Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Best practices and patterns for Next.js App Router, Server Actions, and Routing in this project.
.claude/skills/aiskillstore-next-js-patterns/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-15 | ✗→✓ | ▲ Improved | -46% | 0% |
| case-01 | ✓→✓ | ＝ Same ✓ | -7% | 0% |
| case-02 | ✓→✓ | ＝ Same ✓ | -9% | 0% |
| case-03 | ✓→✓ | ＝ Same ✓ | -41% | 0% |
| case-04 | ✓→✓ | ＝ Same ✓ | -26% | 0% |
We use the Next.js 15 App Router located in app/.
app/[route]/page.tsx"use client" directive at the top for components requiring state (useState, useEffect) or browser APIs. otherwise default to Server Components.app/layout.tsx (Root), app/[route]/layout.tsx (Nested).Link from next/link for internal navigation.useRouter from next/navigation for programmatic navigation (inside Client Components).tsximport Link from "next/link"; import { useRouter } from "next/navigation"; // Link <Link href="/dashboard">Dashboard</Link> // Router const router = useRouter(); router.push('/login');
await fetch() or DB calls.useEffect or SWR/TanStack Query (if added later). Currently using standard fetch in useEffect.next/font/google (e.g., Poppins) in app/layout.tsx.body className.export const metadata: Metadata = { ... } in page.tsx or layout.tsx for SEO.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | pass→pass | 11,811 | 8,722 | -26% | 1 | 1 | 0% | 1,933 | 1,807 | -7% | 0 | 0 | — |
case-02 | pass→pass | 5,054 | 2,350 | -54% | 1 | 1 | 0% | 831 | 754 | -9% | 0 | 0 | — |
case-03 | pass→pass | 7,607 | 2,963 | -61% | 1 | 1 | 0% | 1,391 | 818 | -41% | 0 | 0 | — |
case-04 | pass→pass | 6,802 | 3,134 | -54% | 1 | 1 | 0% | 1,176 | 868 | -26% | 0 | 0 | — |
case-05 | pass→pass | 3,403 | 1,991 | -41% | 1 | 1 | 0% | 454 | 670 | +48% | 0 | 0 | — |
case-06 | pass→pass | 4,664 | 3,708 | -20% | 1 | 1 | 0% | 835 | 938 | +12% | 0 | 0 | — |
case-07 | pass→pass | 5,297 | 2,511 | -53% | 1 | 1 | 0% | 941 | 710 | -25% | 0 | 0 | — |
case-08 | pass→pass | 7,410 | 3,347 | -55% | 1 | 1 | 0% | 1,328 | 901 | -32% | 0 | 0 | — |
case-09 | pass→pass | 6,170 | 2,314 | -62% | 1 | 1 | 0% | 1,250 | 700 | -44% | 0 | 0 | — |
case-10 | pass→pass | 2,973 | 2,215 | -25% | 1 | 1 | 0% | 491 | 771 | +57% | 0 | 0 | — |
case-11 | pass→pass | 2,347 | 2,448 | +4% | 1 | 1 | 0% | 360 | 614 | +71% | 0 | 0 | — |
case-12 | pass→pass | 11,687 | 5,716 | -51% | 1 | 1 | 0% | 2,181 | 1,360 | -38% | 0 | 0 | — |
case-17 | pass→pass | 6,867 | 5,183 | -25% | 1 | 1 | 0% | 1,167 | 1,200 | +3% | 0 | 0 | — |
case-13 | pass→pass | 8,735 | 8,121 | -7% | 1 | 1 | 0% | 1,719 | 1,965 | +14% | 0 | 0 | — |
case-14 | pass→pass | 3,437 | 2,506 | -27% | 1 | 1 | 0% | 608 | 798 | +31% | 0 | 0 | — |
case-15 | fail→pass | 13,664 | 4,987 | -64% | 1 | 1 | 0% | 2,212 | 1,193 | -46% | 0 | 0 | — |
case-16 | pass→pass | 8,604 | 5,590 | -35% | 1 | 1 | 0% | 1,664 | 1,266 | -24% | 0 | 0 | — |
case-18 | pass→pass | 5,812 | 3,516 | -40% | 1 | 1 | 0% | 1,026 | 951 | -7% | 0 | 0 | — |
case-19 | pass→pass | 10,075 | 5,896 | -41% | 1 | 1 | 0% | 1,781 | 1,389 | -22% | 0 | 0 | — |
case-20 | pass→pass | 12,227 | 9,203 | -25% | 1 | 1 | 0% | 2,274 | 2,103 | -8% | 0 | 0 | — |
case-21 | pass→pass | 15,173 | 9,503 | -37% | 1 | 1 | 0% | 2,981 | 2,047 | -31% | 0 | 0 | — |
case-22 | pass→pass | 12,193 | 11,638 | -5% | 1 | 1 | 0% | 2,234 | 2,572 | +15% | 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.