Backlink Profile Analyzer

The skill requires source detection and distinguishes free, optional, and premium APIs; it also limits the no-DataForSEO fallback to Common Crawl and documents capability limits for new/lost links. Points are deducted because it does not fully disclose URL, credential, and crawl-data flows, least-privilege requirements, user confirmation, credential isolation, or recovery. The verifier and external APIs issue requests to arbitrary targets and third-party services.

The document provides a source fallback chain, an error table, rate-limit handling, and a data-sufficiency gate, making the happy path reasonably clear. Points are deducted because there are no dedicated tests or reproducible output evidence for this skill's key scripts; failures involving missing scripts, MCP tools, or shared references are not sufficiently diagnosable. The mandatory pre-delivery checklist also includes unrelated schema, H1, and platform checks. Static assessment caps this at 10.

Trigger phrases, commands, URL inputs, seven analysis sections, and no-data behavior are reasonably specified for single-site, competitor-gap, toxic-link, and verification scenarios. Points are deducted because non-fit boundaries, Chinese-language output, and mainland-China reachability are not addressed; the core free data path depends on Common Crawl, while API and verification access may depend on overseas connectivity.

Front matter supplies the name, description, author, version, category, license, and compatibility; command references, analysis framework, error handling, fallback strategy, and on-demand references are documented. MIT licensing and repository maintenance context are present. Points are deducted because installation, relative script paths, MCP prerequisites, parameter contracts, changelog information, and maintenance responsibility for this sub-skill are not fully specified, and the output is primarily Markdown rather than a stable machine-readable interface.

The skill clearly requests seven analysis sections, a health score, prioritized issues, and link-building opportunities; the data-sufficiency gate and Common Crawl score cap reduce some risk of misleading conclusions. Points are deducted because the files provide no verified representative output for the backlink key paths; toxic-link detection, TLD/country judgments, and anchor benchmarks may require substantial manual review, and results depend on configured or third-party data sources. Static assessment caps this at 7.

The document names concrete scripts and data sources, specifies confidence weights and source labels, and includes fact-checking requirements, providing useful audit hooks. Points are deducted because key SEO benchmarks, Penguin-risk claims, TLD authority assumptions, and toxic patterns lack verifiable primary citations in the selected file; the supplied repository CI and tests mainly cover shared security and other modules, not core backlink-analysis results. Static assessment caps this at 5.