Evidence before explanation
Deterministic checks produce the findings first. AI can explain and prioritise them, but it cannot invent scan results.
BetaApplications
AI Security Assistant
Inspect security headers, transport, cookies, CORS, source-code patterns and dependency hygiene with evidence-backed findings. The scanner is deliberately conservative: it asks for authorization, limits requests, blocks internal targets by default, and states what it did not test.
Start a reviewYou’ll be asked to create an account, then land straight in the workspace.
WHO IT IS FOR
- Developers reviewing a site before launch
- Teams triaging source and dependency risk
- Owners who need actionable remediation rather than a raw scanner dump
WHAT YOU CAN BUILD
- Passive website security reviews
- Source-code security analysis
- Dependency hygiene and advisory checks
KEY CAPABILITIES
03 ENTRIESSafe by default
Website scans use low-volume requests, strict timeouts, scope enforcement and SSRF protections. No credential attacks or destructive tests.
Honest coverage
Every scan records passed checks, limitations, scanner version and whether advisory intelligence was available.
EXAMPLE OUTPUTS
BRIEF → RESULTBRIEF
“Review the headers and cookies on my staging site.”
RESULT
A scoped, low-impact website review with evidence, severity, confidence and fix verification.
BRIEF
“Check this package.json before release.”
RESULT
Pinned-version hygiene checks plus real OSV advisory lookups when the advisory service is reachable.
NOTEThese describe what the generator produces for a brief of that shape. Output varies by run — the specification is planned each time, not retrieved from a template.
Build your first security review.
One account covers every SXX product. Your projects are saved as you work and reopen exactly where you left them.