vc-web-testing

Run the init-playwright script to generate a best-practice setup including playwright.config.ts, authentication fixtures, example tests, and global setup. It supports an optional component testing configuration via the --ct flag.

Use the analyze-test-results script with --playwright and --vitest flags pointing to JSON result files. It outputs a unified summary in text, JSON, or Markdown, including pass rates, failures, and flaky test counts.

Playwright component testing runs real browsers for cross-browser and visual verification, while Vitest browser mode is faster for logic-focused tests. Choose Playwright CT when CSS, layout, or multi-browser fidelity matters.

Integrate @axe-core/playwright by creating an AxeBuilder instance with the page object and calling analyze(). Assert that results.violations is empty, and filter by WCAG tags like wcag2aa for targeted compliance checks.

Flakiness usually comes from hard waits, shared state between tests, or animation timing. Replace fixed sleeps with explicit waits or Playwright auto-waiting, isolate test data per worker, and disable CSS animations during test runs.

Use the k6 setup function to authenticate once and return a token, then pass it as a Bearer header in the default function. Define stages for ramp-up and thresholds like p95 latency under 500ms to gate performance.