SE Ranking AI Visibility Analyst

The skill requires an explicitly configured SERANKING_API_KEY and instructs the caller to check for it before use; unit costs are disclosed. Deductions apply because it does not explain key transmission, third-party data flows or retention, request confirmation, revocation, rollback, or sensitive-data handling for URLs and brand data. The installer and external API create supply-chain and external-access exposure, but the supplied material shows no red-line behavior.

Routing, prerequisites, and core output fields are described, while repository CI, dependency bounds, and security policy provide limited shared-context support. Deductions apply because the selected skill lacks endpoint details, request parameters, authentication-failure handling, quota handling, retry behavior, and diagnosable error feedback. Static review cannot establish key-path execution, so the score remains at or below the static cap of 10.

The scenario is clear: AI visibility, SERP, backlink, and competitor queries, with delegation guidance to related SEO skills. Deductions apply because input formats, output contracts, language/platform support, non-fit boundaries, and precise trigger conditions are absent. Core functionality depends entirely on the SE Ranking API, with no evidence of reachability from mainland-China networks.

The skill provides a name, version 2.2.0, compatibility requirement, prerequisites, command table, cost note, and delegation guidance; repository context additionally shows an MIT license, maintainer/security policy, version metadata, and CI. Deductions apply because this SKILL.md lacks examples, complete installation instructions, parameter definitions, troubleshooting, known limitations, changelog details, and clear extension ownership/update routing. The reference to a dataforseo cost script also creates potential maintenance ambiguity.

The four query types and AI Share-of-Voice fields are directly specified and could produce useful percentages with confidence notes, reducing manual cross-platform aggregation. Deductions apply because there is no representative real output, request/response example, or statically verifiable result, and API dependency, sampling methodology, platform coverage, and approximately five-unit cost may constrain practical value. User review of data and interpretation is still required, so the static score is capped at 7.

The README, dependency file, CI, and security policy provide auditable peripheral evidence, and the skill defines its output fields. Deductions apply because the claims of five-platform coverage, Profound/Peec-like scoring logic, and superior coverage are unsupported within the selected skill by third-party sources, committed tests, fixed samples, or reproducible requests. Evidence is therefore limited and remains at or below the static cap of 5.