fix: merging dev changes to main branch

.github/workflows/deploy.yml

@@ -21,7 +21,7 @@ jobs:
           export AZURE_TENANT_ID=${{ secrets.AZURE_TENANT_ID }}
           export AZURE_CLIENT_SECRET=${{ secrets.AZURE_CLIENT_SECRET }}
           export AZURE_SUBSCRIPTION_ID="${{ secrets.AZURE_SUBSCRIPTION_ID }}"
-          export GPT_MIN_CAPACITY="50"
+          export GPT_MIN_CAPACITY="140"
           export AZURE_REGIONS="${{ vars.AZURE_REGIONS }}"
 
           chmod +x infra/scripts/checkquota.sh

src/backend/app_kernel.py

@@ -1,6 +1,7 @@
 # app_kernel.py
 import asyncio
 import logging
+import os
 import uuid
 from typing import Dict, List, Optional
 
@@ -9,6 +10,7 @@
 from auth.auth_utils import get_authenticated_user_details
 
 # Azure monitoring
+from azure.monitor.opentelemetry import configure_azure_monitor
 from config_kernel import Config
 from event_utils import track_event_if_configured
 
@@ -32,19 +34,19 @@
 # Updated import for KernelArguments
 from utils_kernel import initialize_runtime_and_context, rai_success
 
-# # Check if the Application Insights Instrumentation Key is set in the environment variables
-# connection_string = os.getenv("APPLICATIONINSIGHTS_CONNECTION_STRING")
-# if connection_string:
-#     # Configure Application Insights if the Instrumentation Key is found
-#     configure_azure_monitor(connection_string=connection_string)
-#     logging.info(
-#         "Application Insights configured with the provided Instrumentation Key"
-#     )
-# else:
-#     # Log a warning if the Instrumentation Key is not found
-#     logging.warning(
-#         "No Application Insights Instrumentation Key found. Skipping configuration"
-#     )
+# Check if the Application Insights Instrumentation Key is set in the environment variables
+connection_string = os.getenv("APPLICATIONINSIGHTS_CONNECTION_STRING")
+if connection_string:
+    # Configure Application Insights if the Instrumentation Key is found
+    configure_azure_monitor(connection_string=connection_string)
+    logging.info(
+        "Application Insights configured with the provided Instrumentation Key"
+    )
+else:
+    # Log a warning if the Instrumentation Key is not found
+    logging.warning(
+        "No Application Insights Instrumentation Key found. Skipping configuration"
+    )
 
 # Configure logging
 logging.basicConfig(level=logging.INFO)
@@ -137,9 +139,8 @@ async def input_task_endpoint(input_task: InputTask, request: Request):
         # Convert input task to JSON for the kernel function, add user_id here
 
         # Use the planner to handle the task
-        result = await group_chat_manager.handle_input_task(input_task)
+        await group_chat_manager.handle_input_task(input_task)
 
-        print(f"Result: {result}")
         # Get plan from memory store
         plan = await memory_store.get_plan_by_session(input_task.session_id)
 
@@ -176,7 +177,6 @@ async def input_task_endpoint(input_task: InputTask, request: Request):
         }
 
     except Exception as e:
-        logging.exception(f"Error handling input task: {e}")
         track_event_if_configured(
             "InputTaskError",
             {
@@ -779,18 +779,17 @@ async def delete_all_messages(request: Request) -> Dict[str, str]:
     authenticated_user = get_authenticated_user_details(request_headers=request.headers)
     user_id = authenticated_user["user_principal_id"]
     if not user_id:
+        track_event_if_configured(
+            "UserIdNotFound", {"status_code": 400, "detail": "no user"}
+        )
         raise HTTPException(status_code=400, detail="no user")
 
     # Initialize memory context
     kernel, memory_store = await initialize_runtime_and_context("", user_id)
 
-    logging.info("Deleting all plans")
     await memory_store.delete_all_items("plan")
-    logging.info("Deleting all sessions")
     await memory_store.delete_all_items("session")
-    logging.info("Deleting all steps")
     await memory_store.delete_all_items("step")
-    logging.info("Deleting all agent_messages")
     await memory_store.delete_all_items("agent_message")
 
     # Clear the agent factory cache
@@ -840,6 +839,9 @@ async def get_all_messages(request: Request):
     authenticated_user = get_authenticated_user_details(request_headers=request.headers)
     user_id = authenticated_user["user_principal_id"]
     if not user_id:
+        track_event_if_configured(
+            "UserIdNotFound", {"status_code": 400, "detail": "no user"}
+        )
         raise HTTPException(status_code=400, detail="no user")
 
     # Initialize memory context

src/backend/context/cosmos_memory_kernel.py

@@ -84,7 +84,6 @@ async def initialize(self):
                 id=self._cosmos_container,
                 partition_key=PartitionKey(path="/session_id"),
             )
-            logging.info("Successfully connected to CosmosDB")
         except Exception as e:
             logging.error(
                 f"Failed to initialize CosmosDB container: {e}. Continuing without CosmosDB for testing."

src/backend/handlers/runtime_interrupt_kernel.py

@@ -55,10 +55,6 @@ async def on_message(
         Returns:
             The original message (for pass-through functionality)
         """
-        print(
-            f"NeedsUserInputHandler received message: {message} from sender: {sender_type}/{sender_key}"
-        )
-
         if isinstance(message, GetHumanInputMessage):
             self.question_for_human = message
             self.messages.append(
@@ -67,7 +63,6 @@ async def on_message(
                     "content": message.content,
                 }
             )
-            print("Captured question for human in NeedsUserInputHandler")
         elif isinstance(message, GroupChatMessage):
             # Ensure we extract content consistently with the original implementation
             content = (
@@ -81,7 +76,6 @@ async def on_message(
                     "content": content,
                 }
             )
-            print(f"Captured group chat message in NeedsUserInputHandler - {message}")
         elif isinstance(message, dict) and "content" in message:
             # Handle messages directly from AzureAIAgent
             self.question_for_human = GetHumanInputMessage(content=message["content"])
@@ -91,7 +85,6 @@ async def on_message(
                     "content": message["content"],
                 }
             )
-            print("Captured question from AzureAIAgent in NeedsUserInputHandler")
 
         return message
 
@@ -111,7 +104,6 @@ def get_messages(self) -> List[Dict[str, Any]]:
         """Get captured messages and clear buffer."""
         messages = self.messages.copy()
         self.messages.clear()
-        print("Returning and clearing captured messages in NeedsUserInputHandler")
         return messages
 
 
@@ -133,38 +125,28 @@ async def on_message(self, message: Any, sender_type: str = None) -> Any:
         Returns:
             The original message (for pass-through functionality)
         """
-        print(
-            f"on_message called in AssistantResponseHandler with message from sender: {sender_type} - {message}"
-        )
-
         if hasattr(message, "body") and sender_type in ["writer", "editor"]:
             # Ensure we're handling the content consistently with the original implementation
             self.assistant_response = (
                 message.body.content
                 if hasattr(message.body, "content")
                 else str(message.body)
             )
-            print("Assistant response set in AssistantResponseHandler")
         elif isinstance(message, dict) and "value" in message and sender_type:
             # Handle message from AzureAIAgent
             self.assistant_response = message["value"]
-            print(
-                "Assistant response from AzureAIAgent set in AssistantResponseHandler"
-            )
 
         return message
 
     @property
     def has_response(self) -> bool:
         """Check if response is available."""
         has_response = self.assistant_response is not None
-        print(f"has_response called, returning: {has_response}")
         return has_response
 
     def get_response(self) -> Optional[str]:
         """Get captured response."""
         response = self.assistant_response
-        print(f"get_response called, returning: {response}")
         return response
 
 
@@ -201,7 +183,6 @@ def register_handlers(kernel: sk.Kernel, session_id: str) -> tuple:
     kernel.set_variable(f"input_handler_{session_id}", user_input_handler)
     kernel.set_variable(f"response_handler_{session_id}", assistant_handler)
 
-    print(f"Registered handlers for session {session_id} with kernel")
     return user_input_handler, assistant_handler