gentle-design
Pass
Audited by Gen Agent Trust Hub on Oct 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes potentially untrusted data from multiple sources to influence its design output, creating a surface for indirect instructions to be executed by the agent.
- Ingestion points: The agent reads user-provided comments from
design/<slug>/comments.jsonand scans project source files (such as CSS custom properties, Tailwind configurations, and component libraries) during the design system synchronization process. - Boundary markers: There are no explicit boundary markers or instructions to the agent to disregard embedded natural language instructions within the UI comments or within the design tokens extracted from source code.
- Capability inventory: The skill possesses the ability to create and modify files (HTML, JSON, Markdown) and interact with the
gentle_designtool for state management and local previews. - Sanitization: No evidence of sanitization, escaping, or validation of the text content within comments or design tokens is provided before the data is used to generate or edit HTML artboards.
Audit Metadata