Every check, with its sources.
67 automated checks and 9 manual checks, each graded by evidence.
Evidence grades
- Official: The vendor's own documentation or statements.
- Observed: Independent measurement or reporting.
- Speculative: Inference, including the research team's own.
1 · Eligible
- <token> is blocked for this URL (Official)
- Brave rule: Googlebot access ⇒ Brave ⇒ Claude (Official)
- Training bots (your choice) (Official)
- User-triggered / ads / agent fetchers disallowed (Official)
- Cloudflare Content Signals present (Speculative)
- HTTP status (Official)
- noindex (meta robots / X-Robots-Tag) (Official)
- Snippet controls (nosnippet / max-snippet) (Official)
- Copilot directives (NOARCHIVE / NOCACHE) (Official)
- data-nosnippet spans (Official)
- HTML size vs 4 MB (Observed)
- Per-bot user-agent probe (WAF/CDN) (Official)
- robots.txt was refused (Official)
- Page content not assessed (Official)
- robots.txt returns a server error (Official)
- Not an HTML page (Official)
- No robots.txt rules found (Official)
- Search & answer bots are allowed (Official)
- robots.txt could not be fetched (Official)
2 · Indexed-readiness
- Readable without JavaScript (Observed)
- Canonical URL (Official)
- Freshness signals (Observed)
- Structured data (reported, not a ranking lever) (Observed)
- XML sitemap with lastmod (Official)
- RSS / Atom feed (Observed)
- IndexNow key file (Official)
- llms.txt (reported, no measured effect) (Observed)
3 · Retrievable / Quotable
- How ChatGPT quick mode reads it (Observed)
- Title is descriptive (Observed)
- Natural-language URL slug (Observed)
- Meta description (Observed)
- Single H1 & heading hierarchy (Observed)
- Headings match sub-questions (Observed)
- Answer-first under question headings (Observed)
- Self-contained passages (Official)
- Concrete facts: numbers, prices, dates (Observed)
- Quotable fact sentences (≤150 chars) (Official)
- Lists and tables (Official)
- Content inside tabs / accordions (Official)
- Page length (Observed)
- No hidden instructions to AI (Official)
- Pricing behind "contact us" (Observed)
4 · Search foundations
- Title fits in results (Official)
- Mobile viewport (Official)
- Images have alt text (Official)
- Images reserve their space (Official)
- Links to your other pages (Official)
- Descriptive anchor text (Official)
- hreflang alternates (Official)
- Open Graph share tags (Official)
- Structured data parses (Official)
- Organization details on the home page (Official)
- HTML is compressed (Official)
- Each page has its own title (Official)
- Each page has its own description (Official)
- No broken links (Official)
- Crawler was rate limited or blocked (Official)
- No broken pages or internal links (Official)
- Links go straight to the final URL (Official)
- Every page is distinct (Official)
- Sitemap pages are linked from the site (Official)
- Sitemap lists only canonical, indexable pages (Official)
- Content is in the HTML (Official)
- Every page has a title and one H1 (Official)
- Pages have enough text to answer something (Speculative)
- Important pages are a few clicks from home (Speculative)
- Crawl finished early (Official)
Manual checks
- Google Search Console: Search generative AI control = Include
- Google Search Console: Generative AI performance report
- Bing Webmaster Tools: verify + AI Performance
- WAF / CDN: allowlist the published bot IP ranges
- Server logs: AI bot hits by purpose
- Analytics: AI referrals + branded-search lift
- Brave Search ranking (Claude)
- If you sell products: feeds
- If you are a local business: Google Business Profile & reviews
Machine-readable: /api/method