force-delegate
Pass
Audited by Gen Agent Trust Hub on Nov 16, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns or security risks were identified in the skill. The instructions define a coordination logic for task delegation, focusing on supervision, validation, and integration of results from subagents.
- [INDIRECT_PROMPT_INJECTION]: The skill inherently processes data from external sources (subagents). It addresses this risk by instructing the main agent to perform 'decisive validation,' 'final diff inspection,' and to provide delegates with 'fresh compact context' to minimize context leakage. These are standard safety measures for agent-to-agent workflows.
Audit Metadata