coding-agent

Warn

Audited by Snyk on Feb 21, 2026

Risk Level: MEDIUM
Full Analysis

MEDIUM W013: Attempt to modify system services in skill instructions.

  • Attempt to modify system services in skill instructions detected (high risk: 0.90). The skill explicitly encourages disabling sandboxes and approvals (flags like --yolo, --dangerously-skip-permissions, and examples using --yolo for installs/modifying system files) which instructs agents to bypass security controls and perform unrestricted system writes, creating a high risk of compromising the host state.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 21, 2026, 02:20 PM