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 Load scripts with defer, async, or type=module. Validate the final browser-facing markup, not just the source framework abstraction.
.claude/skills/thedaviddias-defer-async/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 82% | 0% |
| case-02 | ✓→✓ | ＝ Same ✓ | -1% | 0% |
| case-03 | ✓→✓ | ＝ Same ✓ | 9% | 0% |
| case-04 | ✓→✓ | ＝ Same ✓ | -9% | 0% |
| case-05 | ✓→✓ | ＝ Same ✓ | -24% | 0% |
A plain <script> tag in the document head blocks all HTML parsing until the script downloads, parses, and executes. On a slow network this can add seconds of white-screen time before any content renders. defer and async allow the browser to continue parsing HTML while the script downloads, reducing Time to First Contentful Paint dramatically.
Find all script tags in this HTML file. Flag any in the <head> without defer or async, and any at the bottom of <body> that could be in <head> with defer.
Add defer or async to script tags in the document head, or convert to type=module where ES modules are used.
Explain the difference between defer, async, and type=module script loading, and when to use each.
Review templates, server-rendered HTML, and shared components that output markup related to Load scripts with defer, async, or type=module. 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/defer-async
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 5,357 | 7,449 | +39% | 1 | 1 | 0% | 995 | 1,806 | +82% | 0 | 0 | — |
case-02 | pass→pass | 8,068 | 4,816 | -40% | 1 | 1 | 0% | 1,324 | 1,306 | -1% | 0 | 0 | — |
case-03 | pass→pass | 5,255 | 5,474 | +4% | 1 | 1 | 0% | 1,114 | 1,213 | +9% | 0 | 0 | — |
case-04 | pass→pass | 6,775 | 4,910 | -28% | 1 | 1 | 0% | 1,419 | 1,288 | -9% | 0 | 0 | — |
case-05 | pass→pass | 7,251 | 3,953 | -45% | 1 | 1 | 0% | 1,385 | 1,053 | -24% | 0 | 0 | — |
case-06 | pass→pass | 6,066 | 4,941 | -19% | 1 | 1 | 0% | 1,191 | 1,381 | +16% | 0 | 0 | — |
case-07 | pass→pass | 5,585 | 4,417 | -21% | 1 | 1 | 0% | 1,149 | 1,176 | +2% | 0 | 0 | — |
case-08 | pass→pass | 8,002 | 6,185 | -23% | 1 | 1 | 0% | 1,529 | 1,513 | -1% | 0 | 0 | — |
case-09 | pass→pass | 17,742 | 3,197 | -82% | 1 | 1 | 0% | 3,333 | 1,017 | -69% | 0 | 0 | — |
case-10 | pass→pass | 8,910 | 5,886 | -34% | 1 | 1 | 0% | 1,618 | 1,415 | -13% | 0 | 0 | — |
case-11 | pass→pass | 4,910 | 4,128 | -16% | 1 | 1 | 0% | 1,078 | 1,150 | +7% | 0 | 0 | — |
case-12 | pass→pass | 10,373 | 9,933 | -4% | 1 | 1 | 0% | 2,154 | 2,131 | -1% | 0 | 0 | — |
case-13 | pass→pass | 8,343 | 5,871 | -30% | 1 | 1 | 0% | 1,549 | 1,438 | -7% | 0 | 0 | — |
case-14 | pass→pass | 6,417 | 4,652 | -28% | 1 | 1 | 0% | 1,143 | 1,241 | +9% | 0 | 0 | — |
case-15 | pass→pass | 4,330 | 3,825 | -12% | 1 | 1 | 0% | 818 | 1,038 | +27% | 0 | 0 | — |
case-16 | pass→pass | 6,660 | 5,113 | -23% | 1 | 1 | 0% | 1,346 | 1,269 | -6% | 0 | 0 | — |
case-17 | pass→pass | 3,671 | 4,012 | +9% | 1 | 1 | 0% | 630 | 1,133 | +80% | 0 | 0 | — |
case-18 | pass→pass | 10,096 | 8,893 | -12% | 1 | 1 | 0% | 1,716 | 1,720 | +0% | 0 | 0 | — |
case-19 | pass→pass | 5,778 | 4,116 | -29% | 1 | 1 | 0% | 1,174 | 1,165 | -1% | 0 | 0 | — |
case-20 | pass→pass | 8,215 | 3,927 | -52% | 1 | 1 | 0% | 1,341 | 1,168 | -13% | 0 | 0 | — |
case-21 | pass→pass | 5,825 | 5,678 | -3% | 1 | 1 | 0% | 1,383 | 1,444 | +4% | 0 | 0 | — |
case-22 | pass→pass | 8,242 | 5,447 | -34% | 1 | 1 | 0% | 1,625 | 1,403 | -14% | 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.