ask-someone-else
Pass
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill instructions are focused on structured content generation and user interaction within the agent's safe environment.
- [SAFE]: No network access, external tool usage, or shell command execution is present in the skill definition.
- [SAFE]: Evaluation cases demonstrate that the skill maintains boundaries, such as refusing to autonomously send messages to external platforms like Slack.
- [SAFE]: No obfuscation, prompt injection, or malicious persistence mechanisms were discovered.
Audit Metadata