amp-review
Warn
Audited by Snyk on Feb 16, 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 instructs using the --dangerously-allow-all flag to bypass permission prompts, which directs the agent to circumvent security checks and potentially perform privileged or state-changing actions on the host.
Audit Metadata