 1,135 | 1,176 | +4% | 0 | 0 | — |
case-24 | pass→pass | 4,264 | 5,285 | +24% | 1 | 1 | 0% | 783 | 1,186 | +51% | 0 | 0 | — |
case-25 | pass→pass | 7,437 | 4,107 | -45% | 1 | 1 | 0% | 1,222 | 950 | -22% | 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. 25 cases were attempted. The headline lift of +4 percentage points is the difference between those two pass rates over the 25 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.