The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes the codex exec command (SKILL.md) to interact with an external implementation reviewer. This pattern involves spawning a subprocess, which is a sensitive operational capability.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface. Ingestion points: Codebase context and user implementation plans are ingested to form the consultation prompt (SKILL.md). Boundary markers: No explicit boundary markers or instructions to ignore embedded commands are present in the command template to isolate the prompt content from the shell command structure. Capability inventory: Subprocess execution via codex exec (SKILL.md). Sanitization: The skill lacks automated sanitization or shell-escaping for the strings interpolated into the command execution block.

ask-codex