PE

perf-theory-tester

Executes controlled performance benchmarks to test specific optimization hypotheses.

Install

mkdir -p .claude/skills/perf-theory-tester && curl -L -o skill.zip "https://agentskills.codes/api/skills/download/3357" && unzip -o skill.zip -d .claude/skills/perf-theory-tester && rm skill.zip

Installs to .claude/skills/perf-theory-tester

Activation

This is the description your AI agent reads to decide when to run this skill — the better it matches your request, the more reliably it fires.

Use when running controlled perf experiments to validate hypotheses.
68 chars✓ has a “when” trigger
Intermediate

Key capabilities

  • Execute controlled performance experiments
  • Validate optimization hypotheses
  • Run benchmark validation passes
  • Record experiment results and deltas

How it works

The skill manages the experiment lifecycle by enforcing a clean baseline, applying a single change, running benchmarks, and recording the results.

Inputs & outputs

You give it
Performance hypothesis and single code change
You get back
Experiment verdict and benchmark delta metrics

When to use perf-theory-tester

  • Validating optimization hypotheses
  • Benchmarking code changes
  • Running controlled performance regression tests

About this skill

perf-theory-tester

Test hypotheses using controlled experiments.

Follow docs/perf-requirements.md as the canonical contract.

Required Steps

  1. Confirm baseline is clean.
  2. Apply a single change tied to the hypothesis.
  3. Run 2+ validation passes.
  4. Revert to baseline before the next experiment.

Output Format

hypothesis: <id>
change: <summary>
delta: <metrics>
verdict: accept|reject|inconclusive
evidence:
  - command: <benchmark command>
  - files: <changed files>

Constraints

  • One change per experiment.
  • No parallel benchmarks.
  • Record evidence for each run.

When not to use it

  • Running non-controlled or parallel benchmarks
  • General code testing

Prerequisites

Clean baseline environmentDefined performance hypothesis

Limitations

  • Strictly one change per experiment
  • No parallel benchmarking allowed

How it compares

It enforces a strict experimental methodology and output format, whereas manual testing often lacks consistent baseline management.

Compared to similar skills

perf-theory-tester side by side with the closest alternatives in the catalog.

SkillInstallsUpdatedSafetyDifficulty
perf-theory-tester (this skill)16moNo flagsIntermediate
chrome-devtools417moReviewIntermediate
performance-profiling66moReviewIntermediate
performance-benchmark34moNo flagsIntermediate

Try saying

Example prompts that trigger this skill in your AI assistant.

Search skills

Search the agent skills registry