Conciseness

The body is dense and assumes Claude's competence — no basic-concept explanations — with instruction-packed lines like 'A partial audit is valid; do not describe it as comprehensive.' It falls short of anchor 5 because of deliberate duplication: the scanner caveat ('It can miss dynamic metadata... Never equate missing regex matches with missing site functionality') and the verification caveat ('implemented, verification incomplete') each appear twice, in the numbered sections and again in Common mistakes, and could be trimmed.

Actionability

For an instruction-only skill the guidance is fully executable: copy-paste-ready script invocations with real flags ('scan_repo.py --root', 'init_audit_folder.py --root ... init --folder docs/seo-audit-shop', 'next-id --folder ... --count 2') — all verified to match the actual scripts — plus a concrete mode-selection table, an explicit audit order, and an itemized field list for each finding (ID, status, priority/reason, confirmation, ...). Specific guidance covers the common audit and implementation cases; anchor 4's 'minor gaps' do not apply.

Workflow Clarity

Clear sequenced workflow (choose mode → map site → evidence/coverage → records → implement → verify) with explicit validation checkpoints: 'Run proportionate documented project checks, wait for results', 'Compare before/after on the affected route', 'plus a negative route', and recorded outcomes ('passed, failed, pending or unavailable'). Validation is present (not merely implied), so the destructive/batch cap at 3 does not apply; conditional handling ('If output cannot be checked, report implemented, verification incomplete') and a checklist reference match anchor 5.

Progressive Disclosure

The body is a clear overview with well-signaled, one-level-deep references, each introduced at its point of use with its purpose ('Read [audit-checklist.md]... For platform-specific claims read [platform-evidence.md]') — and all six referenced files plus both scripts exist. The long reference is explicitly scoped as 'optional background... Do not load it wholesale for routine tasks', and references cross-link only laterally, with no nested chains; this matches anchor 5.

Specificity

Concrete actions are explicit — 'Audit or improve SEO, AEO... and AI search visibility', 'content audits, technical SEO remediation, and answer-engine readiness', 'evidence-backed findings and verified fixes' — going well beyond the 1–2 actions of anchor 3. It stops short of anchor 5 because implementation-level actions (e.g., fixing metadata, generating sitemaps) are implied by 'remediation' rather than itemized.

Completeness

The 'what' is explicit ('Audit or improve SEO, AEO... in websites and codebases') and the 'when' is explicit with concrete trigger phrases ('Use for crawl/indexing problems... or AI search citations'), plus supported modes ('Supports codebase-only, website-only, audit, implementation, and continuation requests'). It clearly matches anchor 5, which requires both answers with concrete trigger phrases; a missing 'Use when/Use for' clause cap does not apply.

Trigger Term Quality

The 'Use for' clause lists comprehensive natural trigger terms users would actually say: 'crawl/indexing problems, crawl budget, indexability, metadata, canonical tags, hreflang, robots.txt, sitemaps, structured data, schema.org markup, Core Web Vitals, rendering, search performance, or AI search citations', including tool/standard names and synonyms (AEO/GEO answer-engine readiness). No anchor-4 gaps identified — coverage of natural phrasing is thorough.

Distinctiveness Conflict Risk

A clear niche (SEO/AEO/GEO and AI search visibility) with distinct triggers like 'canonical tags, hreflang, robots.txt, sitemaps, structured data' that no neighboring skill would claim. Minimal conflict risk — anchor 5, not anchor 4, because there is no meaningful overlap risk with closely related skills.