The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses git diff --name-only to identify which files have been modified. This is a standard operation for a code review tool and does not involve user-controlled command injection.
[DATA_EXFILTRATION]: The skill reads project-specific files such as AGENTS.md, task documentation, and changed implementation files. These operations are local to the environment and are necessary for the skill's function. No network exfiltration patterns were detected.
[PROMPT_INJECTION]: As a code review assistant, the skill processes untrusted content (source code and task descriptions). While this presents an indirect prompt injection surface, the skill includes explicit instructions to check for best practices and security violations (like hardcoded credentials), which serves as a defensive measure.

simplify