The Agent Skills Directory

[SAFE]: The skill implements a legitimate workflow for workspace organization and template generation, following the principle of least privilege by restricting bash tools.
[COMMAND_EXECUTION]: Shell command usage is strictly limited to directory creation (mkdir) and listing (ls) within the local project structure, presenting no privilege escalation or arbitrary execution risk.
[PROMPT_INJECTION]: The skill contains standard procedural instructions for the agent and does not exhibit patterns of safety guideline bypass or role-play injection.
[DATA_EXFILTRATION]: No network operations or access to sensitive system paths (such as credentials or SSH keys) were detected. File access is confined to the project's skill directory.

create-skill