# performance-benchmark-specialist

Sets up benchmarks for shell tools: warmup runs, statistics, and target checks like <100ms and >90% cache hit rate.

- skill · `type=skill` · Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, OpenCode · 456 installs · ★ 62
- Author: manutej
- License: not stated · policy=link-only · description only; the file stays with the author
- Source: https://github.com/manutej/luxor-claude-marketplace/tree/HEAD/plugins/luxor-design-toolkit/skills/performance-benchmark-specialist/SKILL.md
- Updated: 2026-06-13
- HTML: https://affpapa.org/en/nevibecode/skills/performance-benchmark-specialist

## What it does

The skill lays out a benchmark structure for shell scripts: warmup, a batch of runs, min/max/mean/median/stddev, then a baseline-versus-optimized comparison. Results land in CSV, and target values get asserted automatically. Handy when you changed a script and need to know whether it actually got faster.

Inside is a `bench-helpers.sh` library with millisecond timing, warmup, run and statistics functions, plus a benchmark template. Targets come from the unix-goto project: cached navigation under 100ms, bookmark lookup under 10ms, cache hit rate above 90%, cache build under 5s.

It is not for application profiling, production monitoring or load testing web services — those need their own tools. A one-off measurement is fine with plain `time`. The source text is cut off, so the full set of library functions cannot be verified from it.

## Install

### Claude Code (native platform)

1. **Skills CLI (Vercel)**
```bash
npx skills add manutej/luxor-claude-marketplace --skill performance-benchmark-specialist -a claude-code
```

2. **Manually: put the files in place**
- File: `~/.claude/skills/performance-benchmark-specialist/SKILL.md` — Copy the whole skill folder; for one project use .claude/skills/performance-benchmark-specialist/

### Codex

1. **Skills CLI (Vercel)**
```bash
npx skills add manutej/luxor-claude-marketplace --skill performance-benchmark-specialist -a codex
```

2. **Manually: put the files in place**
- File: `.agents/skills/performance-benchmark-specialist/SKILL.md` — For all projects use ~/.agents/skills/

### Cursor

1. **Skills CLI (Vercel)**
```bash
npx skills add manutej/luxor-claude-marketplace --skill performance-benchmark-specialist -a cursor
```

2. **Manually: put the files in place**
- File: `.cursor/skills/performance-benchmark-specialist/SKILL.md` — Cursor also reads .agents/skills/ and .claude/skills/

### GitHub Copilot

1. **GitHub CLI**
```bash
gh skill install manutej/luxor-claude-marketplace performance-benchmark-specialist
```

2. **Manually: put the files in place**
- File: `.github/skills/performance-benchmark-specialist/SKILL.md`

### Gemini CLI

1. **Skills CLI (Vercel)**
```bash
npx skills add manutej/luxor-claude-marketplace --skill performance-benchmark-specialist -a gemini-cli
```

2. **Platform CLI**
```bash
gemini skills install https://github.com/manutej/luxor-claude-marketplace --path plugins/luxor-design-toolkit/skills/performance-benchmark-specialist
```

### Windsurf (Devin Desktop)

**Manually: put the files in place**
- File: `.devin/skills/performance-benchmark-specialist/SKILL.md` — The legacy .windsurf/skills/ path still works

### OpenCode

**Skills CLI (Vercel)**
```bash
npx skills add manutej/luxor-claude-marketplace --skill performance-benchmark-specialist -a opencode
```

### Claude.ai

**Claude.ai**
- Zip the folder and upload it in Settings → Capabilities → Skills

Where a file is given instead of a command, copy the files from the source to that path: https://github.com/manutej/luxor-claude-marketplace/tree/HEAD/plugins/luxor-design-toolkit/skills/performance-benchmark-specialist/SKILL.md

## When to use it
- measure a shell script before and after optimization
- compute min, max, mean, median and stddev across runs
- check whether a command stays under the 100ms target
- generate a test workspace with folders
- save benchmark results to CSV and compare against baseline

## Example prompts
- Write a benchmark for my navigation script: warmup, 10 runs, statistics and a 100ms target check
- Compare the baseline and optimized versions of my bookmark lookup command and report the speedup
- Create a medium workspace and benchmark the cache, then save the results to CSV

## Similar
- [golang-benchmark](https://affpapa.org/en/nevibecode/skills/golang-benchmark.md) — skill · 39500 installs
- [V3 Performance Optimization](https://affpapa.org/en/nevibecode/skills/v3-performance-optimization.md) — skill · 1100 installs
- [m10-performance](https://affpapa.org/en/nevibecode/skills/m10-performance.md) — skill · 3600 installs
- [go-performance](https://affpapa.org/en/nevibecode/skills/go-performance.md) — skill · 1200 installs
- [bash-scripting](https://affpapa.org/en/nevibecode/skills/sickn33-bash-scripting.md) — skill · 2400 installs
- [cost-benchmark](https://affpapa.org/en/nevibecode/skills/cost-benchmark.md) — skill · 679 installs

From the same set: [manutej/luxor-claude-marketplace](https://affpapa.org/en/nevibecode/by/manutej/luxor-claude-marketplace.md) — 31 cards

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md