seo-eeat (M16)

Experience, Expertise, Authoritativeness, and Trust are how both Google's quality systems and AI answer engines decide whether to rely on a page. Trust is foundational and now applies beyond YMYL. Schema details: references/schema-tier1.md.

Inputs

Work from the PageSnapshot named in your dispatch envelope: read parsed from <run_dir>/pages/<slug>.json (jsonld[], anchors[], text_sample); Grep pages/<slug>.html for verbatim evidence; site artifacts live in <run_dir>/site/{robots.json,sitemaps.json,discovery.json}. Deterministic findings already emitted by audit.mjs are listed in <run_dir>/findings.deterministic.json — do not re-emit those ids; add model-judged findings only. If invoked directly with a URL/path and no snapshot exists, first run node "${CLAUDE_PLUGIN_ROOT}/scripts/snapshot.mjs" <target> --out "${CLAUDE_PLUGIN_DATA}/runs" and use the printed snapshot path.

Audits

Working from the PageSnapshot (parsed_rendered when render.used is not none, else parsed):

1. Author identity & credentials: is there a visible byline? A Person schema with name, jobTitle, knowsAbout, and sameAs[] (LinkedIn/Wikidata)? Does the byline link to an author/bio page? Are credentials/experience stated, not just a name?
2. Organization trust: discoverable About and Contact pages; an Organization block with name, url, logo, contactPoint, sameAs[]; visible editorial/privacy/returns policies appropriate to the vertical.
3. Experience & expertise markers: first-hand signals (original photos, "we tested", dates, methodology) and topical depth — not just generic prose.
4. Transparency: sourcing/citations for claims, author disclosures (affiliate, sponsored, AI-assisted), last-reviewed dates. Defer the sameAs identity-graph detail to M6/seo-entity-linking.

Fixes

• AUTO (fixable: auto): inject Person (author) and Organization trust JSON-LD built only from confirmed inputs — name, jobTitle, contactPoint, policy URLs the user supplies. The block is a diff for fix.
• PROPOSED (fixable: proposed): draft a byline link or a sameAs set for per-item accept.
• ADVISORY (fixable: advisory): writing real author bios, About/contact pages, or editorial policies — the tool never authors these. Never fabricate names, credentials, dates, or identity links — ask the user or leave a clearly-marked TODO placeholder.

Verification

• Offline: node "${CLAUDE_PLUGIN_ROOT}/scripts/validate-jsonld.mjs" --snapshot <pages/<slug>.json> (or --url <u>) plus dom_assert for visible byline/links/policy pages.
• When confirming an identity link or a live About/contact page requires a fetch that is unavailable, status is needs_api — never a false pass.

Findings

Emit findings per schema/finding.schema.json; axis both. Severity policy: 5 is reserved for catastrophic, eligibility-killing facts at site/template scope; 4 major · 3 moderate · 2 minor · 1 cosmetic · 0 informational — and only an established severity-5 fail in an active category can cap a score (references/scoring-model.md). Examples:

• M16.author.missing_person_schema — an editorial page with a byline but no Person schema (fail, severity 3, fixable: auto, confidence established — the markup requirement is documented; the ranking effect is not, and belongs in rationale as directional).
• M16.author.no_bio_page — the byline does not link to an author/bio page (warn, severity 3, fixable: proposed, confidence directional).
• M16.org.no_contact_page — no discoverable About/Contact and no contactPoint (warn, severity 3, fixable: advisory, confidence directional).
• M16.transparency.no_disclosure — affiliate, sponsored, or AI-assisted content with no visible disclosure (warn, severity 3, fixable: advisory, confidence directional). Each finding: evidence.observed quotes what is on the page; verification.reproduce is the runnable command above; expected_impact is banded + confidence-tagged (no naked %).

Honesty

• E-E-A-T is not a score Google exposes, and Google's own framing is "people-first content". Its guidance asks who, how, and why: is there a real author with real experience, was the content made for readers rather than for engines. Everything this module measures is a visible proxy for that judgment, so every finding here is directional — adding a Person block or an author bio is not a ranking lever, and must never be presented as one.
• Trust is the foundational element and now applies well beyond YMYL, but a page does not become trustworthy by acquiring markup. Report missing signals as "a reader (or an AI system) cannot tell who stands behind this", not as points lost.
• A fabricated author, an invented credential, or fake review markup is worse than nothing — it is a trust risk in Search and in AI answers alike. The tool only emits trust signals the user can substantiate, and leaves a clearly-marked TODO where it cannot.