advise-legal-work-skills
Pass
Audited by Gen Agent Trust Hub on Jun 22, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns detected. The skill functions solely as a set of instructions for a 'Legal Work Skill Advisor' and does not contain scripts, binaries, or automated shell commands.
- [SAFE]: The skill incorporates domain-specific safety measures, explicitly instructing the agent to require human legal review for conclusions and to avoid presenting AI-generated output as final legal advice.
- [SAFE]: Privacy and confidentiality are addressed through instructions to protect privileged material and minimize exposure of confidential data during the design process.
- [SAFE]: The workflow emphasizes a 'read-only inspection first' policy, which ensures the agent asks for explicit user permission before making edits, running commands, or publishing content.
Audit Metadata