A single-page SEO analysis is a focused review of one URL across content, metadata, technical signals, structured data, images, and user experience. Claude SEO evaluates six dimensions in one command and returns a score card with prioritized findings for that page.
The command reviews one reachable page and keeps observed HTML separate from optional provider evidence and proposed changes.
OUTCOME
A single-page diagnostic across on-page, content, technical, schema, image, and performance-risk signals.
REQUIRED INPUT
One page address supplied to /seo page <url>.
READINESS
The page must be reachable. Client-rendered pages may need a browser snapshot. DataForSEO enrichment is optional and paid.
LOCAL OUTPUT
A local scorecard, supporting evidence, prioritized findings, limitations, and review-ready schema drafts where applicable.
SIDE EFFECTS
Fetches the page, can call an approved provider, and writes local guidance. It does not edit the target page.
COST + APPROVAL
Estimate and approve paid enrichment. Verify every generated value and obtain owner approval before publishing markup or page changes.
// HOW IT WORKS
What does a single-page analysis cover?
When you run /seo page, Claude SEO fetches the URL and analyzes it across 6 dimensions. Unlike the full-site audit, which can delegate to up to 17 specialists, this skill focuses deeply on a single page with granular checks for every element.
The output is a score card with an overall 0-100 score, 5 sub-scores, a prioritized issues list, and draft JSON-LD for detected opportunities. Verify every value, fill any placeholder, and validate the markup before publishing. Optional DataForSEO calls can add live SERP and backlink data.
What each dimension covers
Every dimension produces its own sub-score and specific, actionable recommendations. The overall score aggregates all dimensions into a single 0-100 number.
7
ON-PAGE SEO
Reviews whether titles and descriptions are clear and likely to fit common result displays, whether the page has a descriptive main heading and useful hierarchy, and whether URLs and links help users navigate. Character bands are preview heuristics, not ranking rules.
5
CONTENT QUALITY
Reviews topical completeness, readability for the intended audience, repetitive keyword use, observable E-E-A-T signals, sources, and content freshness. It does not enforce universal length, density, or readability targets.
5
TECHNICAL ELEMENTS
Canonical tag (present, self-referencing or correct), meta robots (index/follow unless intentionally blocked), Open Graph (og:title, og:description, og:image, og:url), Twitter Card metadata, and hreflang tags for multi-language sites.
JSON
SCHEMA MARKUP
Detects existing schema formats, checks applicable properties, and drafts JSON-LD for possible additions. It avoids retired Google rich-result types. FAQ rich results are retired for all sites, while QAPage is only for one-question pages where users can submit answers.
5
IMAGE ANALYSIS
Alt text appropriate to context, file size against a selected page budget, format and measured quality, explicit dimensions for layout stability, and lazy loading for suitable below-fold images.
CWV
CORE WEB VITALS
Flags potential LCP issues (huge hero images, render-blocking resources), potential INP issues (heavy JS, no async/defer), and potential CLS issues (missing image dimensions, injected content). Reference indicators from HTML analysis, not lab measurements.
// SCORING
How does the page score card work?
The score card provides an overall 0-100 score with 5 sub-scores displayed as progress bars. Issues are organized by priority from Critical (fix immediately) to Low (backlog). Each issue includes a specific, actionable fix.
Read the score as a diagnostic summary, not a search ranking prediction. Start with critical issues that can block crawling, indexing, or correct rendering, then address high-priority content and metadata gaps. Review every sub-score independently because a strong overall result can still hide a weak schema or image dimension. Re-run the same URL after making changes to confirm which checks improved. The report keeps visible on-page evidence beside each recommendation, so you can distinguish detected markup, measured page elements, and optional data-provider results before deciding what to implement.
ON-PAGE SEO
30%
CONTENT QUALITY
25%
TECHNICAL
20%
SCHEMA
15%
IMAGES
10%
Output includes
Overall Score - 0-100 with visual progress bars per dimension
Issues Found - organized by priority: Critical, High, Medium, Low
Recommendations - specific, actionable improvements with expected impact
Schema Suggestions - review-ready JSON-LD drafts for detected opportunities
Optional DataForSEO integration
If DataForSEO MCP tools are available, the analysis can fetch live SERP positions via serp_organic_live_advanced and backlink data via backlinks_summary. These calls require the user's credentials, consume paid credits, and should run only after a cost estimate and confirmation. The core analysis works without DataForSEO.
Claude SEO fetches the page, parses the HTML, and runs all 6 dimensions of analysis. You get a score card with sub-scores, a prioritized issues list, and review-ready schema drafts that still require value verification and validation.
When to use page vs audit
Use /seo page for deep analysis of a single URL, such as a key landing page, blog post, or product page. Use /seo audit for site-wide analysis with up to 17 specialists. The page skill goes deeper on one URL; the audit skill goes wider across your entire site.
// FAQ
QUESTIONS ABOUT PAGE ANALYSIS
It reviews six areas: titles, descriptions, headings, URLs, and links; content completeness, readability, repetition, evidence, and freshness; canonical, robots, social, and hreflang tags; structured data; image delivery and alternatives; and potential Core Web Vitals issues. Its character bands and scores are review heuristics, not Google thresholds.
The page card combines internal sub-scores for on-page elements, content quality, technical metadata, structured data, and images. The weights organize review priorities. They are not a Google score, ranking probability, traffic forecast, or substitute for inspecting the supporting evidence.
The analysis checks for Experience, Expertise, Authoritativeness, and Trustworthiness signals per the September 2025 Quality Rater Guidelines. It looks for author bio, credentials, first-hand experience markers, publication and last-updated dates, and external links to authoritative sources.
Yes. It detects existing schema formats, checks applicable properties, and identifies possible additions. It does not recommend retired Google rich-result types. FAQ rich results are retired for all sites; existing FAQPage can remain for non-Google semantics, while QAPage is only for one-question pages where users can submit answers.
The analysis works on the HTML returned by the server. For JavaScript-rendered content (empty body in HTML), it notes that key content may be rendered client-side, analyzes the available HTML, and flags that results may be incomplete. It suggests using a browser-rendered snapshot if available.
// RELATED SKILLS
EXPLORE MORE
FULL SITE AUDIT
The full-site audit delegates to up to 17 specialists and combines 7 weighted categories in one prioritized action plan.
