007
Warn
Audited by Socket on Jul 15, 2026
1 alert found:
AnomalyAnomalySKILL.md
LOWAnomalyLOW
SKILL.md
SUSPICIOUS: the skill is purpose-aligned and shows no obvious exfiltration or malicious install path, but it materially increases an AI agent's offensive/security-review capability and includes automatic monitoring behavior over arbitrary projects. Risk is driven more by powerful audit/red-team scope than by malicious data flows.
Confidence: 83%Severity: 68%
Audit Metadata