interface-design
Pass
Audited by Gen Agent Trust Hub on Oct 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill is a collection of design principles, workflows, and evaluation criteria aimed at improving the quality of AI-generated user interfaces. It does not contain executable scripts, external dependencies, or network-enabled tools. The content is purely instructional and focused on UI/UX craft.
- [INDIRECT_PROMPT_INJECTION]: The skill defines a workflow where the agent ingests user-supplied "Intent" and "Product Domain" descriptions to generate UI code and writes design patterns to a local
.interface-design/system.mdfile. This creates a surface where user-provided instructions could influence the agent's output or system state, though the impact is restricted to design-related configuration. - Ingestion points: User prompts describing design requirements, target users, and product domain concepts (SKILL.md).
- Boundary markers: The skill requires the agent to explicitly state the "Intent" and design tokens before writing code, providing a clear context for interpreting user requirements and separating metadata from implementation.
- Capability inventory: File-write access for creating and updating
.interface-design/system.md(validation.md, SKILL.md). - Sanitization: No explicit sanitization or validation of user-supplied intent strings or design tokens before they are persisted to the system file.
Audit Metadata