Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Use when auditing PWA readiness, adding an install prompt, or preparing a web app for submission to Microsoft Store or Google Play via PWA Builder.
.claude/skills/thedaviddias-pwa-installability/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-22 | ✓→✗ | ▼ Worse | 20% | 0% |
| case-20 | ✓→✓ | ＝ Same ✓ | -27% | 0% |
| case-01 | ✓→✓ | ＝ Same ✓ | -1% | 0% |
| case-02 | ✓→✓ | ＝ Same ✓ | -17% | 0% |
| case-03 | ✓→✓ | ＝ Same ✓ | -23% | 0% |
Installable PWAs appear in the browser's address bar install prompt and in app stores, giving users a native-app experience without an app store listing. Studies consistently show that installed PWAs have higher engagement and retention than browser-only equivalents — users who install spend 3× more time in the app on average.
Check whether this site meets the PWA installability criteria: HTTPS, valid manifest, service worker, and appropriate icons.
Add or fix the web app manifest, ensure a service worker is registered, and provide correctly sized maskable icons.
Explain what makes a web app installable as a PWA and what the manifest display modes mean for the user experience.
Review the manifest.json and icon assets. Flag missing required fields, incorrect icon sizes, absent maskable icons, and any mismatch between start_url and the actual app root.
For full implementation details, code examples, and framework-specific guidance, see references/rule.md.
Rule page: https://frontendchecklist.io/en/rules/html/pwa-installability
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-20 | pass→pass | 13,882 | 9,125 | -34% | 1 | 1 | 0% | 2,907 | 2,122 | -27% | 0 | 0 | — |
case-01 | pass→pass | 10,571 | 9,560 | -10% | 1 | 1 | 0% | 1,979 | 1,969 | -1% | 0 | 0 | — |
case-02 | pass→pass | 11,350 | 8,071 | -29% | 1 | 1 | 0% | 2,224 | 1,857 | -17% | 0 | 0 | — |
case-03 | pass→pass | 13,507 | 7,471 | -45% | 1 | 1 | 0% | 2,243 | 1,724 | -23% | 0 | 0 | — |
case-04 | pass→pass | 7,247 | 6,329 | -13% | 1 | 1 | 0% | 1,506 | 1,588 | +5% | 0 | 0 | — |
case-05 | pass→pass | 7,108 | 7,395 | +4% | 1 | 1 | 0% | 1,582 | 1,927 | +22% | 0 | 0 | — |
case-06 | pass→pass | 5,178 | 4,664 | -10% | 1 | 1 | 0% | 999 | 1,174 | +18% | 0 | 0 | — |
case-07 | pass→pass | 8,564 | 9,112 | +6% | 1 | 1 | 0% | 1,916 | 2,340 | +22% | 0 | 0 | — |
case-08 | pass→pass | 11,687 | 11,671 | -0% | 1 | 1 | 0% | 2,114 | 2,438 | +15% | 0 | 0 | — |
case-09 | pass→pass | 9,909 | 10,327 | +4% | 1 | 1 | 0% | 1,799 | 2,126 | +18% | 0 | 0 | — |
case-10 | fail→fail | 10,681 | 7,888 | -26% | 1 | 1 | 0% | 1,726 | 1,697 | -2% | 0 | 0 | — |
case-11 | pass→pass | 10,841 | 9,256 | -15% | 1 | 1 | 0% | 2,096 | 2,292 | +9% | 0 | 0 | — |
case-12 | pass→pass | 7,184 | 3,910 | -46% | 1 | 1 | 0% | 1,313 | 1,098 | -16% | 0 | 0 | — |
case-13 | pass→pass | 15,869 | 12,355 | -22% | 1 | 1 | 0% | 2,697 | 2,582 | -4% | 0 | 0 | — |
case-14 | pass→pass | 8,632 | 6,512 | -25% | 1 | 1 | 0% | 1,654 | 1,550 | -6% | 0 | 0 | — |
case-15 | pass→pass | 1,558 | 2,071 | +33% | 1 | 1 | 0% | 270 | 692 | +156% | 0 | 0 | — |
case-16 | pass→pass | 11,908 | 11,776 | -1% | 1 | 1 | 0% | 2,021 | 2,414 | +19% | 0 | 0 | — |
case-17 | pass→pass | 11,147 | 8,692 | -22% | 1 | 1 | 0% | 1,984 | 2,060 | +4% | 0 | 0 | — |
case-18 | pass→pass | 6,447 | 6,432 | -0% | 1 | 1 | 0% | 1,169 | 1,557 | +33% | 0 | 0 | — |
case-19 | pass→pass | 3,291 | 5,212 | +58% | 1 | 1 | 0% | 634 | 1,355 | +114% | 0 | 0 | — |
case-21 | pass→pass | 6,273 | 6,844 | +9% | 1 | 1 | 0% | 1,244 | 1,879 | +51% | 0 | 0 | — |
case-22 | pass→fail | 11,164 | 11,748 | +5% | 1 | 1 | 0% | 2,196 | 2,636 | +20% | 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 -33 percentage points is the difference between those two pass rates over the 22 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.
Other measured skills in the registry, with their headline benchmark lift.