What the scan checks, and how it scores
Every check, its weight and its limits. No black box.
What this is and isn't. This scan does not query ChatGPT or any other AI system, and it can't tell you whether an AI tool mentions, ranks or recommends your business. It checks public readiness signals on your own website: robots.txt rules for AI and search crawlers, structured data, whether robots.txt lists your sitemap, crawlability, and whether your name, address and phone number are machine-readable. Results are automated, can be wrong or incomplete, and reflect only what our fetcher saw at that moment. AI companies change how their crawlers and products work without notice. Use the results as a starting checklist, not a guarantee.
Scored checks
| Check | Points | How we judge it |
|---|---|---|
| Homepage loads over HTTPS | 15 | The homepage returns a normal HTML page over HTTPS, after at most 3 redirects. |
| Not marked noindex | 10 | No robots meta tag or X-Robots-Tag header says noindex. We also flag nosnippet. |
| OAI-SearchBot allowed | 15 | Your robots.txt lets OpenAI's search crawler reach your homepage. OpenAI says sites that opt out won't be shown in ChatGPT search answers. |
| Other AI search crawlers allowed | 5 | Claude-SearchBot and PerplexityBot can reach your homepage. |
| Search engine crawlers allowed | 5 | Googlebot and Bingbot can reach your homepage. |
| Sitemap listed in robots.txt | 10 | A Sitemap: line in robots.txt pointing to a same-site sitemap. The quick scan doesn't open the sitemap file itself. |
| JSON-LD structured data | 10 | At least one JSON-LD block on the homepage that parses as valid JSON. |
| Business or organization schema | 10 | An Organization or LocalBusiness-type entity (including subtypes like Plumber or Dentist) with a name. |
| Machine-readable name, address and phone | 10 | An address in schema or an HTML address tag, and a phone number in schema or a tap-to-call link, on the homepage. |
| Title and description | 5 | A page title and a meta description of at least 50 characters. |
| Readable without JavaScript | 5 | At least about 500 characters of text in the first 64 KB of raw HTML. |
Reported but not scored
We also show how your robots.txt treats these crawlers and tokens. Blocking training crawlers is a business choice, so it doesn't change your score. User-requested fetchers may not follow robots.txt at all.
- GPTBot (AI training)
- ChatGPT-User (User-requested fetch)
- ClaudeBot (AI training)
- Claude-User (User-requested fetch)
- Perplexity-User (User-requested fetch)
- Google-Extended (AI training)
- Applebot (Search engine)
- Applebot-Extended (AI training)
- CCBot (AI training / datasets)
How we read robots.txt
We follow the Robots Exclusion Protocol (RFC 9309): we pick the group that names the crawler (or the * group if none does), apply the longest matching rule, and let Allow win a tie. We test your homepage path only, and report "some paths blocked" when other Disallow rules exist. If robots.txt is missing, crawlers treat everything as allowed. If it returns a server error, we report the result as unknown. For Applebot we follow Apple's documented fallback to your Googlebot rules.
What we fetch
- Your homepage, after up to 3 redirects (we read the first 64 KB)
- /robots.txt (the first 32 KB)
That's all. To keep each scan small, we don't open your sitemap file, /llms.txt or inner pages: we check whether robots.txt lists a sitemap, and we look for your name, address and phone on the homepage only.
Each fetch is limited to 8 seconds. We only fetch public websites on the standard ports, and we refuse addresses that point to private or internal networks. Read about our fetcher.
What we can't see
- Whether ChatGPT, Claude, Gemini, Perplexity or any other AI tool mentions or recommends you. We don't ask them.
- Your Google Business Profile, Yelp, Apple or Bing listings.
- Firewall or bot-protection rules that block AI crawlers by network or IP even when robots.txt allows them. If your host or CDN blocks AI bots, our scan may still look fine.
- Content that only appears after JavaScript runs.
Last checked against the vendors' public documentation on November 17, 2026. Crawler names and policies change; follow the source links for the current rules.