principle-never-block-on-the-human

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 instructs the agent to autonomously "proceed" and perform work without asking the human (e.g., "Do X, explain why" and "Reversible actions ... should proceed without blocking"), which encourages making system changes without explicit consent and could lead to modifying system files, creating users, or attempting privilege escalation.