When an agent represents a prospective Prover, a client choosing who to engage, or a journalist checking a claim, it should be able to read this Institute and represent it correctly. We have built the site so it can — without scraping guesswork, and without anything important trapped in an image or a script.
Two files anchor it: a plain-language map at /llms.txt and a deliberately welcoming /robots.txt.
What that means in practice
Real headings, nav, main, article and section. No content hidden in pictures.
A structured plain-markdown summary of who we are, the tiers, and the canonical links.
Well-behaved crawlers and AI scouts are explicitly allowed, not inadvertently blocked.
The organisation and the credentials it offers are described in structured data.
Addresses match the sitemap, so a link means the same thing tomorrow.
The same content renders cleanly for a person and for a machine.