What does this skill do, and when should you use it?
This skill evaluates one domain’s trust, identity, link, and citation-authority signals relative to a locked peer cohort. It runs a 40-item CITE audit and records evidence coverage, Unknown states, and verified manipulation or penalty vetoes. A complete score requires the market, entity stage, domain type, peer cohort, and dated evidence; missing context produces NEEDS_INPUT/UNDECIDED instead of a total score. It fits structured SEO/GEO authority reviews, not page-level content-quality audits or backlink-only diagnosis.
Reads one normalized domain, market, entity stage, domain profile, locked peer cohort, and dated link, identity, trust, eminence, and citation evidence; assigns all 40 CITE items an evidence-backed state, Unknown, or authorized N/A; verifies CITE-T03, CITE-T05, and CITE-T09 veto conditions; runs the Python scorer when the verified runtime is available; preserves score, coverage, and confidence; and produces SHIP, FIX, BLOCK, or UNDECIDED outcomes. After explicit authorization, it validates and persists a v3 audit artifact.
- An SEO lead needs a peer-relative CITE baseline for a SaaS domain against same-stage U.S. competitors.
- A GEO team needs to assess whether a website has domain-level evidence supporting citation trust.
- A brand team sees unusual link growth, traffic incoherence, or indexing changes and needs a structured manipulation or penalty review.
- An auditor has backlink, entity, and search-visibility exports and needs a machine-checkable domain authority report.
How do you install this skill?
- Do not emit a numeric score or gate verdict without a locked peer cohort, market, entity stage, and domain type; return NEEDS_INPUT/UNDECIDED instead.
- This review did not execute scripts, fetch data, or verify external citations; deployment should confirm that all referenced files, runtime scripts, and the repository root are reachable.
- Before crawling third-party domains, confirm authorization and respect robots.txt, applicable TDM reservation signals, and site terms; lack of private-console access must not be treated as proof of a penalty.
- Shell / CLI
- Local filesystem
Python 3Git
Install the repository with: git clone https://github.com/aaron-he-zhu/aaron-marketing-skills. In Claude Code, use /plugin marketplace add aaron-he-zhu/aaron-marketing-skills, then /plugin install aaron-marketing@aaron. The README does not document a dedicated single-skill installation command; the skill identifies Claude Code and compatible Agent Skills hosts as compatible.
tmp="$(mktemp -d)"
git clone --depth 1 https://github.com/aaron-he-zhu/aaron-marketing-skills.git "$tmp"
mkdir -p ~/.claude/skills
cp -R "$tmp/seo-geo/evaluate/domain-authority-auditor" ~/.claude/skills/
rm -rf "$tmp"Generated from the source repository and skill path; it copies only this skill's folder. If the author's install steps above differ, follow those first. To scope it to one project, replace ~/.claude/skills with that project's .claude/skills.
How do you use this skill?
Once installed, send your agent any of these to trigger it:
- Audit example.com against these five same-stage U.S. SaaS peers. Run the CITE content-publisher profile; mark every private-data gap Unknown.
Use a prompt such as: "Audit example.com against these five same-stage U.S. SaaS peers. Run the CITE content-publisher profile; mark every private-data gap Unknown." Provide the canonical domain, observation date, market, entity stage, domain type, cohort inclusion rules, and evidence window. Without the market, stage, domain type, or peer cohort, the skill should return NEEDS_INPUT/UNDECIDED.
What are this skill's strengths and limitations?
- Uses a 40-item CITE framework covering identity, links, trust, and eminence through peer-relative comparison.
- Explicitly distinguishes Unknown, N/A, estimated, proxy, and user-provided evidence instead of treating missing data as failure.
- Defines positive verification rules for link/traffic incoherence, manipulation networks, and active manual actions or material deindexing.
- Supports structured scoring, validation, coverage, confidence, and permission-controlled v3 artifacts.
- A complete score depends on a locked peer cohort and dated evidence; missing backlink or private-console data leaves items Unknown.
- It does not perform page-level content-quality audits, backlink-only diagnosis, or entity repair; those are routed to other skills.
- The supplied source does not include independent test results for this individual skill or the full contents of its referenced runtime files.
How does this skill compare with similar options?
Side by side with related skills; every score comes from the same FSRS standard.
| Skill | FS score | Stars | Last updated | License |
|---|---|---|---|---|
| Domain Authority Auditor this page | 65 · Recommended | ★ 3k | 13d ago | Apache-2.0 |
| Off-Site & AI Referral Analyzer | 59 · Recommended | ★ 3k | 13d ago | Apache-2.0 |
| Content Quality Auditor | 57 · Use with care | ★ 3k | 13d ago | Apache-2.0 |
| Competitor SEO & GEO Analyst | 51 · Use with care | ★ 3k | 13d ago | Apache-2.0 |
| DataForSEO Live SEO Intelligence | 47 · Use with care | ★ 19k | 23d ago | MIT |
Compared with content-quality-auditor, this skill evaluates domain-level authority and citation trust rather than page-level content quality. Compared with offsite-signal-analyzer, it applies CITE scoring and veto gates instead of performing backlink diagnosis without a gate. entity-registry is the alternative for canonical entity repair.
How did FollowSkills review this skill?
The skill specifies least-scope inputs, permissioned v3 artifact writes, explicit authorization before persistence, and fail-closed handling for unknown evidence, fetched content, robots.txt, penalties, and data-egress risks. The Apache-2.0 license, security policy, zero-dependency design, and no-key-storage statements add supporting safeguards. Deduction: the file does not fully expose runtime permissions, rollback implementation, or sensitive-data classification, and publisher identity is unverified by the FollowSkills enterprise registry.
Inputs, prerequisites, 40-item state handling, Unknown semantics, NOT_SCORED behavior when the runtime is unavailable, validation checkpoints, and failure routing are explicit and internally consistent. Deduction: the workflow depends on several external reference files, scripts, and a deterministic runtime whose reachability is not demonstrated here; no execution reproduction is available, so static calibration keeps this below the 10-point ceiling.
Trigger conditions, non-fit cases, argument hint, required market/stage/domain-type/cohort context, and routing to adjacent auditors are clear. It is keyless by default, targets Claude Code and compatible agent-skill hosts, and includes Chinese metadata and summary text. Deduction: mainland-China reachability of external evidence sources is not established, Chinese-language support is limited mainly to metadata/short labels, and boundary examples are sparse.
The skill has version 18.0.0, Apache-2.0 licensing, a Quick Start, Skill Contract, data-source table, layered procedure, validation checkpoints, persistence limits, references, and Next Best Skill links. The README, license, security policy, version signals, and CI provide shared maintenance evidence. Deduction: the selected skill does not itself identify a maintenance owner, changelog, or update path, and dependency/install failure guidance is distributed across runtime rules rather than consolidated.
The skill turns domain-authority auditing into a locked peer cohort, a 40-item evidence state model, veto verification, and a structured report with actionable follow-up, which is directly useful for reducing manual omissions. Deduction: no real execution result is available, and there is no verified representative output across domain profiles; the example is reference prose rather than proof of correctness, so static calibration limits effectiveness to 7.
The skill defines a fixed CITE framework, 40 stable items, dated/source-backed evidence requirements, Unknown/N/A rules, validation commands, and an example report. Shared repository CI and security/architecture tests provide some corroborating maintenance evidence. Deduction: there is no skill-specific real CI coverage, committed execution artifact, or independent third-party corroboration in the supplied material; the example is author-provided, so static calibration limits verifiability to 5.
Open a dimension to read why it scored that way
Evidence confidence：Low — Mostly static review, author material or a limited demo; useful for discovery, not high-risk decisions.
See the full review method →