Is Lovable secure? A Lovable app security checklist
Is Lovable secure? Not by default. Lovable app security starts with Quick and Deep scans, a signed-out Supabase test, then what the published site shows.
This page didn't finish loading. The writing below is still here. Reload this page.
A security check for apps made with AI builders
Free security scanner for Lovable, Bolt and Cursor apps.
Built a phone app?Scan your APK, AAB or IPA in your browserSigned-in scans are only for a site you have verified you own. You provide one or two test accounts you control. SniffPup signs in as those accounts, sends read-only requests, reports counts and table or endpoint names only, and deletes the logins right after the scan. You are responsible for having authority to test the site. The free scan never signs in. Run a signed-in scan
Taking payments with Stripe? Check your Stripe code free →

Paste your site's URL and SniffPup checks it the way a visitor's browser would: 15 passive checks. It never logs in, never sends attack traffic and never keeps your data. You get a report sorted by severity, with fixes.
Is Lovable secure? Not by default. Lovable app security starts with Quick and Deep scans, a signed-out Supabase test, then what the published site shows.
How to secure Cursor: leave terminal and MCP approval on, keep secrets out of .env and the browser bundle, then scan the live site you own for leaked keys.
Windsurf security risks on Devin Desktop: .env in git, secrets in VITE_ or NEXT_PUBLIC_, Supabase without RLS, and what the live site shows a stranger.
Check whether your live site leaks secret API keys: search your JavaScript bundles, test /.env and /.git paths, check source maps, and rotate anything you find.
Check Row Level Security in two parts: SQL for tables with RLS off and open policies such as using (true), then a signed-out test with your public anon key.
Leaked keys, exposed files, headers, HTTPS, CORS, database exposure and email setup.
Learn more →We only look at what any visitor's browser can see. Secrets we find are always masked.
Learn more →Copy a prompt made for Lovable, Cursor, Bolt or Claude Code, then re-check.
Learn more →Repo scan for a public GitHub repo, and the Payments Doctor for your Stripe code.
Learn more →Built withLovableBoltCursorReplitv0Supabase· Scan it.
Works on any public http(s) website, whatever made it. We're not affiliated with these companies.
Add the SniffPup connector, then ask Claude to scan a site you own. You'll get the score, the issues found and plain-English fixes for the most serious ones, right in the chat. How to connect it →
Listed in Claude's directory
How it works
Three steps, no security jargon.
Drop in your live URL. We load it like a visitor would. No login, no code access needed.

Every issue gets a severity and a short explanation in normal words.
Copy a prompt made for Lovable, Cursor, Bolt or Claude Code, let it fix the issue, redeploy, and re-check.

Example report
Example report · demo-app.example is a made-up site with fake keys, not a real customer
Everything in one place
Every feature, one click away. The same list lives in the Features menu at the top of every page.
One scan for everything: keys, exposed files, source maps, headers, Supabase and Firebase, speed and crashes. Passive, read-only, no signup.
Open →FreeSee a full report before you scan.
Open →Paste your link, see what's wrong, paste the fix.
Open →FreeAdd the SniffPup connector, then ask Claude to scan a site you own. You get the score, the issues found and plain-English fixes for the most serious ones, right in the chat.
Open →Start at 100, minus points per finding. No letter grades.
Open →Paste-in checks for other common bugs in AI-built apps.
Open →Free quick checkPaste your Stripe checkout and webhook code. Quick check free; full check on Solo and up.
Open →Free quick checkCheck a public GitHub repo for committed keys, .env files, tables without RLS and Stripe mistakes, plus a licence check and an SBOM of your npm packages.
Open →Free accountConnect your Supabase project for a free, read-only check of its security settings. You get a score out of 100, and critical issues come with their fix.
Open →FreePaste a database migration, or scan a repo, to catch steps that delete or rewrite data before they run. It also flags migrations that run on deploy with no backup first, and the Supabase project check shows whether you have a backup to restore from.
Open →FreePart of the free repo scan: packages and files you likely don't use, copied code, very large files and errors that get hidden, with a cleanup score out of 100 and a step-by-step cleanup plan. Every finding is free; Solo and up get every fix step.
Open →FreeQuick checks, code and repo checks, the RLS policy explainer and builder tools. Everything else is in the free scan.
Open →Every fix with prompts for Lovable, Cursor, Bolt and Claude Code, plus 30 days of re-scans.
Open →Full reportsTurns your findings into a rules file for Cursor, Claude Code, Lovable or Bolt, so your AI helper is told not to repeat them. In the full fix report.
Open →On unlocked findingsFixed something? Proof it worked re-checks just that finding and shows before and after.
Open →Full reportsDownload any full report.
Open →Free accountEvery scan, score trends and before/after history.
Open →PlansWeekly on Free, daily on Solo, and every hour on Starter, Pro and Ultimate for sites you've verified, with an email only when something new and serious shows up or gets worse.
Open →DNS TXT record or a file. Unlocks the Pro deep checks.
Open →ProRead-only: which tables, collections and buckets a stranger can read. Counts only.
Open →ProAdvisor + outside scan + deep check + fix prompts in one report.
Open →Verified ownersScore 80+? Show a badge and public score page. Score and date only.
Open →Every check SniffPup runs, in one list.
Open →Pre-launch checklists for Lovable, Bolt, Cursor, Replit and Supabase.
Open →What to do in the next 10 minutes: OpenAI, Stripe, Supabase, AWS, GitHub.
Open →Everything we shipped, newest first.
Open →What we check and what others check. Facts only.
Open →What the free scan checks
The same way a stranger would look. You can watch each one run live when you scan. The anonymized write-up of those scans, with no site named, is the state of vibe-coded app security.
We'd rather you know this up front


Yes, Ember is a real dog
She's a small, very fluffy black Pomeranian, and she notices everything: a new smell on the walk, a dropped snack, a door left open.
That's the job we built SniffPup to do for your app. AI builders are great at shipping fast, and they sometimes leave a key in the JavaScript or a .env file on the server. We find those before someone else does.
Ember doesn't write the code. She supervises from the couch.
Read Ember's story →What it costs
See every issue. Full fixes free for every critical issue (no critical? one other fix is free).
Scan freeEvery fix for one app, plus 30 days of free re-scans. No account, no subscription.
See the reportSolo $7, Starter $15, Pro $29, Ultimate $49. Daily or hourly re-scans and alerts.
Compare plansFAQ
Yes. Free scans are passive. SniffPup loads your page and the JavaScript it ships, like a normal browser, plus a short fixed list of well-known paths such as /.env and /.git/config. It never logs in, never sends attack payloads and never uses a secret key it finds. Full scanner details.
No. We store findings with secrets masked, never the full value. For Pro database checks we store only whether a table or bucket is readable and its row count, never row contents. See the Privacy Policy.
Every finding for one app, with a plain-English explanation and a copy-paste fix prompt for Lovable, Cursor, Bolt and Claude Code, plus 30 days of free re-scans so you can check your fixes. You pay once. No account, no subscription.
You can cancel a subscription bought on our website anytime from the billing portal; it ends at the end of the billing period. Purchases made in our app are cancelled and refunded through the store you paid. All sales are final, except where required by law or if we charged you in error. Here's the refund policy.
No scanner can promise that. A passive scan catches common, costly mistakes quickly. It is not a penetration test or a code audit.
Email abuse@sniffpup.com with your domain and we'll block scans of it. More about the SniffPupScanner bot.