Free until we catch a medium-or-above vulnerability in your code or live site. No card to start. No seat counts. No trial timer.
Until we catch your first real (medium+) vulnerability.
For founders shipping live apps.
For founders with real users.
For indie hackers shipping multiple apps.
| Feature | Day 1 ($0) | Solo | Startup | Business |
|---|---|---|---|---|
| Unlimited scans | ✓ | ✓ | ✓ | ✓ |
| Full findings + auto-fix | ✓ | ✓ | ✓ | ✓ |
| MCP + VS Code + git hook + CLI | ✓ | ✓ | ✓ | ✓ |
| Projects | unlimited | unlimited | unlimited | unlimited |
| Manual probes / mo | — | 1 | 2 | unlimited |
| Probe → IDE feedback loop | ✓ | ✓ | ✓ | ✓ |
| CVSS + CWE per finding | ✓ | ✓ | ✓ | ✓ |
| Auto-scan (CVE / supply / secrets) | — | — | weekly | daily |
| Authenticated probe | — | — | opt-in | default-on |
| Priority probe queue | — | — | — | ✓ |
| Support | community | priority < 24 hr |
What's a probe?
A live security scan against your deployed site. We discover routes, fuzz inputs, check for common runtime vulns (XSS, IDOR, SSRF, broken auth), then route the findings back to your IDE so your AI fixes them in-source.
Why are scans unlimited?
Scans are cheap for us to run. We'd rather you scan everything than worry about quota. The real cost (and value) is in probing your deployed app — that's where pricing lives.
Can I switch plans anytime?
Yes. Upgrade applies immediately; downgrade takes effect at the end of your billing cycle. No prorating headaches.
Scan vs probe?
Scan = look at code (your editor, your git, your AI's output). Probe = look at your deployed app from outside. Scans catch "I shouldn't have written that"; probes catch "I didn't realize the deployed config was wrong."
Can I cancel?
Anytime, in your dashboard. You keep access until the period ends; after that the scanner keeps working (still free), but probes stop until you re-subscribe.
Do you read my code?
We scan it on a stateless backend. We don't persist file contents — only the audit log of decisions + findings ties back to your account. We never train models on your code.