Project archive
The earlier experiment
This project began as an eleven-file proposal for publishing business information to AI systems. We no longer recommend that setup. These resources remain available for existing links and historical reference.
The current guide evaluates established formats and optional proposals. It is not a reduced edition of this specification. Read what “discovery files” means today.
Original proposal
Earlier implementation notes
Archived authoring resources
Retired site files
We have removed this site's experimental discovery files: ai.txt, ai.json, identity.json, brand.txt, robots-ai.txt, faq-ai.txt, developer-ai.txt and llms.html. Their old URLs redirect here.
Current information lives on our pages, with structured metadata describing the site. We retain an optional llms.txt index, robots.txt and the sitemap. Original examples remain in the template archive.
The “Required” and “Recommended” labels in the archive belong to the original proposal. They are not requirements imposed by AI providers. Raw templates preserve their original content. Archived authoring resources include scoped guidance for optional work.
The independent llms.txt proposal is maintained upstream. Category naming and early file definitions credited Mark McNeece, AI Visibility (CC BY 4.0). See license and attribution.