Conciseness

The body is lean and assumes competence: no explanation of what canonical tags, schema, or WCAG are; almost every token is snippet, criterion, or skill-specific judgment. Not a 5 because of real duplication: the smoke and standard snippets repeat the same fields (title, titleLen, canonical, h1Count, missingAlts, brokenImages, schema), and the 'Specific QA snippets' section (image audit, heading hierarchy, external link audit) re-covers logic already present in the tier snippets — several dozen tokens that could be consolidated.

Actionability

Fully executable copy-paste console snippets for every tier and check, with numeric pass criteria (title 30-60 chars, meta description 120-160, Lighthouse above 90, INP under 200ms) and a concrete output artifact. The snippets are complete and runnable as written, matching the anchor-5 'copy-paste ready, covers common cases' standard.

Workflow Clarity

The 6-step workflow is clearly sequenced with an explicit validation gate ('If any of these fail, do not proceed with deeper testing until the smoke issue is fixed'), a remediation loop ('Each failure either gets fixed before ship or filed as a known issue'), and checklists via the report template's sign-off section. This matches the anchor-5 pattern of explicit validation steps, error-recovery loops, and checklists; QA is non-destructive so the missing-validation cap does not apply.

Progressive Disclosure

One reference file (`references/qa-report-template.md`) exists on disk, is linked twice in the body plus listed in a dedicated 'Reference files' section, and is one level deep with no nesting — the report template is appropriately externalized. Not a 5 because the ~300-line body inlines content that arguably belongs in references: the 'Specific QA snippets' section (contrast spot-check, form audit, external link audit) is Full-tier-only material that could be split out, leaving the SKILL.md as a leaner overview.

Specificity

Names the domain ('Run QA testing on a page, feature, or full site at one of three depth tiers (smoke, standard, full)') and several concrete actions (run a smoke test after a deploy, verify a page before launch, run a regression sweep, check broken links/404s/images), with only minor gaps in coverage. Not a 5 because the core capability is a single umbrella action ('run QA testing') differentiated by tier rather than multiple distinct concrete operations like the anchor-5 example; not a 3 because it goes well beyond 1-2 actions with tier structure, trigger scenarios, and specific failure classes.

Completeness

Explicitly answers both: what ('Run QA testing on a page, feature, or full site at one of three depth tiers') and when ('Use this skill whenever the user asks to QA a page or site, run a smoke test after a deploy, verify a page before launch, or run a regression sweep'). Matches the anchor-5 example pattern exactly; the 4 anchor requires a 'when' that could be more explicit, which is not the case here.

Trigger Term Quality

Comprehensive natural-term coverage: 'QA, QA sweep, smoke test, regression test, post-deploy check, pre-launch check, verify the deploy, test this page, does the page render, broken link, 404, image not loading' — all phrases a user would naturally say, including synonyms (smoke/post-deploy check/regression sweep). Not a 4 because no common variation is missing; the anchor-5 'file extensions' element is inapplicable to a QA workflow skill, and synonym coverage is complete.

Distinctiveness Conflict Risk

Clear niche (surface-level pre-launch/regression QA) with explicit conflict avoidance: 'deep audits belong to `accessibility-audit`, `performance-optimization`, and `seo-technical`, and code-level debugging to `code-review-web`'. Active de-scoping of adjacent skills minimizes wrong-skill triggering; no overlap concern remains.