Start Authorized Assessment

Dry run is selected by default. Real run executes bounded passive recon plus AI Gateway Hunt/Validate work after authorization, protected context, and typed confirmation are present.

Setup Wizard

Target base URLRepository slugAllowed host

Selected model and attack classes

All 44

Fable 5 · 44 classes · injection, broken-access-control, cryptographic-failures, security-misconfiguration, supply-chain-failures + 39 more

Max run budget (USD)

Runtime Credentials

Server credentials are used first. Browser-session keys are optional one-action overrides and are never rendered back.

0 session keys

Checking server credential presence before showing browser-session overrides.

PlatPhorm API keyemptyVercel AI Gateway keyemptyVercel tokenempty

Optional direct-provider session overrides

Anthropic API keyemptyOpenAI API keyempty

Result

No run launched yet.