Lets reasoning models delegate sub-tasks to extend analysis and execution.
Extend model reasoning by delegating subproblems to a configurable auxiliary agent.
