Learn-skills.dev quality-gates

Auditing, testing, and architecture-governance instructions for Black-Tortoise, covering when to run `architecture:gate`, update docs/AUDIT-*, ship tests, and satisfy `.github/instructions/64-quality-testing-copilot-instructions.md` before merge.

install
source · Clone the upstream repo
git clone https://github.com/NeverSight/learn-skills.dev
Claude Code · Install into ~/.claude/skills/
T=$(mktemp -d) && git clone --depth=1 https://github.com/NeverSight/learn-skills.dev "$T" && mkdir -p ~/.claude/skills && cp -r "$T/data/skills-md/7spade/black-tortoise/quality-gates" ~/.claude/skills/neversight-learn-skills-dev-quality-gates && rm -rf "$T"
manifest: data/skills-md/7spade/black-tortoise/quality-gates/SKILL.md
source content

Quality Gates & Audit Master Guide

Intent

Document the pre-merge checklist: what gates must pass, which audit artefacts to update, and which tests the repo expects, per

.github/instructions/64-quality-testing-copilot-instructions.md
.

Gates to Run

  • Run
    pnpm run architecture:gate
    after any structural change (new dependencies, cross-layer imports, new modules); this enforces the dependency cruiser rules described in
    .github/instructions/61-architectural-governance-copilot-instructions.md
    .
  • Before PR, execute
    pnpm run lint
    ,
    pnpm run build -- --noEmit
    , and the relevant unit suites described in the capability instructions.
  • Ensure Husky/lint-staged hooks triggered by
    git commit
    pass (
    eslint
    and
    tsc --noEmit
    per
    package.json
    ).

Documentation & Audit Tracking

  • Update
    docs/AUDIT-<module>.md
    whenever you change a capability, event, or critical workflow; note invariants, boundaries, and audit findings as specified in
    .github/instructions/64-quality-testing-copilot-instructions.md
    .
  • Document requirements/status in
    requirements.md
    and
    tasks.md
    (status values: not-started, in-progress, completed, blocked) before completion.
  • Capture
    architecture:gate
    failures’ logs and include them in your notes so reviewers can see what changed since the last passing gate.

Testing Expectations

  • Cover every critical path and domain invariant touched by your change with tests (Jasmine + Karma for unit stores/domains, Playwright for happy-path flows).
  • Use Playwright with resilient selectors (
    data-testid
    , roles) and emulator resets so tests are deterministic (
    .github/skills/e2e-playwright
    already describes these patterns).
  • When introducing async or signal transitions, do full Arrange-Act-Assert cycles and clean up subscriptions with
    takeUntilDestroyed()
    or
    rxMethod
    disposal.

When to Apply This Skill

  • Anytime you add/modify tests, instructions, or GA/QA tasks that must mention gating commands, auditors, or docs.
  • Whenever you touch documentation or pipeline scripts so that audit assets remain in sync with the
    docs/AUDIT-*
    expectations.