feat: add get_current_state workflow query to BaseWorkflow

Adds a default @workflow.query handler named "get_current_state"
to BaseWorkflow. Automatically returns the state machine's current
state if self.state_machine exists (covers all StateMachine-based
agents without any override needed). Falls back to _workflow_state
string for non-state-machine agents.

Enables external callers to detect turn completion:
  "waiting_for_input" = agent is done, ready for next message
  "researching" = agent is still working

Companion to scaleapi/scale-agentex#173 which adds
GET /tasks/{task_id}/query/{query_name} endpoint.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

Author: gabrycina

src/agentex/lib/core/temporal/workflows/workflow.py

@@ -15,6 +15,22 @@ def __init__(
         display_name: str,
     ):
         self.display_name = display_name
+        self._workflow_state: str = "initialized"
+
+    @workflow.query(name="get_current_state")
+    def get_current_state(self) -> str:
+        """Query handler for the current workflow state.
+
+        If the workflow uses a StateMachine (self.state_machine),
+        returns the state machine's current state automatically.
+        Otherwise returns the fallback _workflow_state string.
+
+        This enables external callers to detect when the agent has
+        finished processing a turn (state == "waiting_for_input").
+        """
+        if hasattr(self, "state_machine") and self.state_machine is not None:
+            return self.state_machine.get_current_state()
+        return self._workflow_state
 
     @abstractmethod
     @workflow.signal(name=SignalName.RECEIVE_EVENT)