ostack Benchmark provides a structured performance engineering workflow for capturing baselines, running page-level benchmarks, and detecting regressions across branches or over time. It automates page discovery, data collection via a browse daemon, extraction of navigation and resource timing (TTFB, FCP, LCP, DOM timings), and produces human-readable performance reports comparing current metrics to saved baselines.
Use this skill during PR reviews, before releases, or when investigating sudden load-time regressions. It is suitable for web engineers who want deterministic measurements (bundle sizes, request counts) and trend analysis across multiple runs. Use flags like --baseline, --quick, --diff and --trend to control behavior.
Targeted at Claude Code environments and other agent runtimes that provide a browse/perf daemon or JS evaluation capabilities.
This skill has not been reviewed by our automated audit pipeline yet.
