dispatching-parallel-agents
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is a set of markdown instructions providing a conceptual framework for task delegation. It does not include any executable scripts, network operations, or file system modifications.
- [SAFE]: No credential exposure or sensitive data access patterns were identified. The instructions focus on managing agent context and isolation.
- [SAFE]: No obfuscation or prompt injection attempts were found. The use of instructional emphasis (e.g., 'Core principle') is standard for defining workflows.
- [SAFE]: While the skill describes a pattern for processing external data (test failures) which constitutes an indirect prompt injection surface, it specifically advocates for best practices like isolated contexts and precisely scoped instructions, which mitigates these risks.
Audit Metadata