PE

perf-benchmarker

This tool runs performance tests sequentially to avoid noise, enforces minimum durations, and requires metrics in a specific JSON format.

Install

mkdir -p .claude/skills/perf-benchmarker && curl -L -o skill.zip "https://agentskills.codes/api/skills/download/3499" && unzip -o skill.zip -d .claude/skills/perf-benchmarker && rm skill.zip

Installs to .claude/skills/perf-benchmarker

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 performance benchmarks, establishing baselines, or validating regressions with sequential runs. Enforces 60s minimum runs (30s only for binary search) and no parallel benchmarks.
195 chars✓ has a “when” trigger
Intermediate

Key capabilities

  • Executes sequential benchmarks for latency
  • Enforces minimum run durations
  • Validates regression results
  • Forces metric output in structured JSON
  • Identifies performance anomalies

How it works

Runs command-line tasks in a single-threaded loop, enforcing cooldown and warmup periods to collect statistically significant timing data.

Inputs & outputs

You give it
Command to benchmark and target duration
You get back
Structured JSON performance metrics block

When to use perf-benchmarker

  • Establishing baseline performance metrics
  • Validating code performance regressions
  • Running sequential API latency tests
  • Verifying system stability under load

About this skill

perf-benchmarker

Run sequential benchmarks with strict duration rules.

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

Required Rules

  • Benchmarks MUST run sequentially (never parallel).
  • Minimum duration: 60s per run (30s only for binary search).
  • Warmup: 10s minimum before measurement.
  • Re-run anomalies.

Output Format

command: <benchmark command>
duration: <seconds>
warmup: <seconds>
results: <metrics summary>
notes: <anomalies or reruns>

Output Contract

Benchmarks MUST emit a JSON metrics block between markers:

PERF_METRICS_START
{"scenarios":{"low":{"latency_ms":120},"high":{"latency_ms":450}}}
PERF_METRICS_END

Constraints

  • No short runs unless binary-search phase.
  • Do not change code while benchmarking.

When not to use it

  • When parallel benchmarks are needed
  • For short, low-overhead code blocks

Limitations

  • Slow execution due to mandatory 60s minimum run time
  • Incompatible with asynchronous benchmarking tools

How it compares

It prohibits parallel execution and enforces specific output contracts, ensuring that benchmark noise is minimized compared to casual testing.

Compared to similar skills

perf-benchmarker side by side with the closest alternatives in the catalog.

SkillInstallsUpdatedSafetyDifficulty
perf-benchmarker (this skill)16moNo flagsIntermediate
chrome-devtools417moReviewIntermediate
code-coverage-with-gcov154moReviewIntermediate
angular-best-practices213moNo flagsAdvanced

Try saying

Example prompts that trigger this skill in your AI assistant.

Search skills

Search the agent skills registry