ai-agent-drill-evidence-and-cadence

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

• Attempt to modify system services in skill instructions detected (high risk: 0.80). The skill explicitly instructs platform-level state changes—flipping a global/tenant kill-switch, restoring audit logs, running a cadence enforcer that opens exceptions and pages on-call, and writing/uploading signed evidence packs—operations that alter runtime/operational state even if they don't request sudo, so it should be flagged.