00 scan complete

Static security for
AI & agentic code.

Seven threat layers. Zero execution. Rendered straight from the last scan in this workspace.

— total findings

— critical

— high

— medium

— low

overview

distribution of detected vulnerabilities by severity

—

critical

data exfiltration, blind agent execution, shell exec, capability inheritance.

—

high

format-string injection, PII patterns, unbounded retrieval, system-prompt exposure.

—

medium

configuration nudges and conservative warnings.

—

low

notes worth a glance during review.

seven static layers — one row per layer

filter by severity, layer, or code — click a row for full detail

codeseveritylayer locationmessage

files with the most findings

every rule shipped with this version