Stats
Actions
Tags
From perf
Guides controlled performance experiments to validate hypotheses on code changes. Enforces clean baselines, single changes, multiple runs, evidence recording, and verdict output.
How this skill is triggered — by the user, by Claude, or both
Slash command
/perf:theory-testerThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Test hypotheses using controlled experiments.
Test hypotheses using controlled experiments.
Follow docs/perf-requirements.md as the canonical contract.
hypothesis: <id>
change: <summary>
delta: <metrics>
verdict: accept|reject|inconclusive
evidence:
- command: <benchmark command>
- files: <changed files>
npx claudepluginhub composiohq/awesome-claude-plugins --plugin perfMulti-phase performance investigation workflow for establishing baselines, profiling, and making evidence-gated optimization decisions. Use when debugging latency, throughput regressions, or the question "why is this slow?"
Creates and runs reliable benchmarks to measure code change impacts on performance, including latency, throughput. Supports Node.js (vitest, tinybench), Python (pytest-benchmark), frontend (Lighthouse CI), with warmup, stats.
Use when a backpressured loop needs to run benchmarks on a performance-sensitive project and decide whether a change is a regression, an improvement, or a wash — per-iteration sanity checks and the full pre-done run.