# qa-execution

qa-execution — a skill that walks personas through a product's public interfaces, records friction and findings, and writes everything into a living QA docs tree.

- skill · `type=skill` · Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, OpenCode, Windsurf (Devin Desktop), Claude.ai · 568 installs · ★ 655 repo stars
- Author: Pedro Nauck
- License: not stated · policy=link-only · description only; the file stays with the author
- Source: https://github.com/pedronauck/skills/tree/HEAD/skills/mine/qa-execution/SKILL.md
- Updated: 2026-10-20
- Security check: automated, 2026-11-16 (text) — no signs of malicious code found.
- HTML: https://affpapa.org/en/nevibecode/skills/qa-execution

## What it does

The skill tests a product the way a person meets it. Each persona enters through a real entry point with its own device, network and locale, then walks a journey to its true end state. No dev-tools shortcuts, no reading code to decide what should happen. A `Pass` means the expected observable was seen, confirmed through an independent read path, and survived a refresh.

The QA docs tree is the skill's memory and its only output location. Before running, it reads `README.md`, scenario files, open bugs and the cycle's charters. The report is created on disk before the first session, with every matrix row set to `Pending`. After each session and each fix, the report updates. The rule is blunt: write back or it didn't happen.

Seven steps hold the process together: resolve tree and scope, build the matrix and report, walk each journey in persona, run tours and edge probes, apply experience lenses, file findings into the bug registry, then run the governed fix loop. Each auto-fix passes the governor's bounds and gets a regression test. Anything else goes to Decisions for a Human with options and a recommendation.

The skill does not bootstrap the tree. If `docs/qa/` does not exist, run `qa-report` first — it owns the tree and its setup. Legs that only a human can complete, like real payment or external OAuth, get marked `Blocked (needs human verify)` with exact instructions, never faked.

## Install

### Claude Code (native platform)

1. **Skills CLI (Vercel)**
```bash
npx skills add pedronauck/skills --skill qa-execution -a claude-code
```

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

### Codex

1. **Skills CLI (Vercel)**
```bash
npx skills add pedronauck/skills --skill qa-execution -a codex
```

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

### Cursor

1. **Skills CLI (Vercel)**
```bash
npx skills add pedronauck/skills --skill qa-execution -a cursor
```

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

### GitHub Copilot

1. **GitHub CLI**
```bash
gh skill install pedronauck/skills skills/mine/qa-execution
```

2. **Manually: put the files in place**
- File: `.github/skills/qa-execution/SKILL.md`

### Gemini CLI

1. **Skills CLI (Vercel)**
```bash
npx skills add pedronauck/skills --skill qa-execution -a gemini-cli
```

2. **Platform CLI**
```bash
gemini skills install https://github.com/pedronauck/skills --path skills/mine/qa-execution
```

### Windsurf (Devin Desktop)

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

### OpenCode

**Skills CLI (Vercel)**
```bash
npx skills add pedronauck/skills --skill qa-execution -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/pedronauck/skills/tree/HEAD/skills/mine/qa-execution/SKILL.md

## When to use it
- walk a product through its public interfaces the way a real user would
- catch paper cuts that functional checks never flag
- maintain a QA docs tree with reports, bugs and scenarios
- decide which fixes are auto-applied and which need a human decision
- reuse existing evidence for behavior that has not changed

## Example prompts
- Run the QA pass for the branch I just pushed and update the report under `docs/qa/`
- Walk the `checkout-smoke` charter as a persona on a slow mobile network
- Build the release-run matrix and mark every row `Pending` before the first session

## Similar
- [dogfood-cli](https://affpapa.org/en/nevibecode/skills/dogfood-cli.md) — skill · 120 installs
- [bmad-agent-dev](https://affpapa.org/en/nevibecode/skills/bmad-agent-dev.md) — skill · 17700 installs
- [distributed-task-execution](https://affpapa.org/en/nevibecode/skills/distributed-task-execution.md) — skill · 134 installs
- [qe-test-execution](https://affpapa.org/en/nevibecode/skills/qe-test-execution.md) — skill · 122 installs
- [firecrawl-qa](https://affpapa.org/en/nevibecode/skills/firecrawl-qa.md) — skill · 31700 installs
- [create-verification-skill](https://affpapa.org/en/nevibecode/skills/create-verification-skill.md) — skill · 10400 installs

From the same set: [pedronauck/skills](https://affpapa.org/en/nevibecode/by/pedronauck/skills.md) — 52 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
