delegation-intelligence

To coordinate multiple AI agents without context loss, apply explicit delegation requirements including task, scope, return format, success metrics, and constraints. This enforces structured subagent dispatch and mandatory result awaiting before proceeding to the next workflow step.

Choosing between parallel and sequential subagent dispatch depends on task dependencies. Sequential processing enforces mandatory result awaiting and context-first gatekeeping, while parallel processing handles independent tasks concurrently to optimize workflow efficiency.

When delegating tasks to subagents, you must define explicit parameters: task, scope, return format, success metric, acceptance criteria, constraints, and evidence. These requirements ensure proper result handling and prevent context loss during multi-agent coordination.

To integrate context-first gatekeeping before dispatching subagents, ensure you have loaded and applied the gatekeeping skill prior to any delegation. This ensures subagent coordination respects existing planning and context, preventing workflow errors.

Your AI agent skips processing subagent outcomes because it lacks mandatory result awaiting enforcement for sequential tasks. Implementing strict delegation patterns ensures the workflow processes subagent returns and export_cycle usage before proceeding to subsequent steps.