landing-page-design
Pass
Audited by Gen Agent Trust Hub on Nov 11, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill instructions and associated visual system reference file do not contain malicious code, obfuscation, or unauthorized network access patterns.
- [DATA_EXFILTRATION]: The instructions include safety boundaries that prohibit the agent from sending private files to external services or exposing credentials. It explicitly mandates user authorization for any publishing or account permission changes.
- [REMOTE_CODE_EXECUTION]: No remote code execution patterns were detected. The skill notes that it includes no scripts and that fetching any external assets or fonts requires task authorization.
- [PROMPT_INJECTION]: No safety bypass or system prompt extraction attempts were detected. The instruction for local rules to take precedence over framework defaults is limited to design implementation details and is explicitly superseded by user prompts.
Audit Metadata