Find API keys and secrets exposed on your website
Front-end bundles are public, and keys end up in them more often than you would expect. ScanAnt reads your page, scripts and source maps, matches them against patterns for many providers and reports each hit with its location and a masked value. We never use a key we find.
Free, no signup. We only read what your site already shows every visitor.
What we check
Each item is tagged with the scan that reports it. The quick scan is free and needs no signup. The others run only on sites you have proved you own.
Payment keys
Quick scanSecret Stripe keys, Paddle and Lemon Squeezy keys and webhook signing secrets in the browser bundle. Test-mode Stripe keys are flagged separately from live ones.
AI provider keys
Quick scanKeys for OpenAI, Anthropic, Groq, Google Gemini, OpenRouter, Hugging Face and similar services that bill per call.
Cloud and developer tokens
Quick scanAWS access keys, GitHub and npm tokens, and private keys pasted into code.
Messaging and e-mail keys
Quick scanSendGrid, Resend, Mailgun, Postmark, Twilio, Slack, Discord and Telegram tokens or webhook URLs.
Database credentials
Quick scanConnection strings with passwords, signing secrets for login tokens and the Supabase service_role key, which bypasses all row-level security.
Source maps
Quick scanPublic .map files hand over your original source, including comments and variable names where secrets often hide.
Keys in a browser run
Deep scanKeys or tokens sent in request addresses or kept in browser storage while the app runs in a real browser.
Secrets in uploaded source
Code scanHard-coded keys, committed .env files and server secrets placed in variables that ship to the browser, found in a ZIP of your project.
- Quick scan: No signup. Reads only what every visitor receives.
- Deep scan: After you prove ownership and tick the deep option.
- Code scan: You upload your project as a ZIP.
No scan can promise to find every problem. A clean result means no known problems were found on what the scan could see.
Common mistakes
Using a server key behind a NEXT_PUBLIC_ or VITE_ variable
These prefixes tell the build tool to embed the value in public JavaScript.
Calling a paid API straight from the browser
The key must travel with every request, so anyone can copy it from the network tab.
Deleting a key from the code but not revoking it
The old value stays in cached bundles, git history and source maps.
Committing .env to git
Once pushed, the secret lives in history even after the file is removed.
How to fix it
Revoke and regenerate every exposed key at the provider first. Cleaning the code does not make the old key safe.
Move the call behind your own server endpoint or edge function that holds the key and checks who is calling.
Use publishable or anon keys in the browser only when the provider is designed for it, and restrict them by domain or policy.
Add rate limits and spending caps at the provider, and stop publishing source maps in production.
Questions
Do you store or use the keys you find?
Never. Findings keep only a masked value and the place where it was found. A key is not sent to an AI model and not used to make any request.
Is a Supabase anon key or Firebase config a leak?
No. They are public by design. The risk is whether the rules behind them are strict. The report labels them as normal and the full scan checks the rules.
How is a Google API key judged?
It is reported with a caveat, since many Google keys are meant for the browser. What matters is whether it is restricted to your domain and to specific APIs in the Google console.
What about keys that are not in the page but in my repo?
Upload your project as a ZIP for the code scan. It finds hard-coded secrets and committed env files that a public-page scan cannot see.
I found a leak. What first?
Rotate the key, then review the provider's usage log for calls you did not make. Only then clean up the code and the git history.
Related checks
Supabase security check: RLS, storage and auth
What anyone can read or write with your anon key, and whether service_role leaked.
Open →Vibe coding security scanner for AI-built apps
AI endpoints, keys in the browser and client-side paywalls in AI-built apps.
Open →Bolt.new security check for apps you built with Bolt
VITE_ secrets, open backend tables, source maps and headers in Bolt apps.
Open →
Run the scan now
Free for any public site, no signup. Need the verified, deep or code scan? Choose the level on the scan page.
Choose a scan level →