Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Next.js 16+ and Turbopack — incremental bundling, FS caching, dev speed, and when to use Turbopack vs webpack.
.claude/skills/loulanyue-nextjs-turbopack/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-06 | ✗→✓ | ▲ Improved | -56% | 0% |
| case-12 | ✗→✓ | ▲ Improved | -41% | 0% |
| case-15 | ✗→✓ | ▲ Improved | -23% | 0% |
| case-16 | ✗→✓ | ▲ Improved | -73% | 0% |
| case-18 | ✗→✓ | ▲ Improved | -59% | 0% |
Next.js 16+ uses Turbopack by default for local development: an incremental bundler written in Rust that significantly speeds up dev startup and hot updates.
--webpack (or --no-turbopack depending on your Next.js version; check the docs for your release).next build) may use Turbopack or webpack depending on Next.js version; check the official Next.js docs for your version.Use when: developing or debugging Next.js 16+ apps, diagnosing slow dev startup or HMR, or optimizing production bundles.
next dev runs with Turbopack unless disabled..next; no extra config needed for basic use.bashnext dev next build next start
Run next dev for local development with Turbopack. Use the Bundle Analyzer (see Next.js docs) to optimize code-splitting and trim large dependencies. Prefer App Router and server components where possible.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-09 | pass→pass | 2,447 | 1,858 | -24% | 1 | 1 | 0% | 354 | 810 | +129% | 0 | 0 | — |
case-01 | pass→pass | 4,216 | 1,815 | -57% | 1 | 1 | 0% | 709 | 820 | +16% | 0 | 0 | — |
case-02 | pass→pass | 4,942 | 2,732 | -45% | 1 | 1 | 0% | 804 | 1,000 | +24% | 0 | 0 | — |
case-03 | pass→pass | 3,992 | 2,870 | -28% | 1 | 1 | 0% | 568 | 935 | +65% | 0 | 0 | — |
case-04 | pass→pass | 10,600 | 3,704 | -65% | 1 | 1 | 0% | 1,609 | 1,157 | -28% | 0 | 0 | — |
case-05 | pass→pass | 3,208 | 2,150 | -33% | 1 | 1 | 0% | 486 | 805 | +66% | 0 | 0 | — |
case-06 | fail→pass | 12,905 | 3,031 | -77% | 1 | 1 | 0% | 2,186 | 969 | -56% | 0 | 0 | — |
case-07 | pass→pass | 7,927 | 4,822 | -39% | 1 | 1 | 0% | 1,265 | 1,357 | +7% | 0 | 0 | — |
case-08 | pass→pass | 2,637 | 2,248 | -15% | 1 | 1 | 0% | 378 | 887 | +135% | 0 | 0 | — |
case-10 | pass→pass | 5,935 | 3,080 | -48% | 1 | 1 | 0% | 946 | 994 | +5% | 0 | 0 | — |
case-11 | pass→pass | 12,522 | 2,191 | -83% | 1 | 1 | 0% | 2,085 | 873 | -58% | 0 | 0 | — |
case-12 | fail→pass | 12,668 | 4,390 | -65% | 1 | 1 | 0% | 2,137 | 1,259 | -41% | 0 | 0 | — |
case-13 | pass→pass | 12,015 | 8,456 | -30% | 1 | 1 | 0% | 1,970 | 2,026 | +3% | 0 | 0 | — |
case-14 | pass→pass | 8,038 | 4,325 | -46% | 1 | 1 | 0% | 1,243 | 1,200 | -3% | 0 | 0 | — |
case-15 | fail→pass | 13,115 | 6,881 | -48% | 1 | 1 | 0% | 2,239 | 1,713 | -23% | 0 | 0 | — |
case-16 | fail→pass | 15,109 | 1,386 | -91% | 1 | 1 | 0% | 2,744 | 741 | -73% | 0 | 0 | — |
case-17 | pass→pass | 6,495 | 2,602 | -60% | 1 | 1 | 0% | 1,083 | 934 | -14% | 0 | 0 | — |
case-18 | fail→pass | 13,058 | 2,044 | -84% | 1 | 1 | 0% | 1,962 | 795 | -59% | 0 | 0 | — |
case-19 | pass→pass | 2,421 | 2,382 | -2% | 1 | 1 | 0% | 389 | 856 | +120% | 0 | 0 | — |
case-20 | pass→pass | 11,108 | 9,881 | -11% | 1 | 1 | 0% | 2,139 | 2,606 | +22% | 0 | 0 | — |
case-21 | pass→pass | 2,682 | 3,382 | +26% | 1 | 1 | 0% | 446 | 1,139 | +155% | 0 | 0 | — |
case-22 | pass→pass | 11,236 | 7,073 | -37% | 1 | 1 | 0% | 2,018 | 1,812 | -10% | 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 +23 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.