tmux-start

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 0.90). This skill explicitly launches an external AI process with the "--dangerously-skip-permissions" flag and offers to copy .env/secret files into the new worktree, creating a deliberate and high-risk pathway for sensitive data exposure or exfiltration by the launched service.

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

• Attempt to modify system services in skill instructions detected (high risk: 0.80). This skill instructs creating and modifying repo worktrees and copying environment files and—critically—launching Claude with the --dangerously-skip-permissions flag (an explicit permission-bypass), which encourages bypassing agent security and can expose secrets even though it doesn't require sudo or edit system files.