Demo — mass Nuxt e2e

Same browser scenarios (Playwright goto + Nuxt hydration + clicks/cookies), two prepare strategies — measured wall time, not estimates.

Think nuxt-i18n-next-scale: many files, real live Nuxt, not a 1s $fetch toy.

Loading demo…

What is compared

	Naive (pnpm test:naive)	Shared (pnpm test:shared)
Recipe	app01…app10 — one id per file	One app for all files
Prepare	Paid per file (cold artifacts wiped before capture)	Once (one recipe id, shared host)
Assertions	Same 150 browser flows	Same 150 browser flows

Try it

cd examples/mass-nuxt
pnpm generate:e2e
pnpm test:shared   # untestutils shared prepare
pnpm test:naive    # ordinary per-file prepare (slow)

Source: examples/mass-nuxt

Refresh captured results

pnpm run demo:capture

Runs both suites after wiping their artifact dirs, writes docs/public/demo/{shared,naive}.log + results.json. Not part of default preflight (naive is intentionally slow).

Next

• Why untestutils
• Getting started
• How it works