FAQ
Questions, answered.
How the test works, what your report means, and what to do about AI agents on your website.
The test
What is Agent Ready?
A free test that shows whether an AI agent can do business on your website. We read your public pages the way an agent does, find every way a customer can buy, book, join or get in touch, and, if you ask, send a real AI agent to walk those journeys step by step.
What does the AI agent actually do?
It opens your site in a real browser and tries a task, such as booking or getting in touch, the way a customer’s assistant would. It may click and read. It never types, submits, signs in or buys, and it stops at the step where the customer would enter their own details, pay or confirm.
How long does it take?
Reading your site takes one to three minutes. If you ask an AI agent to walk your journeys, add a minute or two for each one. Leave the page open, or keep the link and come back.
Will it change anything on my site?
No. It reads public pages only. It never fills in forms, adds to carts, books, signs in or buys, and it blocks analytics and ad pixels so your numbers stay clean.
What does it cost?
Nothing. The full report is free. To have the AI agent walk your journeys, we ask for your name and work email.
Does it work for any kind of website?
Yes, any public website: shops, booking sites, lead generation, software, nonprofits and more. Our accuracy checks cover eight industries. See how we measure it.
What doesn’t it test?
Pages behind a login, and anything after the final button, because we never press Pay, Book or Confirm. Each test reads up to 25 pages and walks up to two journeys.
Your report
What do the verdicts mean?
Agent ready: every journey we walked reached its final step, verified. Partly ready: some did, or a serious problem will trip agents up. Not ready: no journey an agent could follow. Blocked: your site turned our agent away before any journey. Looks promising or Needs work: journeys found but not walked yet. Full rules.
What does “verified” mean?
We don’t take the agent’s word for it. A journey is verified when the browser itself showed the final step: a form waiting for the customer’s details, a payment field, or the final button the agent refused to press.
Why is something marked inconclusive?
We couldn’t observe it, for example because a page wouldn’t load or the check needs a step we never take. We never count it as a pass.
Why does it say my site blocked the agent?
Your site turned our agent away with a bot challenge, an access-denied response, a rate limit or a robots.txt rule. Customers’ assistants often meet the same wall. The report says what we saw, so your developer or security provider can let assistants acting for customers through.
I fixed something. How do I test again?
Open your report and choose Test again. If you enter the same address within 24 hours, we show the last report instead, so use Test again for a fresh one.
Can I share my report?
Yes. Anyone with the link can open it, so you can send it to your developer or agency. Reports are hidden from search engines.
Agents and your website
Should I block AI bots?
Decide separately for two kinds. Training crawlers collect content for AI models, and blocking them is a fair business choice. Assistants acting for a customer, such as ChatGPT-User, Claude-User and Perplexity-User, visit because a person asked them to, so blocking them turns customers away. Your robots.txt can treat the two differently, and your report checks both.
Do I need an llms.txt file?
It is cheap to add, but few agents read it today, and it won’t fix a journey an agent can’t finish. Fix the journeys first. Your report says whether you have one.
What is WebMCP?
A proposed web standard that lets a page hand an AI agent named actions, such as “check availability” or “book an appointment”, instead of making it guess at buttons. Chrome is trialling it now. Every test checks whether your site offers WebMCP tools.
How is this different from SEO or AI visibility tools?
Those tell you whether AI assistants mention your business. Agent Ready tests what happens next: whether an assistant can actually book, buy or get in touch once it reaches your site.
Our agent and your data
Which agent do you use?
Our crawler, AgentReadyBot, reads your pages. For AI walks, Claude by Anthropic drives a real browser with the same user agent. Everything it does and never does.
How do I stop it visiting my site?
Add User-agent: AgentReadyBot and Disallow: / to your robots.txt. We will stop reading your site, and the report will say that agents are not welcome.
What do you keep?
The address you tested, the report, and a one-way hash of your IP address to limit abuse. If you ask for an AI walk, also your name, work email and company. Our privacy notice has the details.
Working with Sneeze It
Who is behind Agent Ready?
Sneeze It, a performance marketing agency for gyms, studios, spas, salons, kids’ programs, home services and the franchise systems behind them.
Can you fix what the report finds?
Yes. Sneeze It works through the list with you, then we test again with the same AI agent so you can see it get further. Book a walkthrough.
We have many locations or client sites.
We can test every location page or every client site and send one fix list, most serious first. Get in touch.
Still have a question?
Call 800-304-1150, email info@sneeze.it, or book a short call with David Steel.