orchestrate

Multi-agent task orchestration automates complex workflows by managing agent spawning and routing tasks based on status. It handles planning, execution, and review phases autonomously without manual intervention.

You can automate a software development task by starting a new named task from the active directory. The orchestration skill will autonomously manage planner, executor, and code reviewer agents until completion or blockage.

Autonomous agent spawning requires the Task tool with subagent_type configured. If the Task tool is unavailable, a Bash CLI fallback can be used to execute the multi-agent workflow.

The best way to route tasks to specialized agents is through autonomous orchestration. It intelligently routes tasks to the correct agent based on the current status of the task throughout its lifecycle.

Multi-agent workflows stop autonomously when a task becomes blocked during the planning, execution, or review phases. The orchestration process manages the lifecycle until the task is either complete or blocked.

Yes, a Bash CLI fallback is supported for agent execution if the Task tool with subagent_type is unavailable. This ensures the multi-agent task workflow can still proceed through its required phases.