Skill Claude CodeCodex
Review aggregate search, analytics, tag-management, or UX-console evidence through a user-authorized signed-in browser or user-supplied export when no provider API is configured. Use for read-only GSC, GA4, Metrika, Yandex Webmaster, Bing Webmaster, GTM, or Clarity review; do not use it to collect raw visitor data or…
Skill Claude CodeCodex
Planner: given a domain, performs minimal reconnaissance in 5 minutes, then produces an informed roadmap specifying exactly what to collect and in what order BEFORE running a heavy audit. It does not perform the audit; it scopes it: site size (number of URLs), stack (whether JS rendering or CWV analysis is needed)…
Skill Claude CodeCodex
Check backlinks against a provided list of donor pages: whether the target domain is linked, what anchor text and rel values the link uses, whether it passes link equity (nofollow/ugc/sponsored prevent it), and whether the donor page itself is indexable. It answers not only whether a link exists, but WHY it is missing…
Skill Claude CodeCodex
Finds near-duplicate and thin pages on a site: filtered pagination, copied content, utility duplicates, and weakly differentiated product pages. It works with a list of page texts (from a Screaming Frog crawl, sitemap+parse, or a prepared list) and groups similar pages using simhash + locality-sensitive hashing…
Skill Claude CodeCodex
Audits website visibility in the age of AI-generated answers (GEO/AEO): which AI crawlers robots.txt allows (GPTBot, ClaudeBot, Perplexity, Google-Extended…), whether /llms.txt exists and how useful it is to models, and how citable the content is (answers that an AI is ready to restate with a link). Triggers on: “AI…
Skill Claude CodeCodex
Builds the complete H1–H6 heading structure of each page in DOM order and checks the hierarchy, which Screaming Frog cannot do (SF sees only H1/H2 and catches only "missing/multiple H1"). It parses live HTML and checks for exactly one H1, no skipped levels (H2→H4), a non-empty H1 that does not duplicate the title, and…
Skill Claude CodeCodex
Compares raw HTML (what a bot receives without rendering) with the rendered DOM to show what appears on the page ONLY after JavaScript: text, internal links, title/canonical, h1, and Schema.org. This is critical when content, internal linking, or directives arrive only during rendering: Google renders with a delay…
Skill Claude CodeCodex
Audits a website's regional structure: subdomains (msk.example.com), directories (example.com/msk/), and satellite sites on separate domains. Finds the city switcher, recognizes approximately 100 Russian cities by slug and Russian anchor text, checks every regional version, and catches issues that destroy regional…
Skill Claude CodeCodex
Perform an in-depth analysis of robots.txt for redundant and harmful directives: what is actually blocked, whether rendering is impaired, Allow/Disallow conflicts, overly broad rules, and a missing Sitemap. Cross-check Disallow rules against live 200-status pages from sf-analyzer and against the sitemap, then return a…
Skill Claude CodeCodex
Analyzes and recommends Schema.org structured data for a page: which type describes the content (Article/Product/Service/LocalBusiness/Event/Recipe/FAQPage/...), which connected @graph to build (Organization ← WebSite ← WebPage ← type), and what the current JSON-LD is missing for a rich result. Use when asked to check…
Skill Claude CodeCodex
Audit a site's security headers from an SEO perspective: HSTS, CSP, X-Frame-Options, X-Content-Type-Options, Referrer-Policy, and Permissions-Policy, with a numeric score and an A–F grade. Also check software-version disclosure (Server, X-Powered-By), cookie flags (Secure, HttpOnly, SameSite), the http→https redirect…
Skill Claude CodeCodex
SINGLE ENTRY POINT and orchestrator for a complete domain SEO audit. Triggers when a site/domain is provided with a request to inspect/analyze/check it — WITHOUT a specified scope. By default, it runs the MAXIMUM: reconnaissance (domain/hosting/CMS), crawl evidence from a separately installed licensed Screaming Frog…
Skill Claude CodeCodex
Investigate the domain and site infrastructure using data Screaming Frog does not provide: domain age and registration, DNS records, hosting and ASN by IP, server location, TLS certificate, CDN and cache operation detection, CMS/framework/stack, analytics, and pixels. Use when asked "what hosting provider does this…
Skill Claude CodeCodex
Machine-readable SEO audit from a Screaming Frog crawl. Use when the user asks to analyze a .seospider file or Screaming Frog exports, find broken links and where they appear in the DOM (source, anchor, position, XPath), check a sitemap for stale lastmod values / 3xx-4xx entries / discrepancies with the site, detect…
Skill Claude CodeCodex
A navigation reference explaining what an SEO audit obtains THROUGH Screaming Frog, what SF CANNOT do, and what must be handled by an agent or other tools. It routes each need to the appropriate skill/tool (status codes/meta/canonicals/ duplicates/sitemaps/broken links use sf-analyzer; WHOIS/hosting/CDN/CMS use…
Skill Claude CodeCodex
Configure Screaming Frog ONCE so mode A can maximize applicable coverage from the 96-check registry for any site. Explains how to create audit.seospiderconfig (enable Spelling & Grammar, Structured Data, Security, Store HTML/Rendered HTML, JS Rendering, Crawl Linked XML Sitemaps, and Crawl Analysis), where to place…
Skill Claude CodeCodex
Produces a readable Markdown report from a Screaming Frog export: instead of merely "collecting and fixing," it formats an audit report that people can understand — health score, issues by severity, broken links with DOM location (source/destination/position/XPath), duplicates, heavy pages, and sitemap. Use when asked…
Skill Claude CodeCodex
Builds a prioritized task backlog from a Screaming Frog export or audit.json, with a configurable pipeline (which severities to include, grouping by issue type or URL, P1/P2/P3 priorities, effort estimates, and limits). For broken links, each task includes the location (source/destination/position/XPath). Use when…
Skill Claude CodeCodex
Determines whether a website has a silo architecture (topical clusters, hubs, URL depth, internal linking, and semantic coverage) or is a flat brochure site. Uses an sf-analyzer crawl (audit.json: crawldepth, inlinks, isinsitemap, summary.sitemap) and the reference/silo-architecture.md guide. Produces a verdict of…
Skill Claude CodeCodex
Runs a bulk site audit with one command and produces a ready-made report: Excel (4 worksheets, filters, chart), Word (for the client), CSV (for the tracker), or Markdown. Site-level tools run once, while page-level tools run for every URL in the sitemap; everything is consolidated into one JSON document from which a…
Skill Claude CodeCodex
Analyze what a site is built with: CMS/site builder, server-side framework and runtime, analytics and advertising pixels, widgets and chats, consent management, payment systems, fonts and libraries from third-party CDNs, protection, and CDN hosting. Make one HTTP request, then inspect the HTML, headers, cookies, and…
MCP server Claude CodeCodexCursor +2
MCP server "seohead" as configured in PavloSEO/seohead-seotools. Launched with seohead mcp.
Instructions file CodexOpenCode
AGENTS.md instructions for PavloSEO/seohead-seotools, covering agents.md — public repository contract, product model, start here, architecture and invariants.
Skill Claude CodeCodex
Write expert bilingual articles (RU+EN) in JSON format for a blog. Always use this skill when asked to write an article, create expert material, write a guide or review, prepare a blog post, generate researched content with sources, write an article with code examples or case studies, generate bilingual article JSON…