Let the cat judge
your website.
Paste your URL. In 60 seconds you'll know exactly what to fix — with prompts you can paste straight into Cursor, Bolt, or whatever you used to build it.
3 free scans. No card. No signup.
How it works
Three steps. Sixty seconds. No signup required.
🔗Paste your URL
Drop any live URL. Vercel, Netlify, Railway — if it loads in a browser, the cat can judge it.
🐱The cat scans your site
A real browser loads your page and AI checks everything — menus, forms, mobile, accessibility, speed, design, and content. Takes about 60 seconds.
📋Get your report + fix prompts
Scores across 7 areas, a list of what’s wrong ranked by importance, and ready-to-use prompts you paste into Cursor, Bolt, or whatever you build with.
A real scan of a real site
1 life spent on this report
“Your navigation has 14 items. Fourteen. A restaurant menu has fewer choices.”
“Someone typed 'asdf@' into your signup form and got a success message. That's not validation.”
“Your hero text is 11px on mobile. I'm a cat and even I squint at that.”
3 critical issues. 7 warnings. 1 thing that was actually fine.
Fix prompts for Cursor, Bolt, Lovable
Copy. Paste. Fixed.
Every finding comes with a prompt. Paste it into whatever you build with.
Your nav has 14 items — cut it to 6. Move the rest to a dropdown. On mobile, collapse into a hamburger menu. Highlight the active page with your primary color.
Eight categories. One obsessive cat.
What the cat actually checks.
Not vibes. A structured audit — every category judged, every finding scored, every fix prompted.
Navigation
Can people find what they need? Menus, links, buttons — is anything buried or confusing?
Visual Design
Does it look right? Colors, spacing, fonts, and whether the important stuff stands out.
Forms & Inputs
Do your forms actually work? Error messages, labels, and what happens when someone types something wrong.
Accessibility
Can everyone use it? Screen readers, keyboard-only users, color contrast, image descriptions.
Mobile
Does it work on phones? Text size, tap targets, layout — everything people actually use.
Performance
Does it feel fast? Loading indicators, heavy images, and anything that makes users wait.
Content
Is the writing clear? Headings, readability, and whether your copy actually says something.
Security
New — ShieldAre your API keys showing? Missing protections, leaked secrets, login holes the AI left behind.
The full picture
See what you get
Everything in one place — your score, what's wrong, how to fix it, and what your site looks like on desktop and mobile.
“Not bad for a vibe-coded app. But that nav needs serious work.”
Category breakdown
Sample finding
Navigation has too many top-level items
Your main menu has 14 links. That's too many choices — people get lost. Most apps need 5-7.
Fix prompt
Reduce main nav to 5-7 items. Move secondary links to a dropdown or footer. On mobile, collapse into a hamburger menu.
Tech detected
Screenshots
Proof, not promises
Fix. Re-scan. Watch it climb.
Re-scan the same URL free within 14 days.
Before
After
The cat went from “yikes” to “acceptable.”
How do you stack up?
Scan any site alongside yours — a competitor or one you admire. 2 lives.
Your app
62
Competitor
71
“You're beating them on forms. But mobile? They're lapping you.”
Sign in, and the cat remembers.
More than a score.
Six things you get on top of every scan — once you're signed in.
- Your dashboard
- All your scans in one place. Start a new scan, check past results, see how your apps are doing.
- Compare with competitors
- Scan any site alongside yours. See exactly where you're ahead and where they're beating you.
- Four scan types
- Quick check, full UX deep-dive, security-only, or everything at once. Pick what you need.
- Scan history
- Every report saved and searchable. Filter by type, sort by score, find any past scan instantly.
- Tech stack detection
- The cat spots what your site is built with — Next.js, Tailwind, Vercel — and tailors its advice.
- Share with your team
- Send a private link to anyone. They see the full report, no account needed.
In June 2024, hackers found hardcoded API keys for ElevenLabs, SendGrid, and Azure buried in Rabbit R1's source code. With one key they could access every voice message ever generated by any of the 130,000 devices sold — and send phishing emails from Rabbit's official domain.
When Rabbit finally rotated the keys, they forgot to update their own server code first — bricking every device for hours. Nobody had checked.
A startup called Enrichlead launched a platform where 100% of the code was written by Cursor AI. The interface looked polished. The security logic lived entirely on the client side.
Within 72 hours, users discovered they could open the browser console and bypass payment to access every premium feature for free. The project shut down entirely. The AI never warned them.
In July 2025, Wiz researchers found that Base44 — a vibe-coding platform acquired by Wix — let anyone bypass authentication on private enterprise apps using only a public app ID from the URL.
Internal chatbots, HR systems, and knowledge bases with sensitive employee data were wide open. The platform's AI-generated auth checks had a gap nobody noticed until a security team looked.
Security researchers found roughly 3,000 production websites exposing their OpenAI API keys in JavaScript bundles and static assets. Unlike repo leaks that get caught, these keys sat in live frontend code for months.
Anyone inspecting the page source could grab the key and rack up charges. The AI tools that built these sites bundled the secret key right into the client. Nobody checked.
In 2024, GitHub detected 39 million leaked secretsacross its repositories — a 25% jump from the year before. 35% of those secrets were found in JavaScript files, sitting in plain sight.
Repos using AI coding tools showed a 40% higher rate of secret exposure. The tools generate working code fast, but they don't think about what they leave behind.
These are real security failures in AI-built apps — found after they shipped.
That's why we built Shield.
A security scan that catches what the AI tools leave behind.
Add Shield to any scan for 5 lives. Sleep better.
Your UX credential — everywhere
The score card is a moment. The badge is a billboard. Put it where people look.
In your GitHub README
## My App
A dashboard for tracking metrics.
In your landing page footer
<!-- Add to your README or landing page --> <a href="https://usability.cat/r/abc123"> <img src="https://usability.cat/badge/abc123.svg" alt="usability.cat - 4 paws" /> </a>
Lives & Paws
How the cat gets paid.
Lives 🐱
Every scan costs lives. You start with 3, free.
What the cat charges
Every account starts with 3 free lives. No credit card. When you need more, grab a pack. Lives never expire — use them whenever you're ready to ship.
Paws 🐾
Earn the cat's approval.
Your paw rating
Your paw rating is your app's UX credential. Put it in your README, share it on Twitter, show it to your users. It means a real AI audit was done and your app earned it.
Get Lives
The cat doesn't work for free. But lives never expire.
Kitten
Enough for 3 full scans with fix prompts.
Hunter
Enough for 9 full scans with fix prompts.
Apex
Scan every project you build this year.
No subscriptions. No recurring charges. The cat respects your wallet.
The cat rewards effort.
Earn free lives.
Four ways to stack lives without spending. The more you share, the more you scan. The more you scan, the better your app gets.
- +1life📤
Share your score
Post your paw rating on X or LinkedIn. When someone scans their app from your post, you both get a life.
- +3lives👥
Refer a builder
Send your referral link. When they scan their first app, you both get 3 lives.
- +1life📈
Improve your score
Re-scan after fixing issues. If your paw rating goes up, the cat gives a life back.
- +3lives🏆
Earn 5 paws
Score 90+ for the first time. The cat is impressed. That's rare. Have 3 lives on the house.
Questions the cat gets asked
Mostly reasonable ones.
Ready for the verdict?
3 free scans. Full reports. Every finding, every fix prompt, your paw rating. No watered-down preview.
No card. No signup. Just a judgy cat.