frontend-design
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill contains instructional Markdown content only. No evidence of obfuscation, malicious code, or unauthorized network activity was found during the analysis.- [COMMAND_EXECUTION]: The skill suggests using the standard
npx shadcn@latest addcommand for UI component initialization. This involves running a well-known tool within a standard development environment.- [INDIRECT_PROMPT_INJECTION]: The skill identifies an attack surface by reading external project context from.tasks/domain.md. - Ingestion points: Instructions specify reading
.tasks/domain.mdfor brand details and terminology. - Boundary markers: The agent is directed to extract specific, limited types of information (colors, roles, statuses), which serves as a natural constraint.
- Capability inventory: The skill permits the agent to write and modify frontend source files (CSS, React).
- Sanitization: Content is parsed and applied to UI tokens rather than being directly executed, reducing the risk of malicious payload triggering.
Audit Metadata