The Agent Skills Directory

[SAFE]: The skill is comprised of documentation and architectural patterns for task delegation and does not include any executable code, scripts, or suspicious operational instructions.
[COMMAND_EXECUTION]: The documentation contains specific safety warnings against using non-interactive shells for external agents, ensuring that user approval prompts remain visible and effective.
[EXTERNAL_DOWNLOADS]: The skill references external design patterns and research from well-known technology organizations and community leaders, which are documented for attribution purposes and do not involve runtime script execution.

subagent-patterns