agentleFS
nandovejernandovejer/luz/.claude/skills/qa-release/SKILL.md

qa-release

Skill1 starsChanged 15 days agoMITView on GitHub

Run the LUZ release QA checklist: lint, size/strings/exceptions/pack/contrast checks, unit tests and Playwright (chromium and chromium-nojs), then write an evidence report and list what needs manual or browser review. Use before a release or when asked for "QA", "release check" or "evidence". Never runs firefox or webkit.

2 other projects use a version of this.

Adopted by

Most recently taken up by

  • Other versions2 projects use a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 2

    1. code23/agentic-qa-toolkit★ 2same name
    2. Myriambs/argus-qa★ 0same name

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

  • In this repositoryOne other file it gives its agents.

The file

We split this skill into the rules it gives an agent

1 rule. About 0 are common, 0 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Run the firefox or webkit Playwright projects, or launch firefox-reviewer or safari-reviewer.

    line 35

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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

QA release

Runs the same gates as CI plus the Playwright suites, and writes an evidence report. It decides nothing about browsers or assistive technology: it lists them.

Do

  1. Run the runner from the repo root (it takes several minutes; run it in the background and wait):

    node scripts/qa-release.js            # full, for a release
    node scripts/qa-release.js --fast     # packages and site already built: no build
    node scripts/qa-release.js --skip-ui  # no Playwright

    It runs, in order: pnpm pages:build (the packages and the Pages site), lint:css, lint:js, check:size, check:strings, check:prompts, check:exceptions, check:license, check:pack, check:contrast, check:ai, test:unit, and Playwright with --project=chromium --project=chromium-nojs (3 workers: more can overload the dev server and give false failures).

  2. Read the report, test-results/qa-release/report-<timestamp>.md. Do not paste the whole file back: give the owner the table and the list of failures.
  3. On a failure, re-run only that step to tell a real failure from a load failure (a Playwright test that fails under load and passes alone is flaky: say so, do not hide it). Fix real failures, or say what blocks them. Never raise a ceiling in budgets.json and never update a visual snapshot to make the suite pass.
  4. Close with the manual list, from the report's "Needs a person" section:
    • the open rows of the pending list in Claude's memory (luz-browser-review-pending.md): the security audit gives no OK while one is open (OWNER.md §4);
    • the components changed since the last tag, which need SR-6 testing: one tests/manual/SR-6-evidence-template.md copy each;
    • changesets waiting, uncommitted changes, changed snapshots.
  5. Legal audit, for a release to npm (not for a QA run that ends in no release): launch the legal-auditor agent (read-only, Sonnet 5.5; call it without model). If .claude/legal/ holds a previous report, hand it the latest so it says what is fixed, open and new. Save its report as .claude/legal/<YYYY-MM-DD>.md (private, git-ignored) and give the owner the table of findings, with what it marks "consultar abogado" listed apart. It is advisory: it never blocks the release and the skill never applies a fix; the owner chooses which to apply (texts to docs-writer, code to frontend-engineer, LICENSE, notices and changesets to chore-runner). Add one line to the evidence report: the date of the legal report and its count of Critical and High findings.

Never

  • Run the firefox or webkit Playwright projects, or launch firefox-reviewer or safari-reviewer. Report the open rows of the pending list as blocking and ask the owner for permission (OWNER.md §3); do not add rows yourself, frontend-engineer does.
  • Apply a legal-auditor proposal, or treat "consultar abogado" text as decided.
  • Say a release is ready. Say which steps passed, which failed, and what still needs a person.

Also useful

  • node scripts/smoke-consumer.js packs the three packages, installs them in an empty project and loads a page from the installed files and from jsDelivr-style paths, offline.
  • pnpm exec playwright test tests/visual.spec.js --project=chromium is the visual regression of the component demos; --update-snapshots only for an intended change, and look at every changed PNG.

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.