← All blog posts

What an Agent-Ready Score Is, and Why You Can Now Share Yours

August 22, 2026 · 6 min read

AI agents read websites differently than people do. The Agent-Ready check scores how clearly yours can be read, and every result now has a permanent, public link you can share with your team, your client, or your audience.

When someone asks an AI assistant "who should I hire for this," the assistant does not browse your site the way a person does. It fetches files. It follows robots.txt. It looks for machine-readable text. If what it finds is thin, blocked, or buried in JavaScript, your business shows up vague or not at all, no matter how good the site looks to humans.

The Agent-Ready check exists to answer one question: how clearly can an AI agent read your website right now?

What the score actually measures

The check fetches your site the way an agent would and reports what was reachable:

• Your homepage, and whether meaningful text comes back without a browser rendering it
• robots.txt, and whether AI crawlers are allowed or accidentally blocked
• sitemap.xml, and whether your pages are discoverable in one pass
• llms.txt, the emerging convention for telling AI tools what your business does and where the important pages live

Those signals produce a score out of 100 and a color tier. Red and orange mean agents are largely guessing. Yellow means the basics are there and the details are missing. Green means you are already ahead of most sites we scan.

It is an access score, not a marketing audit. It does not grade your copy, your design, or your rankings. It tells you whether the machine on the other end got a clean read.

Why "up to 100" is not a vanity number

Every point in the score maps to something an agent either found or did not. A site at 62 is not badly written, it is usually just missing a clean text layer and a pointer file. That is a fixable gap, and it is fixable without touching how your site looks to human visitors. You publish a few files on your own domain and the machine-facing read improves.

That distinction matters because most site owners assume improving AI visibility means rebuilding something. It usually does not. It means adding the layer that was never there.

The new part: your score has a permanent link

Until now, an Agent-Ready result lived in the browser tab where you ran it. Close the tab, lose the result.

Every check now produces a public, read-only report at its own URL. That page shows the domain that was checked, when it was checked, the score, and the same tiered guidance you saw when you ran it. No login. Nothing expires when you refresh.

That small change unlocks a few things people were already trying to do manually:

• Send it to whoever owns the website. Agencies and consultants can run a check on a prospect or client and share the exact result instead of describing it.
• Put it in a proposal or a deck. A live link ages better than a screenshot, because it reflects the domain, not a moment.
• Show your work publicly. If you scored green, that is a legitimate signal worth posting. If you scored low and then fixed it, the before and after is a story.
• Compare across a portfolio. Run every property you manage, keep the links, and you have a readable snapshot of where the gaps are.

What the report page does not do

It is deliberately narrow. A shared report will not run a new scan, will not ask a visitor to log in, and will not expose anything that was not already public on the checked domain. Anyone landing on someone else's shared score gets one obvious next step: check their own site.

That is the whole design. The report is evidence, not a funnel wrapped in a paywall.

How to use this in practice

1. Run the check on your primary domain. Note the score and the tier.
2. Save the report link. It is the version you can send to anyone.
3. Fix the reachable gaps first: unblock AI crawlers in robots.txt, make sure sitemap.xml is current, publish a clean text layer and an llms.txt.
4. Re-run it. The new report gives you a fresh link, and the delta is the proof that the work landed.

Why this matters now

Discovery is moving from "which page ranks" to "what the assistant says when someone asks." In that world, being readable by machines is not a technical nicety, it is whether you get mentioned at all. The score tells you where you stand. The shareable report makes that answer something you can hand to another person.

Want to understand how AI agents actually read your site? Read the full guide →

Check your website's Agent-Ready score