benchmark

Detect web performance regressions in pull requests by capturing baseline metrics and comparing them against current runs. This skill measures TTFB, FCP, LCP, and bundle sizes to surface timing regressions and resource overhead on branches and deploys.

Web vitals benchmarking tracks TTFB, FCP, LCP, DOM Interactive, DOM Complete, and Full Load. It measures resource timing and page load performance to establish baselines and detect regressions across branches and deploys.

Save a performance baseline by running the benchmark against a target URL to capture current timing data. Compare this saved baseline against new deploys or PRs to identify timing regressions and resource overhead.

You can automate performance regression alerts for branches without external dependencies. This skill uses the browse daemon to measure load times and resource timing, then reports regression alerts and trend data.

The best way to monitor bundle size regressions across PRs is to capture baseline resource metrics and compare them on each pull request. This skill reports bundle sizes alongside DOM metrics to flag resource overhead.

This performance benchmarking approach works for continuous regression detection by applying measurements to PRs, branches, and deploys. It saves baselines, compares current runs, and generates trend reporting to help teams act fast.