AI & Agent Discovery
Machine-readable descriptions of this site for automated agents. Everything here is public, read-only and served from the same origin.
Discovery documents
- robots.txt — crawler rules, Content-Signal and Agentmap.
- sitemap.xml — canonical public pages.
- llms.txt and llms-full.txt — site overviews.
- API catalog (RFC 9727 linkset) and OpenAPI description.
- MCP server card.
- Agent Skills index and the site-lookup skill.
- ARD catalog.
- DNS-AID index (ILang) and auth.md.
What an agent can actually do
One read-only operation is published: site-lookup. Given a page name or a
conversion direction, it returns the real public URL, title and a short description from
the site's own page table. It does not accept files, does not write anything and does not
require credentials. The same operation is described in the OpenAPI document, the MCP
server card, the skill and the ARD catalog so the descriptions agree.
Hosting limits, stated plainly
This site is static on Cloudflare Pages. There is no server-side application, no POST handler and no per-request content negotiation at the origin. Markdown variants and custom response headers are served as static files and edge headers. DNS-based discovery records are published in DNS; DNSSEC status is verified separately. The OAuth documents describe a planned contract only — authentication is not available on this public site.