Built it with Lovable, Bolt, v0, Cursor or Replit?
AI tools ship working apps fast, and they ship the same security mistakes again and again. This is the kind of thing a scan turns up.
Example report
my-saas.app
Very hackable. Fix this today.
9 issues. Detected: Next.js, Supabase, Vercel.
Grade F
Critical in database
Anyone can read your profiles, payments tables
Using only the public key shipped in your page, a visitor who is not signed in can read rows from these tables.
profiles (1,204 rows), payments (318 rows)
Critical in leaked secrets
OpenAI API key is visible to every visitor
sk-proj-…x9Qa in /_next/static/chunks/app/page-4f1c.js
In the full report: the fix, written for this stack
alter table profiles enable row level security; create policy "Owners read their own profile" on profiles for select using (auth.uid() = id);
How it works
1
Enter your domain
No install, no account and no access to your code. The scan only needs the address your app is live at.
2
Get your grade in ten seconds
An A to F grade, a count of issues at each severity, and your three most serious issues in full with how to fix them.
3
Fix it with one prompt
The $19 report adds every issue, fixes written for your stack, and a single prompt that tells your AI coding tool to fix them all.
What the scan checks
The mistakes AI-built apps ship with most, sorted by how badly each one can hurt you.
- Critical
- Secret API keys shipped in your JavaScript (Stripe, OpenAI, Anthropic, GitHub, the Supabase service role). Downloadable .env files and Git repositories. Supabase tables anyone can read without signing in.
- High
- Pages that still load over unencrypted HTTP. CORS settings that let any website read a signed-in user's data. Cloud and chat tokens in your bundle.
- Medium
- No Content Security Policy, no HSTS, no clickjacking protection. Login cookies any script can read. A domain anyone can send email from.
- Low
- Public source maps, advertised server versions, missing minor headers, and no way for a researcher to report a hole to you.
Free to scan. $19 for everything.
One payment per domain, plus sales tax where it applies. No account, no subscription. Your report link is emailed to you, and if it is not useful you get a full refund within 14 days.
Free scan
$0
- An A to F grade for your app
- How many issues you have at each severity
- Your three most serious issues in full
- A standard fix for each of those three
Full report
$19
- Every issue the scan found, in full
- Fixes written for your stack and host, with the code to paste
- One prompt that tells Cursor, Claude Code, Lovable or Bolt to fix everything
- A plain-English summary of what an attacker could do today
- Unlimited re-scans of that domain
Would you rather have a person fix it for you? Get in touch.
Questions
Is it safe to run on my live app?
Yes. The scan only reads what your app already serves to every visitor: pages, scripts, headers and a few well-known file paths. It never logs in, submits a form or writes anything. If your app uses Supabase, it counts the rows your public key can see without downloading them.
I am not a developer. Can I still fix what it finds?
That is who it is built for. Every issue is explained in plain English, and the full report gives you one prompt to paste into the AI tool that built your app. A few fixes, like replacing a leaked key, need a couple of clicks in a dashboard, and the report tells you exactly where. If your app lives on a builder's own address, such as a lovable.app subdomain, some settings are not yours to change, and the report says which.
What do I get for $19?
Every issue instead of the top three, fixes written for the stack and host the scan detected, the fix-everything prompt, and unlimited re-scans of that domain. It is one payment per domain plus sales tax where it applies, with no account and no subscription. Your report link is emailed to you. If it was not useful, email us within 14 days for a full refund.
Does a good grade mean my app is secure?
It means an outsider finds nothing obvious. The scan covers what is publicly visible. It cannot see what happens after login, your server code or your payment logic. For those you need a code review or a penetration test.
Can I scan a site I do not own?
No. Only scan apps you own or have permission to test.
What happens to my data?
There are no accounts and no scan history. Results sit in a short-lived cache for a few minutes. For paid reports, the findings are sent to an AI model to write your fix plan. Secrets are shortened before they leave the scanner.