benchmark

Detect Core Web Vitals regressions by comparing current PR build metrics against established baselines. This skill measures page load timings and resource sizes using the browse daemon, generating actionable insights when current performance drops below baseline expectations.

Metrics tracked for web performance baselines include TTFB, FCP, LCP, and DOM timings. The system establishes baselines for these Core Web Vitals and resource sizes, then performs trend analysis over time to reveal slowdowns and performance drift across PRs.

Baseline page load times by invoking the benchmark skill on a target URL. It uses the browse daemon to measure page load timings and resource sizes, establishing the baseline dataset required for future PR comparisons and regression detection.

Yes, you need an existing baseline dataset to compare current PR performance. The skill requires a pre-established baseline and a target URL to perform comparisons, measuring current metrics against the baseline to identify regressions and generate actionable insights.

You can test pages measured with the browse daemon for performance regression. The scope includes pages tested with this daemon, capturing metrics like TTFB, FCP, LCP, and DOM timings to compare current PR builds against established baselines.

Performance trend analysis matters because it reveals slowdowns and performance drift over time. By continuously comparing PR builds against baselines for Core Web Vitals and page load timings, teams can identify regressions early and maintain stable web app performance.