ui-designer
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill is designed to operate within a strictly isolated sandbox environment (_design/ or ui-design/), ensuring that generated design artifacts do not interfere with the main project code.
- [SAFE]: It contains explicit prohibitions against modifying production code, routes, package configurations, or global themes, adhering to the principle of least privilege.
- [SAFE]: There are no signs of obfuscation, malicious network exfiltration, or hardcoded credentials within the instructions or assets.
- [INDIRECT_PROMPT_INJECTION]: The skill processes external design briefs and brand materials as part of its core workflow. Ingestion points: user prompts, PRD documents, and reference materials mentioned in SKILL.md. Boundary markers: absent. Capability inventory: restricted to file writes within the sandbox and UI generation tools. Sanitization: absent. As the skill's capabilities are limited to a sandbox, this attack surface is considered low risk and inherent to the skill's purpose.
Audit Metadata