The Agent Skills Directory

Command Execution: The skill executes a local script, launch_codex_security_mcp, to resolve security policy paths and manage policy chains. This is an intended functionality used to automate the discovery of security-related files within the user's repository.- Indirect Prompt Injection Surface: The skill processes external, untrusted data including source code, existing policy files, and prior findings. While this presents a surface for indirect prompt injection, the instructions include specific guidance to treat these inputs only as evidence and not as a source of authority for privileged actions, which effectively mitigates the risk.

define-security-policy