Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Create or update component documentation pages in packages/ui/docs. Use when adding docs, demos, DocumentPage, usage.vue, or component doc routes.
.claude/skills/indielayer-docs-page/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -9% | 0% |
| case-02 | ✗→✓ | ▲ Improved | -26% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -61% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -40% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -40% | 0% |
Create packages/ui/docs/pages/component/<name>/index.vue:
const title, const description, const demos, const componentsdescription (sentence case) explaining what the example shows?raw code for <document-page>back / next to sibling folder names (preserve casing, e.g. inputGroup)<x-button>Add demo SFCs alongside (usage.vue, variants.vue, …).
Author API copy in the component SFC (not only on the docs page):
description on each prop option object (and shared factories in composables / variantBooleanProps)<script> default export, add docs: { slots, emits, methods } string maps (same place as validators)packages/ui/src/components/button/Button.vueDocumentPage; pnpm gen:llms embeds them in public/md/component/*.mdFrom packages/ui:
bashpnpm gen:search pnpm gen:llms # writes llms.txt, llms-full.txt, and public/md/component/*.md pnpm gen:sitemap
bashpnpm build:docs # vite-ssg prerender pnpm dev
Reference: packages/ui/docs/pages/component/button/index.vue
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 17,657 | 14,307 | -19% | 1 | 1 | 0% | 3,683 | 3,353 | -9% | 0 | 0 | — |
case-02 | fail→pass | 14,751 | 8,966 | -39% | 1 | 1 | 0% | 2,865 | 2,116 | -26% | 0 | 0 | — |
case-03 | fail→pass | 18,558 | 3,967 | -79% | 1 | 1 | 0% | 2,427 | 937 | -61% | 0 | 0 | — |
case-04 | pass→pass | 13,408 | 14,486 | +8% | 1 | 1 | 0% | 2,715 | 2,504 | -8% | 0 | 0 | — |
case-05 | pass→pass | 18,287 | 14,465 | -21% | 1 | 1 | 0% | 3,440 | 2,335 | -32% | 0 | 0 | — |
case-06 | pass→pass | 19,634 | 24,000 | +22% | 1 | 1 | 0% | 2,919 | 3,960 | +36% | 0 | 0 | — |
case-07 | fail→pass | 11,331 | 3,831 | -66% | 1 | 1 | 0% | 1,590 | 949 | -40% | 0 | 0 | — |
case-08 | fail→pass | 8,705 | 3,060 | -65% | 1 | 1 | 0% | 1,349 | 809 | -40% | 0 | 0 | — |
case-09 | pass→pass | 6,975 | 2,443 | -65% | 1 | 1 | 0% | 1,247 | 676 | -46% | 0 | 0 | — |
case-10 | pass→pass | 8,135 | 1,995 | -75% | 1 | 1 | 0% | 1,172 | 698 | -40% | 0 | 0 | — |
case-11 | pass→fail | 11,129 | 3,527 | -68% | 1 | 1 | 0% | 1,755 | 985 | -44% | 0 | 0 | — |
case-12 | fail→pass | 18,952 | 3,758 | -80% | 1 | 1 | 0% | 1,987 | 1,065 | -46% | 0 | 0 | — |
case-13 | fail→fail | 25,544 | 3,477 | -86% | 1 | 1 | 0% | 4,456 | 761 | -83% | 0 | 0 | — |
case-14 | pass→pass | 15,711 | 4,372 | -72% | 1 | 1 | 0% | 1,890 | 1,073 | -43% | 0 | 0 | — |
case-15 | pass→pass | 12,574 | 2,238 | -82% | 1 | 1 | 0% | 1,270 | 802 | -37% | 0 | 0 | — |
case-16 | fail→pass | 9,395 | 3,432 | -63% | 1 | 1 | 0% | 1,350 | 798 | -41% | 0 | 0 | — |
case-21 | fail→fail | 12,995 | 3,050 | -77% | 1 | 1 | 0% | 1,976 | 799 | -60% | 0 | 0 | — |
case-17 | fail→pass | 18,001 | 8,190 | -55% | 1 | 1 | 0% | 2,561 | 1,176 | -54% | 0 | 0 | — |
case-18 | fail→pass | 8,188 | 5,013 | -39% | 1 | 1 | 0% | 1,417 | 843 | -41% | 0 | 0 | — |
case-19 | fail→fail | 9,853 | 1,571 | -84% | 1 | 1 | 0% | 1,718 | 686 | -60% | 0 | 0 | — |
case-20 | fail→pass | 14,151 | 1,636 | -88% | 1 | 1 | 0% | 1,303 | 651 | -50% | 0 | 0 | — |
case-22 | fail→pass | 12,310 | 2,522 | -80% | 1 | 1 | 0% | 1,808 | 828 | -54% | 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 +45 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.
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.