AI visibility starts with the boring question: can the crawler actually fetch the page? The answer is now yes for every documented AI crawler.
What happened
A crawler-accessibility audit tested every documented AI and search crawler against the site's key routes: OpenAI (OAI-SearchBot, GPTBot, ChatGPT-User), Anthropic (ClaudeBot, Claude-User, Claude-SearchBot), Perplexity (PerplexityBot, Perplexity-User), Google (Googlebot, Google-Extended, GoogleOther, Google-Agent), Microsoft (Bingbot), Apple (Applebot, Applebot-Extended), and Meta (meta-webindexer, meta-externalagent, meta-externalfetcher).
One significant finding: Applebot - the actual Apple search crawler - was missing from robots.txt entirely. Only its training opt-out (Applebot-Extended) was configured. Applebot is now explicitly allowed.
Method
The configuration was checked against official crawler documentation fetched live, not against memory or community lists. Each user agent was then verified with real HTTP fetches. xAI/Grok publishes no documented crawler identity - that is a provider limitation, not a site defect.
What this does not prove
Crawler access is necessary, not sufficient. Being fetchable does not mean being cited. The experiment's demonstrated capability claims remain unchanged: none.
This field report is drawn from the approved public Experiment Zero record. Events, metrics, and chronology remain tied to that source record.