Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Use this skill to bootstrap a new Next.js application with TypeScript, Tailwind CSS, React, and an API backend.
.claude/skills/miguelaperez-nextjs-app-scaffold/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-04 | ✗→✓ | ▲ Improved | -41% | 0% |
| case-06 | ✗→✓ | ▲ Improved | -82% | 0% |
| case-15 | ✗→✓ | ▲ Improved | -11% | 0% |
| case-16 | ✗→✓ | ▲ Improved | -63% | 0% |
| case-17 | ✗→✓ | ▲ Improved | -28% | 0% |
This skill provides a standardized way to create premium, production-ready Next.js applications.
When asked to create a new Next.js app, follow these steps:
Use the provided helper script to initialize a new Next.js project with the correct flags and a docs folder:
bash./scripts/scaffold.sh npm install npm run dev
The script will run npx create-next-app@latest with:
src/ directory@/*docs folder creationnpm run githook:install)Once the project is initialized, immediately call the onboarding skill to document the initial state and verify that the docs folder is correctly set up.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-03 | fail→fail | 21,660 | 4,169 | -81% | 1 | 1 | 0% | 5,373 | 481 | -91% | 0 | 0 | — |
case-01 | fail→fail | 3,278 | 4,289 | +31% | 1 | 1 | 0% | 364 | 487 | +34% | 0 | 0 | — |
case-02 | fail→fail | 26,737 | 3,938 | -85% | 1 | 1 | 0% | 6,179 | 468 | -92% | 0 | 0 | — |
case-04 | fail→pass | 6,598 | 2,344 | -64% | 1 | 1 | 0% | 1,219 | 723 | -41% | 0 | 0 | — |
case-05 | fail→fail | 4,595 | 1,900 | -59% | 1 | 1 | 0% | 738 | 586 | -21% | 0 | 0 | — |
case-06 | fail→pass | 13,741 | 1,434 | -90% | 1 | 1 | 0% | 2,770 | 490 | -82% | 0 | 0 | — |
case-07 | pass→pass | 5,930 | 1,502 | -75% | 1 | 1 | 0% | 1,077 | 518 | -52% | 0 | 0 | — |
case-08 | pass→pass | 3,592 | 1,966 | -45% | 1 | 1 | 0% | 693 | 570 | -18% | 0 | 0 | — |
case-09 | pass→pass | 12,665 | 1,696 | -87% | 1 | 1 | 0% | 2,318 | 552 | -76% | 0 | 0 | — |
case-10 | pass→pass | 2,690 | 2,328 | -13% | 1 | 1 | 0% | 473 | 469 | -1% | 0 | 0 | — |
case-11 | pass→pass | 8,807 | 2,520 | -71% | 1 | 1 | 0% | 1,624 | 664 | -59% | 0 | 0 | — |
case-12 | pass→pass | 4,141 | 1,466 | -65% | 1 | 1 | 0% | 696 | 489 | -30% | 0 | 0 | — |
case-13 | pass→pass | 12,184 | 1,660 | -86% | 1 | 1 | 0% | 2,092 | 611 | -71% | 0 | 0 | — |
case-14 | pass→pass | 7,079 | 2,023 | -71% | 1 | 1 | 0% | 1,134 | 587 | -48% | 0 | 0 | — |
case-15 | fail→pass | 6,112 | 3,539 | -42% | 1 | 1 | 0% | 1,044 | 927 | -11% | 0 | 0 | — |
case-16 | fail→pass | 8,656 | 1,831 | -79% | 1 | 1 | 0% | 1,534 | 561 | -63% | 0 | 0 | — |
case-17 | fail→pass | 5,075 | 1,865 | -63% | 1 | 1 | 0% | 803 | 575 | -28% | 0 | 0 | — |
case-18 | fail→pass | 7,695 | 2,273 | -70% | 1 | 1 | 0% | 1,418 | 695 | -51% | 0 | 0 | — |
case-19 | pass→pass | 6,010 | 1,893 | -69% | 1 | 1 | 0% | 973 | 571 | -41% | 0 | 0 | — |
case-20 | pass→pass | 11,956 | 11,142 | -7% | 1 | 1 | 0% | 2,351 | 2,725 | +16% | 0 | 0 | — |
case-21 | pass→pass | 14,134 | 13,163 | -7% | 1 | 1 | 0% | 2,758 | 2,906 | +5% | 0 | 0 | — |
case-22 | pass→pass | 11,772 | 10,951 | -7% | 1 | 1 | 0% | 2,231 | 2,436 | +9% | 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, and 19 counted toward the lift figure. The other 3 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +27 percentage points is the difference between those two pass rates over the 19 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.