agentleFS
Sign inStart free
merllinsbeardmerllinsbeard/merlin-skills/skills/qa/SKILL.md

gstack-qa

Skill1 starsChanged 6 months agoMITView on GitHub

Merlin-adapted QA gate for spec-kit/create-goal web coding. Use after implementation or before ship to verify behavior, run tests, collect browser proof, and fix in-scope bugs.

2 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

9 rules. About 0 are common, 6 are distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈6

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Run the smallest rigorous QA loop for the active spec/goal.

    line 9
  2. Very specific:

    Use after implementation, before gstack-ship, or when the user asks whether the web app works.

    line 13
  3. Very specific:

    Do not use as a replacement for tdd during implementation.

    line 15
  4. Very specific:

    If the spec or goal is missing but the user wants autonomous QA for a feature, ask to run create-goal first unless the user explicitly says …

    line 27
  5. Distinctive:

    Run the repo's relevant automated checks.

    line 32
  6. Distinctive:

    Do not expand product scope during QA.

    line 36
  7. Distinctive:

    Stop when all acceptance checks pass or when a blocker requires external input.

    line 37
  8. Distinctive:

    Write or update a compact QA note next to the active feature when there is a spec directory:

    line 41
  9. Distinctive:

    If there is no spec directory, report inline or use the repo's existing QA/report location.

    line 47

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

gstack-qa

Run the smallest rigorous QA loop for the active spec/goal. This skill is adapted from gstack QA, but it is standalone: do not call gstack runtime helpers, telemetry, browser daemons, review specialists, or gbrain.

When To Use

Use after implementation, before gstack-ship, or when the user asks whether the web app works.

Do not use as a replacement for tdd during implementation. If there is no implemented surface yet, route back to tdd or create-goal.

Inputs

Inspect, in order:

  1. User's newest request and stated QA scope.
  2. specs/<feature>/GOAL.md, then tasks.md, plan.md, and spec.md when present.
  3. Existing test commands in package.json, Makefile, CI config, README, AGENTS.md, or repo-local docs.
  4. Current git diff and recently changed files.
  5. Existing QA reports, screenshots, Playwright traces, or release notes.

If the spec or goal is missing but the user wants autonomous QA for a feature, ask to run create-goal first unless the user explicitly says to QA the current diff only.

QA Loop

  1. Define the acceptance checklist from the spec/goal. Keep it short and concrete.
  2. Run the repo's relevant automated checks. Prefer existing scripts over inventing new commands.
  3. For web behavior, use playwright-cli first. Use playwright-skill only when a reusable script is more appropriate.
  4. Classify each finding as blocker, functional bug, regression risk, visual issue, or test gap.
  5. Fix only in-scope bugs when the user asked for QA-and-fix. Keep fixes small and rerun the failing check after each fix.
  6. Do not expand product scope during QA. If the spec is wrong or incomplete, report the spec gap instead of silently changing behavior.
  7. Stop when all acceptance checks pass or when a blocker requires external input.

Evidence

Write or update a compact QA note next to the active feature when there is a spec directory:

specs/<feature>/qa-report.md

If there is no spec directory, report inline or use the repo's existing QA/report location.

The report should include:

  • scope tested;
  • commands run;
  • browser flows checked;
  • bugs found and fixed;
  • remaining risks or missing coverage;
  • final ship recommendation.

Output

End with:

QA Result: pass | pass-with-risks | blocked | fail
Evidence: <tests/browser/report paths>
Ship readiness: yes | no | needs user decision

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free