3. AI search

Someone describes your problem to an assistant and gets a shortlist without you on it. A search page has ten slots and a scroll below them, so a weak position can still be found. An assistant’s answer has room for a few sources, and outside them you are absent.

Three pages, in this order.

AI crawlers and llms.txt is about access. Which agents exist, who owns them, where a block actually lives — it is often not in robots.txt at all — and what belongs in llms.txt.

Why AI answers cite someone else is about the writing. Answer first, facts with dates, one checkable claim. Nobody else on the results page has that claim.

Who actually crawls you is my own logs. Which agents came, what they fetched, and how rarely anybody asked for llms.txt. I read my own log first, and only after that did I believe what other people write about access and about getting quoted.

Start with access. If you never let a crawler in, it will not see the page, and then it does not matter how well the page is written.