The Agent Skills Directory

[PROMPT_INJECTION]: None detected. The instructions define a specific conversational persona for design review and do not attempt to bypass agent safety filters or override core behaviors.
[DATA_EXFILTRATION]: None detected. There are no network operations, hardcoded domains, or instructions to send data to external services.
[REMOTE_CODE_EXECUTION]: None detected. The skill consists entirely of natural language instructions and does not download or execute external scripts.
[COMMAND_EXECUTION]: None detected. No shell commands or system-level calls are present in the skill.
[CREDENTIALS_UNSAFE]: None detected. No hardcoded API keys, tokens, or sensitive file paths were found.
[NO_CODE]: The skill contains only markdown instructions and configuration metadata, with no executable scripts or complex logic.

grill-me