AGENT READINESS STANDARD 1.0
Is your service agent‑ready?
Your customers are starting to send their own AI assistants to deal with you. ARS shows whether those assistants can get the job done, safely. A person on our team reviews every report.
Your ARS level, checked against the published method
The five fixes that would raise your score the most
In your inbox within 2 business days
OnClarity sells AI agents too, so the method is public and anyone can check a score. Read the standard.
WHAT WE SCORE
Five groups. 100 points. One open method.
ARS scores what happens when a customer’s personal AI assistant contacts your service on their behalf. A full test runs on the live service from the outside, in English and Arabic, with German and French in review. The method is public, so anyone can check a score.
15 pts
Let agents in
Assistants can find and reach your service and read your help: machine-readable help, no CAPTCHA wall, no login for general questions.
25 pts
Give answers agents can use
Right answers grounded in your policies, with exact amounts, dates and links, the same every time, in the customer’s language.
25 pts
Get the job done
Status, changes, cancellations and refunds completed or routed, with a clear next step and a route to a person.
25 pts
Protect the customer
Verifies it’s really the customer, asks them before anything irreversible, and asks only for the data the job needs.
10 pts
Be open
Says it is AI, explains where data goes, and never asks for passwords in chat.
HOW IT WORKS
How you get your score
01
Tell us your website
It takes a minute. Use your work email so the report goes to the right company.
02
We test and review
We run the ARS Level 1 checks on your public pages. If you have a chat assistant, we may also test it as a declared AI assistant, with no real personal data. Someone on our team checks every finding.
03
You get your report
Your ARS level, the evidence and your top fixes, within 2 business days. Book 20 minutes with us if you’d like to go through it.
Questions
Is it free?
Yes. The ARS snapshot is free for any company, one report per company domain.
How do you test our service?
Our test assistant says it’s an AI assistant running an automated evaluation, and uses reserved test details. It never pretends to be a real customer, never confirms an action, and never asks to be connected to your staff.
What is a personal AI assistant, and why does it matter?
It’s an AI assistant a customer sends to deal with a company for them: checking an order, changing a booking, asking for a refund. ARS is an open standard for whether your service lets that assistant get the job done, safely.
OnClarity sells AI agents. Is this fair?
ARS scores deployments, not vendors, and never ranks the companies that build AI agents. The method is public, so you can check every point.
Can I dispute my result?
Yes. Reply to your report email and we’ll look again. We re-test after you make fixes.