Generate the fixes that get you cited by AI.

Why these three fixes?

Being cited by an AI engine is the new being-ranked-by-Google. Three technical signals gate it — and they're the three this tool generates:

• llms.txt — a machine-readable summary + priority-page map at your root. An emerging convention rather than a confirmed ranking input (see below), but it costs one file.
• JSON-LD schema — Organization + Person markup makes you a recognizable entity, so models attribute facts to your brand instead of a competitor's.
• robots.txt allowlist — blocking retrieval bots (OAI-SearchBot, PerplexityBot, Claude-SearchBot) makes real-time citation impossible. This block keeps the door open.

Do AI crawlers actually read llms.txt?

Not demonstrably, and we would rather say so than sell you a myth. No major provider has confirmed using it, and Google has said publicly that it does not. Treat llms.txt as a cheap, zero-risk convention — one static file, no downside if adoption arrives later — and not as the thing that will get you cited.

The other two fixes in this kit are the load-bearing ones. The robots.txt allowlist is binary and immediate: if you disallow the retrieval bots, real-time citation is impossible, full stop. The JSON-LD schema is what makes a model attribute a fact to your organisation instead of to whoever else is mentioned nearby. If you only ship one thing from this page, ship the robots allowlist.

Which bots should I allow?

The distinction that matters is retrieval versus training. Retrieval bots — OAI-SearchBot, PerplexityBot, Claude-SearchBot — fetch a page at the moment someone asks a question, and blocking them removes you from the answer entirely. Training crawlers are a separate decision with different trade-offs, and reasonable publishers land on both sides of it. The generated block separates the two so you can choose deliberately rather than by accident.