The Agent Skills Directory

[SAFE]: The core instructions in SKILL.md are focused on design philosophy and quality control. There are no attempts to override agent behavior, bypass safety protocols, or manipulate the underlying system.
[SAFE]: No sensitive data access or exfiltration patterns were detected. Credentials are not hardcoded, and the skill documentation correctly describes using standard environment variables for API keys in the evaluation pipeline.
[SAFE]: The skill references multiple external resources from well-known and trusted sources, such as GitHub repositories from Vercel, Anthropic, and reputable design experts. These are documented neutrally as design references.
[SAFE]: No obfuscation, hidden content, or deceptive metadata was found. The code examples provided for HTML, CSS, and GLSL are standard implementation patterns for the described visual effects.
[SAFE]: The evaluation script run_eval.py depends on common, legitimate Python packages (anthropic, playwright, scipy, httpx) and is used for design benchmarking as described in the documentation.
[SAFE]: There are no signs of persistence mechanisms, privilege escalation, or dynamic code execution from untrusted sources. The skill adheres to best practices for frontend development and design system documentation.

minoan-frontend-design