The Agent Skills Directory

[SAFE]: The skill is a documentation and framework-heavy tool designed for UX quality assurance. It does not perform any unauthorized file system modifications or network operations.
[PROMPT_INJECTION]: The instructions are strictly focused on framework enforcement (Value, Agency, Identity, Resilience, Echo) and do not contain any instructions to override safety filters or system prompts.
[DATA_EXFILTRATION]: No patterns indicative of credential harvesting or data exfiltration to external domains were found. References to regulatory bodies (FTC, EU) are purely for domain knowledge.
[OBFUSCATION]: All content is in plain text. No Base64, hex-encoding, or hidden Unicode characters were detected.
[REMOTE_CODE_EXECUTION]: There are no external dependencies, package installations, or remote script fetches.
[COMMAND_EXECUTION]: The skill has a strong constraint to never write or modify code, and no shell command patterns were detected.
[INDIRECT_PROMPT_INJECTION]: While the skill processes design specifications for review, its lack of code execution or network write capabilities makes it a low-risk surface for exploitation via indirect injection.

warden