agent-task-shaping
Pass
Audited by Gen Agent Trust Hub on Jun 16, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill contains no executable code or scripts. It functions as a structured reasoning template for scoping AI agent tasks.
- [SAFE]: No external network connections, data exfiltration patterns, or credential harvesting attempts were detected.
- [SAFE]: The content does not utilize obfuscation or prompt injection techniques. It focuses on establishing autonomy levels and tool boundaries for agentic workflows.
Audit Metadata