Use apply_action in the Cortex decision framework?

Is it possible to use apply_action of the Articulation to move the robot joints directly in the Cortex framework? For example, I have the following DfState that tries to move the joints directly:

class dummpyMoveRobots(DfState):
    def __init__(self):
        super().__init__()
        
    def enter(self):
        super().enter()
    
    def step(self):
        appAction = self.context.robot.get_applied_action()
        print(f"prev action: {appAction.joint_positions}")
        action = ArticulationAction(joint_positions=np.array([0.0, 0.0, 0.0, 0.0, 0.0, 0.0]))
        self.context.robot.apply_action(action)
        appAction = self.context.robot.get_applied_action()
        print(f"applied action: {appAction.joint_positions}")
        return self

When I run this DfState it doesn’t apply the specified action at all. Instead it applies actions that are computed in the MotionCommander. In the step function of omni.isaac.cortex.motion.commander, it computes an action that is based on the specified motion command, and applies that action to the robot. I guess that overwites the action specified to the robot in the above code block. Is there any way to control joints directly in the cortex framework?