frontend-design
Pass
Audited by Gen Agent Trust Hub on Feb 21, 2026
Risk Level: SAFE
Full Analysis
- [Prompt Injection] (SAFE): The instructions use words like 'IMPORTANT' and 'CRITICAL', but only to emphasize design quality and aesthetic choices. There are no attempts to bypass AI safety filters or override system instructions.
- [Data Exposure & Exfiltration] (SAFE): No commands for accessing sensitive local files, environment variables, or making external network calls were found.
- [Remote Code Execution] (SAFE): The skill does not download or execute any remote scripts or packages. It is entirely descriptive.
- [Obfuscation] (SAFE): All content is in plain text. No Base64, hidden characters, or encoded commands are present.
- [Indirect Prompt Injection] (SAFE): Although the skill processes user-provided requirements for frontend components, it lacks the technical capabilities (such as file-writing or network access) required to facilitate an exploit chain.
Audit Metadata