autopilot
Warn
Audited by Socket on Aug 25, 2026
1 alert found:
SecuritySecuritySKILL.md
MEDIUMSecurityMEDIUM
SKILL.md
SUSPICIOUS: the skill is coherent as an autonomous workflow, but its actual footprint is still risky because it enables unattended write/exec behavior, repeated local commits, and exposure to indirect prompt injection from repo content. No clear credential theft or malicious exfiltration appears, but the autonomy level is high enough to warrant caution.
Confidence: 88%Severity: 76%
Audit Metadata