The Agent Skills Directory

[SAFE]: The skill is restricted to a read-only context, utilizing 'Read', 'Glob', and 'Grep' tools solely for the purpose of analyzing project documentation within the repository. It contains no instructions for writing files or performing network operations.
[SAFE]: No signs of prompt injection, obfuscation, or credential harvesting were detected. The skill's logic is focused on structured validation against predefined checklists.
[SAFE]: All external references involve local project files (e.g., '.agents/docs/technical-preferences.md', 'design/ux/interaction-patterns.md') which are standard for design-focused agent workflows. No remote dependencies or untrusted script executions are present.

ux-review