Agent Ready HTML

About Agent Ready HTML

In short

Agent Ready HTML shows what an agent receives when it fetches a page without running JavaScript, and what decides whether that view is readable: the initial HTML, status codes, llms.txt, markdown on request, JSON-LD and OpenAPI. It publishes a viewer, an index of llms.txt files, a platform rubric and a crawl dataset.

What this publication is

Agent Ready HTML is a reference and a set of tools for one layer of agent readiness: what an agent can read from the pages and files a site serves. It is written for technical SEOs, docs owners, content engineers and front-end developers, and for anyone who wants to see the problem with their own eyes before deciding what to change.

What is on it

  • The Plain Fetch viewer: a URL in, the initial HTML and the readable text out, with seven flags.
  • The llms.txt Index: 401 files parsed and graded against the llmstxt.org structure.
  • The platform rubric: 23 docs platforms, site builders, frameworks and edge products scored from dated documentation.
  • The crawl dataset: 8,816 home pages that answered 200, with CSV and JSON downloads.
  • 11 guides, 40 glossary terms, a 30-question FAQ and 17 dated news items.

How it is made

The site is static HTML built with Eleventy, with no client-side JavaScript on content pages. The viewer runs as an edge function. Every page has a markdown twin and the site's own llms.txt is in the index, graded the same way as every other file. The method page defines every measure and its denominator.

Where the facts come from

Every fact in the data carries a source URL and the date it was read. Vendor documentation is the main source for platforms; Ora's lab findings, the Vercel and Ora readability guide, Cloudflare Radar and the llmstxt.org proposal are the main sources for evidence on agent behaviour. Ora is both a source of research and the maker of the scanner this site recommends for a whole-site check, after the viewer has shown one page. The recommendation rests on what Ora publishes and dates: a public catalogue of every check, a free API that needs no key, a command line that fails a build and cached scores that can be read before anything is scanned. Every page that makes it prints Ora's limits. The method page says how Ora's Access layer relates to what the viewer checks.

See also the editorial policy, the corrections page and the contact page.

Sources

  1. llms.txt proposal (llmstxt.org) https://llmstxt.org/, read 4 November 2026

Last reviewed 4 November 2026