performance-profiling

To identify web performance bottlenecks, you collect evidence-based metrics using Lighthouse and DevTools traces before optimizing. This approach prioritizes critical metrics like LCP, INP, and CLS to ensure data-driven improvements.

Yes, you can run reproducible performance profiling in CI pipelines. The Skill executes end-to-end audits using the lighthouse_audit.py script, enabling consistent web performance measurements across multi-tenant builds.

Evidence-based performance auditing is the process of collecting Lighthouse metrics and DevTools traces to prescribe optimizations. It ensures actions target actual bottlenecks affecting user experience rather than assumptions.

To start a Lighthouse audit, you execute the lighthouse_audit.py script with a target URL. This initiates the end-to-end performance profiling process and generates the required evidence-based measurements.

Yes, this performance profiling approach tracks Core Web Vitals by focusing on critical metrics. It specifically prioritizes LCP, INP, CLS, and bundle size changes that directly impact the user experience.

You should not rely solely on Lighthouse when you lack DevTools-compatible traces or cannot run the lighthouse_audit.py script. End-to-end profiling requires both to ensure accurate, reproducible evidence-based measurements.