benchmark

To detect performance regressions in a pull request, you compare current runs against stored baselines using the browse daemon. It captures navigation timing, Core Web Vitals, and resource sizes to surface diffs and recommendations.

Performance baseline capture stores initial page load times, Core Web Vitals, and resource metrics under .gstack/benchmark-reports. You need it before detecting regressions to establish a reference point for future PR comparisons.

You track Core Web Vitals and page load trends over time by running the benchmark across PRs to collect navigation and resource timing. It generates trend reports comparing current performance against historical baselines.

Yes, you can run a quick performance check on a single URL to perform a full performance audit. This collects artifact metadata and navigation timing for immediate comparison against existing baselines.

Yes, performance regression detection works without external dependencies. It uses the integrated browse daemon to establish baselines and collect resource timing metadata directly within your workflow.