The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructions utilize standard shell commands such as 'ls' and 'rg' (ripgrep) to help the agent discover existing components and project structures within the local repository. these operations are strictly for codebase exploration and do not involve dangerous parameters or network-connected piping.
[SAFE]: The skill is designed for architectural guidance and component reuse. there is no evidence of malicious patterns, data exfiltration, unauthorized network operations, or obfuscation. the dependencies and hooks described are consistent with standard React and Storybook development workflows.

prototype-system