The Agent Skills Directory

[SAFE]: No prompt injection, bypass instructions, or attempts to override agent safety guidelines were found in the skill content.
[SAFE]: No data exposure or exfiltration risks identified. The skill does not access sensitive local file paths or perform external network requests.
[SAFE]: No obfuscation techniques detected. Content is presented in clear, human-readable markdown without base64, zero-width characters, or hidden encoding.
[SAFE]: No unverifiable dependencies or remote code execution patterns. While the skill conceptually discusses design system implementation and npm packages, it does not perform any installations or execute remote code.
[SAFE]: No privilege escalation or persistence mechanisms found. The skill does not use administrative commands or attempt to modify system configuration files.
[SAFE]: The skill is purely informational, providing design principles and checklists (e.g., pixel perfection, optical alignment) without any dynamic code generation or dangerous execution capabilities.

visual-design