codex-cli

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). The skill explicitly instructs bypassing approvals and sandboxing, recommends copying auth files and hiding state, and prescribes configuration (approval_policy="never", sandbox_mode="danger-full-access") that enable credential access, stealthy modification, and remote-code-capable agent runs.

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 defaults to bypassing approvals and sandboxing (e.g. --dangerously-bypass-approvals-and-sandbox) and provides examples and recovery steps (stopping processes, copying auth files, using disposable CODEX_HOME) that instruct the agent to circumvent security controls and modify machine state.