automation
Warn
Audited by Socket on Apr 12, 2026
1 alert found:
AnomalyAnomalySKILL.md
LOWAnomalyLOW
SKILL.md
SUSPICIOUS. The core skill is purpose-aligned and uses official OpenHands API endpoints, so it is not overtly malicious. Risk comes from autonomous scheduled execution with full sandbox/secrets access and from transitive trust in external plugins fetched from arbitrary repos or git URLs, especially when mutable refs are allowed.
Confidence: 85%Severity: 66%
Audit Metadata