agentation
Warn
Audited by Socket on Apr 16, 2026
1 alert found:
AnomalyAnomalySKILL.md
LOWAnomalyLOW
SKILL.md
SUSPICIOUS. The core functionality is coherent for a UI-annotation skill, and most installs/data flows are plausibly aligned with that purpose. However, risk is materially elevated by transitive skill installation, unpinned `npx -y` execution, auto-running agent hooks, autonomous watch-loop behavior, and optional forwarding of annotation data to arbitrary webhook endpoints.
Confidence: 84%Severity: 56%
Audit Metadata