For AI
This site isn't just built for people clicking through it in a browser — it's built to be legible to AI agents, crawlers, and language models too. Here's exactly how, and what each piece actually does.
XML sitemap
A standard sitemap, regenerated on every build, listing every page on the site so search engines and crawlers can find and index them without guesswork.
robots.txt with Content-Signal
The usual crawler rules, plus a Content-Signal line that gives AI crawlers more specific guidance than a blanket allow or disallow: ai-train=no (don't use this site to train models), search=yes (fine to index for search), ai-input=yes (fine to use as retrieved context when an AI is answering a question about Collecurate).
llms.txt and llms-full.txt
Following the emerging llms.txt convention: llms.txt is a short index of the whole site — page titles, URLs, one-line descriptions — for a language model to skim before deciding what to fetch. llms-full.txt is the same index with the complete text of every page bundled into a single file.
A markdown version of every page
Every page also exists as a clean, chrome-free markdown file — no navigation, no footer, no scripts, just the actual content. /pricing has /pricing.md, /about has /about.md, and so on, generated automatically on every build. This very page is no exception.
Content negotiation
An agent doesn't need to know the .md URL at all — requesting the ordinary page with an Accept: text/markdown header gets the same clean markdown back automatically, with a matching Content-Type and an x-markdown-tokens estimate in the response headers. Everything else — images, stylesheets, ordinary browser requests — is completely unaffected.
Structured data (JSON-LD)
Every page carries schema.org structured data in its
: Organization and SoftwareApplication markup describing Collecurate as a whole, BreadcrumbList on every inner page, and FAQPage markup wherever there's a real FAQ. It's the same machine-readable format search engines use for rich results — a precise, structured description an agent can parse directly, instead of having to infer one from prose.API catalog
A machine-readable catalog at /.well-known/api-catalog, following RFC 9727, bookmarking the real Collecurate app for tools that discover APIs this way.
Start your first collection free
One link or a QR code is all it takes to start gathering everyone's photos.
