capture-all-pages

Capturing high-resolution screenshots for translation QA requires Playwright as a runtime dependency and a running dev server. The script automates navigation across pages and outputs PNGs to an __inbox directory along with a summary.json detailing results.

Yes, you can capture a subset of pages by passing the --pages argument with a specific page count to the capture script. You can also use the --base-url argument to target a custom dev server URL for capturing pages.

Playwright is used as the runtime dependency to batch capture screenshots at 2000x1600 high resolution. It automates browser navigation across all manual pages and ensures consistent high-fidelity rendering for visual translation verification.

The best way to verify layout consistency is to batch capture high-resolution PNG screenshots of all manual pages and review them visually. The skill organizes outputs under an __inbox directory and generates a summary.json file detailing capture results for tracking.

You need a running dev server to capture manual page screenshots because the Playwright script navigates to live URLs to render and capture the pages accurately. A custom base URL can be specified to point the script to the correct local or remote server environment.