How do you know whether AI crawlers actually see your site?

We kept seeing teams jump straight to citation tactics without checking the first constraint: can AI crawlers reach the pages they want cited?

We added Agent Analytics to FixAEO to show which bots—GPTBot, ClaudeBot, PerplexityBot, Googlebot, and others—visit, which pages they read, and how often.

I’m curious how other makers verify this today. Do you trust server logs, Search Console, a crawler tool, or mostly robots.txt checks? What signal would make you confident an AI engine is actually discovering the right pages?