qa-skills
Think like a senior QA engineer: a Claude Code skill (/qa-review) and a one-command quality gate that catch bugs before you push.
Safety scan
A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.
Badge for your README
Markdown
[](/listings/mamunr412--qa-skills)HTML
<a href="/listings/mamunr412--qa-skills"><img src="/api/listings/mamunr412--qa-skills/badge.svg" alt="findsafeskills safety rating" /></a>Install
Install instructions from the README
As written by the repo's author. findsafeskills has not verified these steps.
With the skills CLI:
npx skills add mamunr412/qa-skills --skill qa-review
Or copy it by hand. Clone this repo, then copy the whole `skills/qa-review/` folder:
# This project only New-Item -ItemType Directory -Force <your-project>\.claude\skills Copy-Item -Recurse -Force .\skills\qa-review <your-project>\.claude\skills # All projects New-Item -ItemType Directory -Force "$env:USERPROFILE\.claude\skills" Copy-Item -Recurse -Force .\skills\qa-review "$env:USERPROFILE\.claude\skills"
Restart Claude Code, then type `/qa-review`.
The skill ships its own copy of `qa-check.ps1`, so you do not need to copy the script. If you want to run the gate yourself, copy it into your project root.
What was scanned
The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.
README.md · README · 93 lines
- # qa-skills
- Think like a senior QA engineer: a Claude Code skill and a one-command quality gate that catch bugs before you push.
- ## What is inside
- | Part | What it is | Link |
- |---|---|---|
- | `/qa-review` skill | Reviews a feature, module or project like a senior QA engineer and gives a verdict | [skills/qa-review](skills/qa-review/SKILL.md) |
- | `qa-check.ps1` gate | One PowerShell command: format, lint, types, tests, secrets, build, audit | [qa-check.ps1](skills/qa-review/scripts/qa-check.ps1) |
- | QA Playbook | The guide behind both: setup, mindset, checklists, edge cases, test design | [docs/QA-PLAYBOOK.md](docs/QA-PLAYBOOK.md) |
- ## Install
- With the skills CLI:
- ```bash
- npx skills add mamunr412/qa-skills --skill qa-review
- ```
- Or copy it by hand. Clone this repo, then copy the whole `skills/qa-review/` folder:
- ```powershell
- # This project only
- New-Item -ItemType Directory -Force <your-project>\.claude\skills
- Copy-Item -Recurse -Force .\skills\qa-review <your-project>\.claude\skills
- # All projects
- New-Item -ItemType Directory -Force "$env:USERPROFILE\.claude\skills"
- Copy-Item -Recurse -Force .\skills\qa-review "$env:USERPROFILE\.claude\skills"
- ```
- Restart Claude Code, then type `/qa-review`.
- The skill ships its own copy of `qa-check.ps1`, so you do not need to copy the script. If you want to run the gate yourself, copy it into your project root.
- ## Usage
- | Command | What it does |
- |---|---|
- | `/qa-review` | Reviews the changes on the current branch |
- | `/qa-review src/features/billing` | Reviews that module and the code that uses it |
- | `/qa-review 123` | Reviews pull request 123 (diff and description) |
- | `/qa-review <issue-url>` | Reads the ticket and checks its acceptance criteria |
- | `/qa-review --full` | Reviews the whole project or release, with the full gate |
- | `/qa-review --write-tests` | Writes and runs automated tests for every case |
- | `/qa-review --fix` | Fixes confirmed bugs, adds a regression test for each, never commits |
- Run the gate on its own:
- ```powershell
- # Quick check, 1-3 minutes
- powershell -ExecutionPolicy Bypass -File .\qa-check.ps1
- # Before a PR: coverage, build, audit, end-to-end tests
- powershell -ExecutionPolicy Bypass -File .\qa-check.ps1 -Mode full -E2E
- ```
- Results are written to `.qa/qa-report.md`. Add `.qa/` to your `.gitignore`.
- ## What /qa-review does
- - Finds the scope: your branch diff, a path, a PR, a ticket, or a feature description.
- - Treats acceptance criteria as claims. A ticked box is not proof, so each one is checked.
- - Builds a risk map and spends most of the time on high-risk areas.
- - Runs the `qa-check` gate and reads every failing log.
- - Hunts bugs with proven heuristics: boundaries, bad input, state, timing, permissions.
- - Checks CRUD data flow: do the list, the detail page and the counts update after create, edit and delete?
- - Builds a full checklist of positive, negative and edge cases, tests every one, and marks it PASS, FAIL or NOT RUN.
- - Reviews test quality: weak tests, flaky tests, and mutation testing.
- - With `--fix`, fixes confirmed bugs with a regression test each. It never commits.
- - Ends with a verdict: Ship, Ship with fixes, or Block.
- ## Tested on a sample app
- The test app was a small Tasks app with 8 planted bugs: 6 code bugs, 1 weak test and 1 flaky test. Its own tests all passed.
- | Run | Result |
- |---|---|
- | Read-only | Found 8 of 8 bugs, plus 3 extra real issues. Verdict: Block. 30 of 30 cases tested, 15 FAIL. |
- | `--fix` | Fixed all 8 bugs with 22 regression tests. 32 of 32 tests passed, 5 runs in a row. Left 5 product decisions to the author. Nothing committed. |
- Limit: the sample app had no UI, so the browser checks were not exercised.
- ## Requirements
- - Claude Code, or another agent that supports skills.
- - Windows PowerShell 5.1 or PowerShell 7 for the gate. PowerShell 7 is not yet tested.
- - Optional: [gitleaks](https://github.com/gitleaks/gitleaks) for secret scanning, [Stryker](https://stryker-mutator.io/) for mutation testing.
- ## License
- MIT. See [LICENSE](LICENSE).
Ratings and comments
Sign in to rate this skill.
Sign inFree to use. If it’s helped, pass it on.