profile-browsing

Profile app performance by measuring Web Vitals and React rerenders to identify bottlenecks. The profiler targets development and QA, diagnosing slow navigations, long tasks, and excessive rerenders across complex routes to optimize user-perceived load times.

Yes, you can measure per-route metrics across multiple sessions. The profiler orchestrates parallel playwright subagents to collect navigation timing, long tasks, and per-view metrics, reading react-scan data to diagnose slow interactions across routes.

The best way to diagnose slow navigations and excessive rerenders is using a two-layer profiling approach. It covers browser-level symptoms like Web Vitals and layout shifts, plus React-level diagnosis including commit counts, render bursts, and per-component render data.

Yes, you need exactly one dev server running to start profiling. Once the dev server is active, you run the profiler across desired routes using the predefined batch plan to collect navigation timing and per-view metrics for analysis.

Yes, it works with Playwright by coordinating browser sessions via playwright subagents. These subagents collect navigation timing, long tasks, and per-route metrics across multiple sessions to export per-view metrics for performance analysis.

Use end-to-end browsing performance profiling when you need to diagnose complex routes with slow navigations, long tasks, or excessive rerenders. It coordinates parallel subagents to measure Web Vitals and React commits that manual inspection cannot easily capture.