benchmark

To detect web performance regressions in a pull request, you establish performance baselines for page load times and Core Web Vitals, then compare before and after results to identify slowdowns. This validates PRs and tracks trends over time.

The best way to establish a baseline for Core Web Vitals is by collecting navigation timing and resource data using the browse daemon. This captures page load metrics into baseline JSON files for future trend analysis and regression detection.

Yes, you need the browse daemon to collect page load timing data. The daemon gathers navigation timing and resource data utilizing JavaScript performance APIs to execute the end-to-end workflow from data collection to reporting.

Yes, you can track long-term performance trends across multiple builds. The system compares metrics across PRs and captures baselines over time to identify slowdowns in page load times and resource sizes for long-term analysis.

To measure page load performance for a specific URL, run /benchmark <url>. This measures performance, captures a baseline, and compares the results against previous data to identify any web performance regressions.