ux-designer
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted user input to define product context, user tasks, and data models. This information is used to generate high-fidelity design specifications and information architecture reports. The lack of explicit boundary markers or sanitization logic presents a surface for indirect prompt injection if the output is used in automated downstream processes. Ingestion points: SKILL.md (Section 3: Required Inputs). Boundary markers: None identified. Capability inventory: Generates text-based specifications and references local assets; no file-writing or network exfiltration found. Sanitization: No evidence of input validation or escaping for user-provided context.
- [EXTERNAL_DOWNLOADS]: The skill contains a commented-out reference to the Google Fonts API in assets/design-system-core.css to load the Inter font family, targeting a well-known and trusted service for design assets.
Audit Metadata