What Bolt already checks
Bolt has its own security audit. According to Bolt's announcement, you can run it from the publish menu: it looks for problems such as broken access control, secrets in code and database policies, applies the fixes it can and flags the rest for you to review. Bolt's release notes say the audit is available on paid plans.
Use it before you publish. Bolt describes it as a first pass, and it works from inside your project. Securick is the independent outside check of the published app: the code visitors download, the headers your host sends and what your database answers to the public key.
How Bolt apps reach the internet
Bolt can publish your project for you: Bolt's hosting docs describe free hosting on an address ending in bolt.host. For data, Bolt's database docs describe a database Bolt manages for you, or your own Supabase project. Bolt also has a place for secrets and environment variables, so keys don't have to live in your code.
Once it's published, the address is public. Anyone can open it, read the JavaScript it sends and try the API it talks to. That's the part Securick checks.
Security mistakes to look for in Bolt apps
An AI or payment key in the browser
The quickest way to call OpenAI, Anthropic or Stripe is straight from the page, and that's exactly what leaks the key. In a Vite project, for example, anything named VITE_… is built into the code visitors download (Vite docs). Calls that need a secret key belong in a server function that keeps the key on the server.
A database the public key can read
Whether the database is the one Bolt manages or your own Supabase project, the browser talks to it with a public key. What that key may read depends on the access rules. If they're open, so is your data.
Files that shouldn't be online
Source maps, environment files, backups and directory listings sometimes end up on the host with the rest of the build. Each can give away code, keys or the structure of your app.
Headers left to chance
HTTPS redirects, HSTS, a Content-Security-Policy and clickjacking protection depend on the hosting setup. A custom domain can behave differently from the default address, so check the address your users actually use.
What Securick tests on a Bolt app
These come straight from our list of tests. The free scan runs 46 tests; Pro runs 130.
- #2 OpenAI API key exposed Free
- #3 Anthropic API key exposed Free
- #1 Stripe live secret key exposed Free
- #10 Secret values in public env variables Free
- #11 .env file public Free
- #13 JavaScript source maps public Free
- #14 Backup files public Free
- #16 Directory listing on Free
- #19 HTTP redirects to HTTPS Free
- #20 SSL certificate valid Free
- #22 HSTS header Free
- #23 Content-Security-Policy present Free
- #24 Clickjacking protection Free
- #44 Other AI provider keys exposed Pro
On a domain you verify, Deep tests check the database the app talks to, with the public key it already ships:
- #59 Supabase tables not readable with the public key Deep
- #60 Supabase storage buckets not listable Deep
- #61 Firebase Realtime Database not open to read Deep
- #62 Firestore not open to read Deep
- #63 Firebase Storage not open to list Deep
- #99 Login and sign-up rate limits Deep
What a finding looks like
Your OpenAI API key is visible to anyone
Fix prompt to paste into your builder
Move the OpenAI call out of the front end. Create a server function that reads OPENAI_API_KEY from the project's secrets and calls OpenAI there, and have the page call that function instead. Then revoke the exposed key in the OpenAI dashboard and create a new one, because the old one is public.
Anyone who copies a key like this can run up your bill. That's why it shows as critical, with the masked proof of where we found it.
Free, Pro and Deep
Free
46 tests from your public URL, 2 scans a day, no sign-up. Every finding with proof.
Pro · $10 one time
30 scans within 30 days: all 130 tests, with a fix prompt for every finding and PDF reports.
Deep · part of Pro
17 more tests (147 in total) that touch your database and signed-in users, only on a domain you verify.
Questions
Bolt has its own security audit. Do I still need Securick?
Run Bolt's audit first; it can fix problems inside your project. Securick checks the published app from the outside, the way visitors and attackers see it, and works the same for apps from any builder. The two look from different sides.
Does it work with apps on a bolt.host address?
Yes. Paste the bolt.host address or your custom domain. If you use both, scan the one your users visit: headers and redirects can differ between them.
My app uses the database Bolt set up for me. Can you test it?
Free and Pro scans check what's public, like keys in the code and auth settings. Reading the database itself is a Deep test, which only runs on a domain you verify, using the public key the app already ships.
Is it safe to scan an app that's still in development?
Yes. The free scan only loads what any visitor can load and doesn't write anything. It does need a published, public address: we can't reach a preview that only you can open.
What if my host blocks the scanner?
Some firewalls block automated visitors. The report says so in plain words instead of guessing, and our scanner page explains how to let SecurickBot through.