The Agent Skills Directory

[SAFE]: The skill instructions are focused entirely on providing a structured analysis of a codebase. It does not include any commands for network access, file modification, or external execution.
[DATA_EXFILTRATION]: While the skill instructs the agent to read environment and configuration files, this is done within the context of a security audit to identify risks like hardcoded secrets. No mechanisms for exfiltrating this data are present.
[COMMAND_EXECUTION]: The skill includes an explicit constraint: "Do NOT modify any code during this audit. Analysis only.", which prevents unauthorized file changes or script execution.

project-review