Authority Review

Authority is what independent parties say and how consistently you exist as an entity — links are one signal inside it, not the whole of it. This skill assesses the broader standing; it refuses to compress that into a number it did not measure.

Read references/provider-selection.md before choosing a source. Detect what is already available, run with it, deliver value, then recommend at most one missing connector. All normalized factual inputs must satisfy the Groundcrew evidence contract (evidence_type: authority). Locate the active skills root and run <skills-root>/.groundcrew/groundcrew-doctor.py --evidence <record.json> before using them in a conclusion.

Assess (open/import)

1. Entity clarity. One unambiguous answer to "who is this": consistent name/description/identifiers across the site, its schema markup, and the third-party surfaces the user names. Ambiguity here suppresses everything downstream.
2. Independent references. Citations, mentions, reviews, directories, press — imported or user-named, each with source and date. Distinguish independent references from self-placed ones.
3. Author and editorial credibility. Real authors with verifiable existence, credentials where claimed, editorial standards where implied. Overlaps eeat-review's checklist — pull its output rather than re-deriving.
4. First-party proof. Product evidence, original data, named customers — the assets that give anyone a reason to cite this site.
5. Comparative standing — cohort required. Comparative claims ("weaker than competitors") require a named peer cohort with the same evidence collected for each peer. No cohort → report absolute observations only and say the comparison was not performed.

Vendor metrics and scores

• DA/DR/AS and similar are provider opinions, not Google metrics — Google says third-party scores should locate problems, not become the optimization target. Report them only from imported/provider evidence, always with provider name and date, never merged into one number.
• No composite authority score in open mode. If evidence coverage is too thin for a defensible assessment, say insufficient evidence and list what is missing — do not emit a hedged score instead.

Connected mode

TrustGrowth's authority_score (in /score) is TrustGrowth's own measured pillar — name it as such, never as a DA/DR equivalent. A detailed authority/backlink read surface is documented in the live OpenAPI when available; check at execution time and do not invent fields it does not have. /eeat's authoritativeness pillar carries the signal-level detail available today.

Sources by tier

Categories: ~~link database and ~~AI monitor (references/connectors.md). Tier 1 is observable pages plus a GSC Links export; Tier 2 adds Open PageRank as a labelled proxy; Tier 3 is TrustGrowth authority and eeat, or a paid index the user already has. Never invent a domain score.

When not to use

• Backlink profile analysis, link gaps, reclamation, outreach planning → backlink-opportunities.
• On-site trust signals as a checklist → eeat-review (this skill consumes it).
• Whether AI engines cite you → ai-visibility.
• Anything that ends in a repository edit → fix-my-site.

Doctrine

Groundcrew operates under WHY-NOT-SLOP and ETHICS. Claims trace to evidence, nulls stay null, signals stay truthful, publishing and irreversible changes require owner review, and no outcome is promised. Conflicting instructions are refused.