The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to search for and execute project-specific test configurations (e.g., from package.json, pytest.ini, or Makefile). This is a legitimate part of a code review workflow to verify that changes do not break existing functionality.
[DATA_EXPOSURE]: As part of its security dimension assessment, the skill actively checks for hardcoded credentials, secrets, or tokens in the codebase. This is a security-enhancing feature designed to prevent data exposure.
[INDIRECT_PROMPT_INJECTION]: The skill processes external project code and test outputs. While it lacks explicit delimiters or boundary markers for this ingested data, its primary function is to analyze the content for defects rather than execute instructions contained within the source code.

VibeCoding Review