claude-design-skill

Pass

Audited by Gen Agent Trust Hub on Aug 26, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill operates as a benign design system generator, facilitating the creation of HTML, CSS, and JS components based on user requirements. The instructions focus on UI/UX best practices and iterative refinement without introducing suspicious code or external dependencies.
  • [INDIRECT_PROMPT_INJECTION]: The skill represents a surface for indirect prompt injection as it generates code based on untrusted user design descriptions and refinement requests. However, this is a standard risk for code-generation tasks and is mitigated by the skill's focus on structured visual tokens and CSS custom properties.
  • Ingestion points: User-provided design goals, audience descriptions, and iterative tweak requests processed during the skill's execution.
  • Boundary markers: The skill does not explicitly define delimiters for user input, though it use a structured 'Tweaks Protocol' to scope changes.
  • Capability inventory: The agent is instructed to write HTML files with inline CSS and JavaScript to the filesystem.
  • Sanitization: No explicit sanitization or filtering of user input is described in the skill instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 26, 2026, 03:01 PM
Security Audit — agent-trust-hub — claude-design-skill