feat(mma): Complete Phase 7 implementation: MMA Dashboard, HITL Step Modal, and Memory Mutator

multi_agent_conductor.py

@@ -102,17 +102,50 @@ class ConductorEngine:
                     model_name="gemini-2.5-flash-lite", 
                     messages=[]
                 )
-                run_worker_lifecycle(ticket, context)
+                run_worker_lifecycle(ticket, context, event_queue=self.event_queue)
                 await self._push_state(active_tier="Tier 2 (Tech Lead)")
 
-def confirm_execution(payload: str) -> bool:
+def confirm_execution(payload: str, event_queue: events.AsyncEventQueue, ticket_id: str) -> bool:
     """
-    Placeholder for external confirmation function.
-    In a real scenario, this might trigger a UI prompt.
+    Pushes an approval request to the GUI and waits for response.
     """
-    return True
+    import threading
+    import time
+    import asyncio
+    # We use a list container so the GUI can inject the actual Dialog object back to us
+    # since the dialog is created in the GUI thread.
+    dialog_container = [None]
+    
+    task = {
+        "action": "mma_step_approval",
+        "ticket_id": ticket_id,
+        "payload": payload,
+        "dialog_container": dialog_container
+    }
+    
+    # Push to queue
+    try:
+        loop = asyncio.get_event_loop()
+        if loop.is_running():
+            asyncio.run_coroutine_threadsafe(event_queue.put("mma_step_approval", task), loop)
+        else:
+            event_queue._queue.put_nowait(("mma_step_approval", task))
+    except Exception:
+        # Fallback if no loop
+        event_queue._queue.put_nowait(("mma_step_approval", task))
 
-def run_worker_lifecycle(ticket: Ticket, context: WorkerContext, context_files: List[str] = None):
+    # Wait for the GUI to create the dialog and for the user to respond
+    start = time.time()
+    while dialog_container[0] is None and time.time() - start < 60:
+        time.sleep(0.1)
+    
+    if dialog_container[0]:
+        approved, final_payload = dialog_container[0].wait()
+        return approved
+    
+    return False
+
+def run_worker_lifecycle(ticket: Ticket, context: WorkerContext, context_files: List[str] = None, event_queue: events.AsyncEventQueue = None):
     """
     Simulates the lifecycle of a single agent working on a ticket.
     Calls the AI client and updates the ticket status based on the response.
@@ -150,11 +183,17 @@ def run_worker_lifecycle(ticket: Ticket, context: WorkerContext, context_files:
     
     # In a real scenario, we would pass md_content from the aggregator
     # and manage the conversation history in the context.
+    # HITL Clutch: pass the queue and ticket_id to confirm_execution
+    def clutch_callback(payload: str) -> bool:
+        if not event_queue:
+            return True
+        return confirm_execution(payload, event_queue, ticket.id)
+
     response = ai_client.send(
         md_content="", 
         user_message=user_message,
         base_dir=".",
-        pre_tool_callback=confirm_execution if ticket.step_mode else None,
+        pre_tool_callback=clutch_callback if ticket.step_mode else None,
         qa_callback=ai_client.run_tier4_analysis
     )