Automate web performance benchmarking with browse daemon and JS evaluation.
npx skills add https://github.com/I194/PMTools_2.0 --skill benchmark-i194
Please help me install this Agent Skill. Skill: benchmark Source: https://github.com/I194/PMTools_2.0/tree/main/.claude/skills/gstack/benchmark Command: npx skills add https://github.com/I194/PMTools_2.0 --skill benchmark-i194
Performance regression detection across pages and PRs by collecting real metrics with a browse daemon and JS evaluation, establishing baselines and surfacing trends.
Run the benchmark on your target URL to start collecting performance data.
High-intent search queries and answers about installing and using this skill.
To detect web performance regressions on a PR, you can capture a baseline and compare current metrics against it. This Skill automates benchmarking using a browse daemon to collect real metrics and surface regressions on every PR.
A baseline benchmark captures per-page timing metrics including TTFB, FCP, LCP, and DOM thresholds. It also performs resource analysis to measure real web vitals from running pages using JS evaluation.
You generate a performance trend report by running a simple reporting pipeline that outputs JSON and markdown. This analyzes collected baselines and current metrics to guide optimizations across pages and resources.
Yes, collecting real web performance metrics requires a local setup with the browse daemon. You also need baseline storage and a reporting pipeline to fully automate the end-to-end benchmarking process.
The best way to compare current page performance against a historical baseline is to automate end-to-end benchmarking. This Skill captures baselines, compares current results on every PR, and surfaces trends across pages.
cheng80Detect performance regressions by comparing page load metrics against baselines.
cffrankParse Vitest JSON output to generate test quality metrics and reports.
olaafrossiBenchmark web pages against baselines and report performance regressions.
SergeiGolosAudit web performance with Lighthouse and analyze bundle sizes.
