Single-Page SEO Inspector
Audits one URL across on-page SEO, content quality, technical metadata, schema, images, and performance signals.
The skill explicitly requires factual handling of unreachable, authenticated, and client-rendered pages and forbids guessing, which reduces misleading behavior. However, it does not itself disclose network, credential, or external-API data flows, user confirmation, permission boundaries, or rollback mechanisms, so points are deducted.
It specifies analysis areas and several abnormal-input responses clearly. However, it references an unavailable quality-gates.md, and the selected skill does not reproduce key script paths, dependency availability, or output consistency; under static calibration it remains below 10.
The single-URL SEO review scenario, trigger phrases, input, and main output structure are reasonably clear. Non-fit boundaries, Chinese-language output support, and environment compatibility are not stated, and some optional capabilities depend on external services such as DataForSEO, so points are deducted.
The front matter includes a stable name, MIT license, author, and version, while the README supplies repository-level installation, command, examples, and maintenance context. The selected skill itself lacks independent dependency/install notes, parameter constraints, FAQs, changelog details, and explicit maintenance ownership, and it has an undocumented file dependency.
It covers on-page, content, technical metadata, schema, images, and potential Core Web Vitals issues, with prioritized actionable recommendations. However, Core Web Vitals are explicitly not measurable from HTML alone, results can be incomplete without rendering or external data, and static review cannot verify directly usable outputs, so points are deducted.
Repository CI, tests, methodology, and a sample report provide some auditability. The evidence is nevertheless mostly repository claims and examples, without an independently recorded execution or dedicated key-path tests for seo-page, so only a low static score is justified.
- The skill itself does not state SSRF, redirect, private-network, rate-limit, or robots/authorization boundaries for fetching; verify that the underlying implementation enforces them before using it on untrusted URLs.
- Core Web Vitals are only HTML-based indicators here, not real measurements; client rendering, authentication, or missing DataForSEO data can make findings incomplete.
- The referenced quality-gates.md is not included in the selected skill evidence, reducing reproducibility of its word-count thresholds.
- Claims about FAQPage, HowTo, and rich-result status are time-sensitive and should be rechecked by the maintainer; this assessment is static and uses only the supplied material.
What does this skill do, and when should you use it?
This is the single-page analysis skill inside the claude-seo repository, focused on one URL rather than a whole-site audit. It reviews page elements, content quality, E-E-A-T signals, technical metadata, structured data, images, and possible performance risks. The output is organized as a score card, prioritized findings, actionable recommendations, and schema suggestions. Core Web Vitals are treated as reference signals only because they cannot be measured from HTML alone.
Reads the available content for a supplied URL and checks the title tag, meta description, H1-H6 hierarchy, URL structure, internal and external links, word count, readability, keyword density, E-E-A-T signals, and freshness indicators. It checks canonical, robots, Open Graph, Twitter Card, hreflang, JSON-LD and other Schema.org types. It also reviews image alt text, file size, format, dimensions, and lazy-loading method, then flags HTML clues that may contribute to LCP, INP, or CLS issues. The report contains an overall score, category scores, Critical-to-Low findings, specific recommendations, and ready-to-use JSON-LD suggestions. Optional DataForSEO MCP tools can provide live SERP positions and backlink data.
- An SEO consultant needs a fast, repeatable review of a landing page, article, or About page before presenting findings to a client.
- A content team wants to check headings, metadata, readability, keyword usage, freshness, and trust signals before publishing an article.
- A development team needs a page-level check of canonical tags, robots directives, social metadata, hreflang, and structured data after launch.
- An e-commerce or brand team wants to identify image accessibility, size, format, dimension, and lazy-loading issues on individual pages.
- A site owner needs clear failure handling when a URL is unreachable, protected by authentication, or dependent on client-side rendering.
What are this skill's strengths and limitations?
- Covers the main page-level SEO dimensions: on-page elements, content, technical metadata, schema, images, and potential performance issues.
- Produces a structured score card, prioritized findings, and concrete recommendations.
- Includes explicit handling for deprecated schema guidance, JavaScript lazy-loaders, and client-rendered content.
- MIT-licensed and usable without DataForSEO; the DataForSEO integration is optional.
- It cannot measure real Core Web Vitals from HTML alone; it can only flag potential LCP, INP, and CLS issues.
- Analysis depends on the target URL being reachable and publicly inspectable.
- Pages that load important content after hydration or user interaction may produce incomplete findings.
- The source provides no independent test coverage or cross-platform validation record for this individual skill.
How do you install this skill?
The skill is installed as part of the full Claude SEO plugin from AgriciDaniel/claude-seo. With Claude Code 1.0.33+, run: /plugin marketplace add AgriciDaniel/claude-seo, then /plugin install claude-seo@agricidaniel-claude-seo. Alternatively: git clone --depth 1 https://github.com/AgriciDaniel/claude-seo.git; bash claude-seo/install.sh. The source does not document a separate installation procedure for skills/seo-page/SKILL.md alone.
How do you use this skill?
Provide one public URL or use a trigger such as “analyze this page,” “check page SEO,” “single URL,” or “page analysis.” Through the full plugin, run: /seo page https://example.com/about. If the URL cannot be reached, the skill reports the error instead of guessing. For 401/403 responses, provide rendered HTML or a public URL; for JavaScript-rendered pages, expect a warning that the analysis may be incomplete.
How does this skill compare with similar options?
Compared with a manual page audit, it offers a more repeatable checklist and scoring format. Compared with commercial SEO audit tools, it provides conversational single-page analysis within Claude Code under an MIT license. The source does not claim that it replaces specialist crawlers such as Screaming Frog or Ahrefs.