auto-oversight

CRITICAL E004: Prompt injection detected in skill instructions.

• Potential prompt injection detected (medium risk: 0.30). This skill explicitly instructs fully autonomous system changes (hard-merging PRs, editing git/cron/systemd/harness) and automatic deployments without human approvals, which weakens human oversight and safety boundaries.

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 directs the agent to perform persistent system edits (git/cron/systemd/harness), automated merges/deploys, and writes to on-host state/DBs, i.e. it instructs changing system-level files and services that can alter or compromise the machine's state and may require elevated privileges.