dispatching-parallel-agents
Pass
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides a structured methodology for parallelizing independent tasks using sub-agents. It explicitly recommends isolating sub-agent context ("They should never inherit your session's context or history"), which aligns with the principle of least privilege and reduces risks associated with shared state.
- [SAFE]: No remote code execution or external dependency patterns were detected. The example commands for task creation are standard platform-specific calls for agent delegation.
- [SAFE]: No evidence of obfuscation, credential harvesting, or unauthorized network operations was found. The skill operates entirely within the expected instruction-based scope.
- [SAFE]: The metadata and core instructions are consistent with the stated purpose of improving multi-tasking efficiency during development sessions.
Audit Metadata