frontend-design
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [Prompt Injection] (SAFE): The use of 'CRITICAL' and 'IMPORTANT' tags are functional markers for the AI's creative direction. There are no attempts to bypass safety filters, extract system prompts, or override agent constraints.
- [Data Exposure & Exfiltration] (SAFE): The skill does not access local files, environment variables, or hardcoded secrets. It contains no network operations.
- [Remote Code Execution] (SAFE): No external dependencies are downloaded or executed. While the skill suggests using libraries like 'Motion' in generated code, it does not perform any installations itself.
- [Obfuscation] (SAFE): No encoded strings, zero-width characters, or hidden content were detected.
- [Persistence & Privilege Escalation] (SAFE): The skill does not interact with the operating system, shell profiles, or system services.
Audit Metadata