A generated companion to llms.txt containing a header describing the source and the company, then every published article in full — URL, publication date, author, topic and summary, followed by the complete markdown body. Markdown rather than HTML, so headings and tables survive and navigation furniture doesn't.
Also called: full text for AI · AI corpus · one file for retrieval
- 1Published posts are concatenated with a metadata block above each body.
- 2Each article is fenced with its canonical URL so anything quoting a passage has the citation to hand.
- 3Served as text/markdown with the same caching as llms.txt.
- 4Linked from llms.txt under Optional, and on the public middleware allowlist.
The distinction from llms.txt is the reason it exists: "That one helps a system decide whether we're relevant; this one lets it actually answer without crawling ten pages, which is the difference between being cited and being skipped when a model is working under a time or token budget." Markdown is chosen deliberately — "it strips the navigation and CTA furniture that pollutes an extracted answer, and every heading and table survives intact. Tables in particular matter: they are the shape a retrieval system extracts most reliably."
- Answering from the site required fetching many pages.
- Extracted HTML carries navigation and CTA noise into the answer.
See it on your own jobs
Twenty minutes, your numbers, no slide deck. We’ll build one of your real buildings in front of you and send you the estimate link at the end — yours to keep either way.
or keep browsing features →