Build a spec-compliant llms.txt file — give LLMs a curated map of your site's best content
/llms.txt at your domain root.# Realm by Rook > Realm by Rook builds intelligence systems for the generative era — AI agents, domain strategy, and experience design for ambitious brands. Pricing and availability are in /pricing. For engagement inquiries, use /contact — we typically respond within one business day. ## Docs - [What we do](https://realmrook.com/services): Overview of engagement types - [Utilities hub](https://realmrook.com/utilities): 31+ free tools ## Optional - [Press](https://realmrook.com/resources/press-releases)
llms.txt is the emerging standard (llmstxt.org, 2024) for websites to give LLMs a curated, high-signal map of their content — distinct from robots.txt (access control) and sitemap.xml (exhaustive URL list). Early adopters get mentioned by Perplexity, ChatGPT Search, and Anthropic's Claude more reliably. This tool builds a spec-compliant file in minutes.
A proposed standard (published by Jeremy Howard / Answer.AI in Sep 2024 at llmstxt.org) that sits at yourdomain.com/llms.txt and gives LLMs a concise, curated map of your site's highest-value content. Think of it as robots.txt + sitemap.xml, but for LLMs.
Adoption is early but growing. Anthropic's Claude can follow llms.txt links when browsing; Perplexity has expressed support; several documentation search tools use it. Being an early adopter is a cheap SEO/AEO bet.
At the root of your domain: https://yoursite.com/llms.txt (plain text / markdown, served with text/plain or text/markdown Content-Type).
llms.txt is a curated index (like a sitemap). llms-full.txt is the full document text for LLM training/retrieval — much bigger. Ship llms.txt first; add llms-full.txt if you want RAG systems to ingest you.
No. sitemap.xml is for search engines discovering every crawlable URL. llms.txt is for LLMs getting a human-curated summary of the most important pages. Use both.