Is your siteAI‑ready?
See your site the way AI crawlers and agents read it, check by check, with the exact fix for anything missing.
Ten checks, one hundred points
Each check earns its full points on a pass and half on a partial pass. The checker fetches the homepage, robots.txt, llms.txt, the agent card, the sitemap and a sample of Markdown pages, once each, with an honest user agent and a 20 second limit. It runs no model and keeps nothing.
Open to a plain request
A plain request reaches the page, not a challenge or a bot wall. A wall caps the whole score at 20.
AI crawlers allowed
robots.txt lets GPTBot, ClaudeBot, PerplexityBot and Google-Extended read public pages, with no blanket Disallow left over.
llms.txt
/llms.txt follows the llmstxt.org shape: a title, a one-line summary and sections of links to the pages that matter.
Structured data (JSON-LD)
Structured data in the server HTML, with real schema.org types, so an agent knows what the business is.
Readable without JavaScript
The main text is in the HTML itself, readable without running JavaScript.
Points agents to llms.txt
The homepage points to /llms.txt (rel="describedby") and to Markdown versions of its pages (rel="alternate"), in the HTML or the headers.
Markdown versions of pages
The Markdown pages that llms.txt links to really load. Up to five are sampled.
Agent card
/.well-known/agent-card.json describes what an agent can do on the site. It is a newer standard, so it weighs less.
Open Graph preview
og:title, og:description and og:image, so links unfurl with a real preview.
Sitemap
A sitemap, found through robots.txt or at /sitemap.xml.
Every Laarpi site is AI-ready on publish
Laarpi compiles the agent door with the site, from the same source as its pages, on every plan, Free included. There is nothing to set up and no file to write by hand.
- An llms.txt that maps the siteOne plain index agents read first, written from the real pages.
- A Markdown twin of every pageThe same words without the layout, linked from llms.txt.
- Structured data in the server HTMLWhat the site is, who runs it and what it sells, in the first byte.
- Forms agents can useEach form registers as a tool where the browser supports WebMCP.
- An agent card/.well-known/agent-card.json says what an agent can do there.
- A robots file for AI crawlersGPTBot, ClaudeBot, PerplexityBot and Google-Extended let in.