interface-design
Pass
Audited by Gen Agent Trust Hub on Aug 26, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is composed of architectural guidelines and stylistic instructions designed to improve the quality of generated UI code (CSS, HTML, React components). It does not contain executable scripts or binary files.
- [DATA_EXPOSURE_AND_EXFILTRATION]: No network operations, credential harvesting, or access to sensitive local environment files (e.g., .ssh, .aws) were detected in the instructions.
- [REMOTE_CODE_EXECUTION]: There are no patterns involving external package installations or the downloading and execution of remote scripts.
- [INDIRECT_PROMPT_INJECTION]: While the skill ingests user input to generate design proposals and updates a local design system file (
.interface-design/system.md), this is an intended state-management feature for design consistency and does not expose high-privilege capabilities or unsafe data processing. - [OBFUSCATION]: The content is written in clear markdown with no signs of encoding, hidden characters, or steganographic patterns.
Audit Metadata