SEO + AI Visibility Audit Prompt Library (9 Master Prompts) ⚡

Act as a senior technical SEO and search-visibility auditor for this website. This is a READ-ONLY baseline. Do not change or publish anything. Work from the current repository/project and its configured live production deployment. Discover the production URL, site purpose, primary outcomes, audiences, topics, markets/languages, platform/CMS/hosting, available connected data, prior audit artifacts, and recent changes from the repository, configuration, documentation, CMS, deployment tooling, version history, and connected services. Do not ask the user to fill in a setup card or provide a URL. If multiple sites or production domains exist, resolve the active public target from canonical/deployment evidence and state the choice. If anything cannot be verified, label it 'Needs external data,' request the smallest useful export or access in the report, and continue with what can be verified.

EVIDENCE RULES:
- Inspect the live public site; do not treat a preview, successful build, configuration value, or tool score as proof of the live result.
- For every finding, give the exact URL/template/surface, observed evidence, source/tool, date, confidence, and validation method.
- Separate Verified, Inferred, Needs external data, and Not applicable. Never invent metrics, rankings, backlinks, citations, conversions, or competitor facts.
- Compare with the previous audit when supplied and separate new, fixed, persistent, and uncertain findings.

Build a reproducible baseline:
1) CURRENT INVENTORY: Enumerate all current static pages, CMS collections/templates/items, generated routes, localization variants, redirects, and any excluded/private/staging pages. Crawl sitemap and public site. Reconcile inventory vs sitemap vs crawlable URLs.
2) SEARCH FOUNDATION: Verify robots.txt, XML sitemap, HTTP/HTTPS and host normalization, response codes, redirect chains, self-referential canonicals, canonical consistency, indexability, and staging/project-domain exposure.
3) PERFORMANCE AND EXPERIENCE: Inspect field Core Web Vitals (LCP <= 2.5s, INP <= 200ms, CLS <= 0.1), semantic landmarks, keyboard access, form labels, and alt text.
4) PERFORMANCE BASELINE: Record Search Console queries/clicks, analytics organic landing sessions, Bing metrics, and referring domains.
5) CONTENT AND AUTHORITY BASELINE: Map important URLs to user intent, funnel role, topic family, and internal link graph.

OUTPUT: Return direct verdict, data coverage, top 5 actions, evidence table with ID/severity/URL/evidence/action, do-not-touch list, exact data requests, and next checkpoint.

Run a READ-ONLY weekly SEO and AI-visibility pulse for this website. Do not edit or publish anything. Compare with the last saved pulse and use complete data periods only. Work from the current repository/project and its configured live production deployment. Discover the production URL, site purpose, declared outcomes, priority audiences/topics/markets, connected search and analytics services, last saved pulse, recent publishes/changes, and stable AI-query panel. Do not ask the user for setup fields or a URL.

Check:
1) INCIDENTS AND REGRESSIONS: Search Console messages, manual actions, sudden indexing changes, sitemap failures, crawl anomalies, Core Web Vitals regressions, and newly published URLs.
2) SEARCH MOVEMENT: Compare last 7 complete days vs prior 7, plus 28-day rolling trend. Report clicks, impressions, CTR by query and landing page. Surface top 3 opportunities or decay signals.
3) AI DISCOVERY: Check Bing AI citations, referral traffic from ChatGPT, Perplexity, Claude, and test the stable AI query panel.
4) DECISION: If healthy, say 'No urgent SEO action this week'. If unhealthy, name the smallest diagnostic next step.

Return: One-paragraph verdict, data coverage, anomalies, top 3 actions max, exact evidence, and next checkpoint.

Run a READ-ONLY monthly technical SEO audit of the website, its available CMS/source/configuration, and the live public deployment. Do not modify or publish anything. Work from the current repository/project and its configured live production deployment.

Audit and provide URL-level evidence for:
A. DISCOVERY AND INDEXING: robots.txt, XML sitemaps, canonical tags, HTTP status codes, redirect loops, orphan pages, crawl stats.
B. RENDERING AND DOM: Verify primary text and links are present in rendered HTML without requiring client-side JS hydration. Accessible button/link semantics.
C. METADATA AND SEARCH APPEARANCE: Unique titles and meta descriptions, logical heading hierarchy (single H1), Open Graph/Twitter tags, image alt text.
D. STRUCTURED DATA: Validate JSON-LD syntax and visible content fidelity (Organization, Person, Article, FAQPage, BreadcrumbList).
E. PERFORMANCE: Core Web Vitals targets (LCP <= 2.5s, INP <= 200ms, CLS <= 0.1).
F. LOCALIZATION: Hreflang tags, reciprocal links, and language-specific canonicals if applicable.

Return: Direct verdict, data coverage, top 5 actions, evidence table, do-not-touch list, and explicit regression tests.

Run a READ-ONLY content-performance and internal-link audit for this website. Do not edit CMS items, source content, or the live site.

1) MAP DEMAND TO PAGES: Group Search Console queries by user intent and topic. Identify high-impression/low-CTR pages, striking-distance queries (positions 4-20), declining pages, and cannibalization.
2) CONTENT QUALITY: Check for firsthand experience, original research/data, screenshots, examples, author bylines, and transparent limitations. Flag thin, stale, or generic AI copy.
3) INTERNAL-LINK GRAPH: Build internal link graph from crawlable  tags. Find orphans, dead ends, and high-authority pages not linking to priority targets. Propose exact contextual link sentences with natural anchors.
4) TITLES & SNIPPETS: Review search appearance against user intent.

Return: Direct verdict, top 5 actions, evidence table, Protect/Improve/Consolidate queues, proposed internal-link changes, and next checkpoint.