overnight
Warn
Audited by Socket on Mar 18, 2026
1 alert found:
SecuritySecuritySKILL.md
MEDIUMSecurityMEDIUM
SKILL.md
SUSPICIOUS. The skill’s behavior matches its stated purpose, so it is not overtly deceptive or credential-harvesting, but it enables high-autonomy code execution: automatic worker spawning, arbitrary shell-based completion checks, broad repo access, and continued self-directed task generation without further approval. This is a coherent but high-risk automation skill rather than confirmed malware.
Confidence: 86%Severity: 76%
Audit Metadata