The Agent Skills Directory

[SAFE]: No security issues detected. The skill uses standard conversational instructions to help users refine designs. It does not contain any executable code, shell commands, or network operations.
[DATA_EXPOSURE_AND_EXFILTRATION]: No network access or sensitive file access patterns detected. The instruction to 'explore the codebase' is a standard agent capability for context gathering and does not involve exfiltration to external domains.
[PROMPT_INJECTION]: No attempts to override system prompts, bypass safety guidelines, or extract sensitive instructions detected.

grill-me