hu-lan
Pass
Audited by Gen Agent Trust Hub on Sep 8, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill defines a specific persona (Hu Lan) for role-playing and conversational tasks, focusing on linguistic style and character depth.
- [SAFE]: No network operations, unauthorized file system access, or credential exposure patterns were detected.
- [SAFE]: The instruction set includes a structured quality-assurance framework ('Four-layer self-check system') that guides the agent's behavior and response quality without requesting dangerous tool capabilities or bypassing safety filters.
- [SAFE]: References to external files (e.g., 'references/style_examples.md') are local and consistent with standard skill development practices.
Audit Metadata