For people who build apps with AI
Your app might be leaking your users' private data right now — and you'd never know.
Apps built fast with AI often let one user read another's private info, or leave secret keys sitting in the open. It's invisible from the outside — unless you know exactly how to look. We look, explain what we find in plain English, and hand you the fix.
No jargon. No install. Free check needs nothing but your web address.
This might be happening in your app right now
The three most common ways AI-built apps leak data. All invisible from the outside.
Your users can read each other's private data
On loads of AI-built apps, one customer can see another's orders, messages or details — just by changing a number in the web address.
Your secret keys are out in the open
Keys for Stripe, OpenAI, your database — accidentally left in your public code, where anyone can copy them and run up your bill.
Private pages that anyone can open
A page that should need a login quietly serves its data to the whole internet — no password required.
Here's the thing: you can't see any of it
You're a builder, not a hacker. You made something great, fast — that's the point of building with AI. But checking it's safe needs a completely different skill: thinking like an attacker. That's the gap we fill, so you don't have to learn any of it.
“Can't Claude or Cursor just check it for me?”
No — and this is the important bit. The AI that built your app is brilliant at writing code, but it can't do what we do. It never opens your live, running app, never signs in as two different real users, and never tries to break in from the outside. It's reading the blueprints; we're rattling the doors and windows of the finished building. That's why we catch things it can't.
How we fix that for you
Give us your web address
That's it to start. No install, no code, no account needed for the free check.
We poke it like a nosy stranger would
We try the exact tricks an attacker uses — safely, changing nothing — and see what opens up.
You get a plain report + the fix
Every problem explained simply, with a ready-made instruction you paste into Claude Code or Cursor. It writes the fix for you.
Simple pricing
Start free. Pay only when you want the deeper checks.
Free check
£0See what a stranger can see, in a minute.
- Just your web address — nothing to install
- Finds exposed keys, public files and missing protections
- A plain-English score out of 100
Full launch check
£49The proper once-over before you go live.
- Everything in the free check, but far deeper
- We actively test your live app, not just look at it
- Every serious problem is proven before it counts
- Plain fixes you paste straight into your AI editor
Deep data check
£99Can your users see each other's data? We find out.
- Everything in the £49 check
- We sign in as two of your own test accounts
- We prove whether one user can read another's private data
- Read-only — nothing changed. Logins deleted after.
We only tell you about problems we've proved
Most tools bury you in scary "maybes". We don't. If we say something's wrong, we actually proved it — so your score is honest, and you're never frightened by a false alarm.
Common questions
I'm not technical. Is this for me?
Especially you. Every result is in plain English with a ready-made fix you paste into the same AI tool you built with. No security knowledge needed.
Will the check break or change my app?
No. We only ever look and read — we never create, change or delete anything. It's completely safe to run on your live app.
The £99 check asks for logins — is that safe?
You give two test accounts you own (never a real customer's). We only read, and your logins are encrypted, used once, then permanently deleted. Never logged, never shown.
How is this different from other scanners?
We only tell you about problems we actually proved — not a wall of scary 'maybes'. The score you get is honest.
Find out if your app is safe
The free check takes about a minute and needs nothing but your web address.