Skip to main content
affpapa.org
RU EN
skill CX Codex

perf-benchmarker — Claude Code skill: what it does, how to install

Runs benchmarks sequentially under strict duration rules to catch performance regressions and set baseline metrics.

Install into Claude Code · Skills CLI (Vercel) All methods and platforms ↓ 
npx skills add composio-community/awesome-claude-plugins --skill perf-benchmarker -a claude-code
npx skills add composio-community/awesome-claude-plugins --skill benchmark -a claude-code
Open on GitHub

What perf-benchmarker does

The skill enforces sequential execution for all benchmarks. Parallel test runs are explicitly disallowed to prevent skew. A minimum warmup period of 10s runs before any measurements begin.

Standard runs must take at least 60s per execution. Shorter 30s passes are permitted only during binary search phases. The workflow requires re-running any anomalies, and code changes are prohibited during an active run. All rules follow the contract defined in docs/perf-requirements.md.

Metrics extraction depends on a strict output contract. The benchmarked command must print a JSON block between PERF_METRICS_START and PERF_METRICS_END markers. If your test harness cannot output wrapped JSON, metric parsing fails.

What it helps with

  • measure latency baselines without noise from parallel executions
  • validate performance regressions across sequential runs
  • isolate anomalies by repeating irregular test executions
  • extract structured metrics delimited by PERF_METRICS markers

Example prompts for the agent

  • Run sequential benchmarks for npm run bench with 60s duration
  • Check for regression using python bench.py with 10s warmup
  • Perform binary search benchmark using cargo bench with 30s runs

How to install perf-benchmarker

1Skills CLI (Vercel)

npx skills add composio-community/awesome-claude-plugins --skill perf-benchmarker -a claude-code
npx skills add composio-community/awesome-claude-plugins --skill benchmark -a claude-code

2Manually: put the files in place

~/.claude/skills/benchmark/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/benchmark/

Step by step, with commands for each platform: How to install skills in Claude Code: five ways

Questions about perf-benchmarker

How to install perf-benchmarker in Claude Code?

Run npx skills add composio-community/awesome-claude-plugins --skill perf-benchmarker -a claude-code in your terminal. You can also place the configuration file at ~/.claude/skills/benchmark/SKILL.md.

What duration constraints does perf-benchmarker enforce?

Runs must last at least 60s, though 30s is permitted for binary search routines. Every benchmark requires at least 10s of warmup before metric recording starts.

How must benchmark results be formatted?

Commands must output JSON metrics enclosed between PERF_METRICS_START and PERF_METRICS_END. The summary records the command, run duration, warmup time, and rerun notes.

Similar skills

All skills for Testing and QA →
rag-perf NVIDIA RAG Runs a profiling pass and an aiperf load test against a deployed NVIDIA RAG Blueprint server, all driven by one YAML config. CX Codex +1 2,200installs llm-evaluation Seth Hobson A skill for evaluating LLM applications: automated metrics, human feedback, LLM-as-judge and benchmarking. Useful when comparing models or catching regressions before deploy. CX Codex 12,755installs web-perf cloudflare Audit Core Web Vitals, diagnose rendering bottlenecks, and inspect network dependencies directly through your agent. CX Codex +1 110,200installs golang-benchmark samber This skill helps write and run Go benchmarks, inspect profiles, and compare measurements. Use it when tracking regressions or investigating a hot path. CX Codex +1 39,500installs tilegym-improve-cutile-kernel-perf TileGym Team Systematically profiles, tunes, and benchmarks cuTile GPU kernels in TileGym. Runs iterations on real hardware to lower latency while preserving accuracy. CX Codex +1 1,600installs principle-explain-the-number backnotprop A skill for the moment before you trust a measured number: name what caps it, then rule out that the test measured something other than your work. CX Codex +1 4,100installs

From the same set8

perf-investigation-logger Appends structured performance investigation notes, quotes, and evidence to project logs. Keeps benchmarks and optimization decisions organized. CX Codex +1 1,600installs perf-theory-tester A skill that keeps perf experiments honest: one change per hypothesis, repeated runs, a recorded verdict. Useful when you need proof that an optimization helped. CX Codex +1 1,500installs perf-profiler Runs profilers against CPU and memory hot paths, captures flame graphs, and records file and line evidence. CX Codex +1 1,600installs perf-code-paths This skill maps likely execution paths, entrypoints, and related files for a performance scenario before profiling. Useful when the hot area is unclear. CX Codex +1 1,600installs perf-theory-gatherer A skill that turns git history and code evidence into up to five hypotheses about a performance problem. CX Codex +1 1,600installs perf-analyzer Turns profiling and experiment results into evidence-backed performance recommendations. Use it when a decision needs a traceable basis. CX Codex +1 1,600installs changelog-generator Turns git history into a user-facing changelog: sorts commits into categories and rewrites developer messages for customers. CX Codex +1 1,500installs

Project sponsors

Companies that keep this analytics open
Gold sponsor
Nutra.Club

Impressions, clicks and code copies — open statistics. We keep the numbers public for the same reason NeBlask itself is public. Become a sponsor