FREE AI-READINESS DIAGNOSTIC
Would AI recommend you, or your competitors?
You can rank #1 on Google and be invisible to ChatGPT, Claude and Perplexity at the same time. We read your site the way an AI agent does, score it across three pillars - found, understood and trusted - and show you exactly what to fix, right down to the answer an AI gives when a buyer asks about you.
Free · 10 seconds · no email · zero cookies
5 free scans a day
- Your score out of 100 - found, understood and trusted, in seconds.
- The exact gaps costing you visibility, clarity and trust - and the deals that follow.
- Top fixes ranked by impact - what to change this week, and what each is worth.
- The answer AI gives about you. Unlock what an assistant actually tells buyers - generated live from your page.
- A shareable report link. Scan a competitor while you're at it - it's free.
What do we check?
More than 40 checks across the three tests an AI agent runs before it recommends anyone:
- Be Found. Can AI reach you at all? We read your robots.txt, llms.txt and sitemaps, then test whether AI crawlers actually get through at the edge - many sites quietly block ChatGPT or Perplexity without knowing it. If your security wall turns automated readers away, that's the headline finding, front and centre.
- Be Understood. Is your offer legible to a machine? Structured data, clear headings, coherent messaging - and whether your content even exists without JavaScript, because most AI crawlers never run it. A JavaScript-only page reads as near-blank.
- Be Trusted. Will an agent stake its recommendation on you? Load speed, security signals, and email anti-spoofing (SPF and DMARC) - so you find out whether anyone can send email as you.
The free scan is deterministic - no LLM, no guesswork - so the same site always scores the same way. Fix something, re-scan, and trust the movement.
Do you scan my whole site, or just the homepage?
The homepage first - because that's the page an AI agent reads first when a buyer asks about you. We read it exactly the way an agent does, then add the files that sit around it: robots.txt, your sitemap, llms.txt and your email DNS (SPF and DMARC).
It's the front door - and most sites fail at the front door. Going deep across every page, then keeping it that way, is what CORE is for.
What does the deep audit add?
The part that stings - and the part that wins the deal. One work email unlocks:
- Your page, two readers. What a person sees, side by side with the raw text an AI agent actually reads. On most sites, the two barely match.
- The answer AI gives about you. How an assistant describes your business to a buyer, generated live from your page - beside the facts it can confirm, and the gaps where it's guessing.
- The full read-out. Every check explained, with the exact fix for each.
No card, no spam. It unlocks instantly, right on the report.
How can you know what AI says about my business - and is it accurate?
We generate the answer live from your own page - the same raw text an AI assistant reads - so it reflects what an agent could actually say about you today, not a canned example.
Then we keep it honest. Beside the answer sits a plain confirm-or-guess ledger: the facts an assistant can verify from your site, and the gaps where it's forced to guess. The prose is the assistant's; the facts are checked, not invented. Where it's guessing about you, so is every real AI - and that's exactly the problem to fix.
I already rank well on Google. Doesn't that cover it?
No, and that's the trap. Google renders your JavaScript before it reads you; most AI crawlers do not. So you can sit at the top of Google and still be blank, or misdescribed, to ChatGPT, Claude and Perplexity - the tools your buyers increasingly ask first.
This scan reads your site the way those agents do, not the way Googlebot does, so you see the gap Google hides. And that gap never shows in your analytics - you simply never see the enquiry you lost.
What if you can't reach my site?
Then that's your result, not an error. If your firewall, bot protection or robots.txt turns our reader away, it turns ChatGPT and Perplexity away too - so we report exactly what blocked us and what it means, instead of a score.
We're a polite, single-page reader that identifies itself and obeys your robots.txt. We never hammer your site or touch anything behind a login. If a security wall is quietly hiding you from AI, that finding alone is worth the scan.
What's the easiest way to improve my score?
Start with the quick wins:
- A clean robots.txt that doesn't accidentally block AI crawlers.
- An llms.txt file that tells agents what you do and where to look.
- Structured data (schema.org) so machines read your offer, not just render it.
- Server-rendered content, so your page still says something when the JavaScript never runs.
- Faster load times, because agents and visitors both abandon slow pages.
Run the scan first - the report tells you exactly which of these you're missing, ranked by impact. Prefer to skip the homework? That's what CORE is for.
Is it really free? What's the catch?
The score is free and needs no email. You get 5 fresh scans a day; each report is cached for a week and lives at its own shareable link - send it to your developer, or run it on a competitor.
The only trade: the deep audit asks for your work email. That's it. No card, no calls you didn't ask for. We built this because businesses that see what an AI sees tend to want it fixed - and fixing it is what we sell.
Can you just fix it for us?
Yes - that's the point. Agentu CORE is our website management service: we build, host and optimise your site end to end, so reports like this become our job, not yours. From £195/mo.
Run the scan, then talk to us about your report, or see what CORE includes. No pitch until there's a fit.