Local_date_fix_updated

Author: UtkarshMishra-Microsoft

src/backend/app_kernel.py

@@ -85,6 +85,25 @@
 async def input_task_endpoint(input_task: InputTask, request: Request):
     """
     Receive the initial input task from the user.
+    
+    ---
+    parameters:
+      - name: body
+        in: body
+        required: true
+        schema:
+          type: object
+          properties:
+            session_id:
+              type: string
+              description: Session ID for the task
+            description:
+              type: string
+              description: Description of the task to be performed
+            user_locale:
+              type: string
+              description: User's locale preference (e.g., 'en_US', 'en_GB', 'de_DE')
+              default: 'en_GB'
     """
     # Fix 1: Properly await the async rai_success function
     if not await rai_success(input_task.description):
@@ -115,6 +134,9 @@ async def input_task_endpoint(input_task: InputTask, request: Request):
     if not input_task.session_id:
         input_task.session_id = str(uuid.uuid4())
 
+    # Extract user locale from request headers or input task
+    user_locale = request.headers.get("X-User-Locale", input_task.user_locale or "en_GB")
+
     try:
         # Create all agents instead of just the planner agent
         # This ensures other agents are created first and the planner has access to them
@@ -132,6 +154,7 @@ async def input_task_endpoint(input_task: InputTask, request: Request):
             user_id=user_id,
             memory_store=memory_store,
             client=client,
+            user_locale=user_locale,  # ✅ Add this
         )
 
         group_chat_manager = agents[AgentType.GROUP_CHAT_MANAGER.value]
@@ -229,6 +252,10 @@ async def human_feedback_endpoint(human_feedback: HumanFeedback, request: Reques
             user_id:
               type: string
               description: The user ID providing the feedback
+            user_locale:
+              type: string
+              description: User's locale preference (e.g., 'en_US', 'en_GB', 'de_DE')
+              default: 'en_GB'
     responses:
       200:
         description: Feedback received successfully
@@ -256,6 +283,9 @@ async def human_feedback_endpoint(human_feedback: HumanFeedback, request: Reques
         human_feedback.session_id, user_id
     )
 
+    # Extract user locale from request headers or feedback object
+    user_locale = request.headers.get("X-User-Locale", human_feedback.user_locale or "en_GB")
+
     client = None
     try:
         client = config.get_ai_project_client()
@@ -268,6 +298,7 @@ async def human_feedback_endpoint(human_feedback: HumanFeedback, request: Reques
         user_id=user_id,
         memory_store=memory_store,
         client=client,
+        user_locale=user_locale,  # ✅ Add this
     )
 
     if human_agent is None:
@@ -476,6 +507,10 @@ async def approve_step_endpoint(
     kernel, memory_store = await initialize_runtime_and_context(
         human_feedback.session_id, user_id
     )
+    
+    # Extract user locale from request headers or feedback object
+    user_locale = request.headers.get("X-User-Locale", human_feedback.user_locale or "en_GB")
+    
     client = None
     try:
         client = config.get_ai_project_client()
@@ -486,6 +521,7 @@ async def approve_step_endpoint(
         user_id=user_id,
         memory_store=memory_store,
         client=client,
+        user_locale=user_locale,  # ✅ Add this
     )
 
     # Send the approval to the group chat manager
@@ -970,6 +1006,187 @@ async def get_agent_tools():
     return []
 
 
+@app.post("/api/tasks")
+async def handle_task(request: Request):
+    """
+    Handle individual tasks with user locale support.
+    This endpoint demonstrates how to pass user_locale to agent tools.
+
+    Expected request body:
+    {
+        "task_type": "schedule_orientation",
+        "employee_name": "John Doe",
+        "date": "2025-07-15",
+        "user_locale": "en_US"
+    }
+    """
+    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")
+
+    try:
+        data = await request.json()
+        user_locale = data.get("user_locale", "en_GB")  # fallback to en_GB
+        task_type = data.get("task_type")
+
+        # Import the HR tools
+        from kernel_tools.hr_tools import HrTools
+
+        if task_type == "schedule_orientation":
+            employee_name = data.get("employee_name")
+            date = data.get("date")
+
+            if not employee_name or not date:
+                raise HTTPException(status_code=400, detail="employee_name and date are required")
+
+            # Example: Call the HR tool function with user_locale
+            result = await HrTools.schedule_orientation_session(
+                employee_name=employee_name,
+                date=date,
+                user_locale=user_locale
+            )
+
+            track_event_if_configured(
+                "TaskExecuted",
+                {
+                    "task_type": task_type,
+                    "user_locale": user_locale,
+                    "employee_name": employee_name,
+                    "date": date,
+                },
+            )
+
+            return {
+                "status": "success",
+                "result": result,
+                "user_locale": user_locale
+            }
+
+        elif task_type == "schedule_performance_review":
+            employee_name = data.get("employee_name")
+            date = data.get("date")
+
+            if not employee_name or not date:
+                raise HTTPException(status_code=400, detail="employee_name and date are required")
+
+            result = await HrTools.schedule_performance_review(
+                employee_name=employee_name,
+                date=date,
+                user_locale=user_locale
+            )
+
+            return {
+                "status": "success",
+                "result": result,
+                "user_locale": user_locale
+            }
+
+        elif task_type == "process_leave_request":
+            employee_name = data.get("employee_name")
+            leave_type = data.get("leave_type")
+            start_date = data.get("start_date")
+            end_date = data.get("end_date")
+
+            if not all([employee_name, leave_type, start_date, end_date]):
+                raise HTTPException(status_code=400, detail="employee_name, leave_type, start_date, and end_date are required")
+
+            result = await HrTools.process_leave_request(
+                employee_name=employee_name,
+                leave_type=leave_type,
+                start_date=start_date,
+                end_date=end_date,
+                user_locale=user_locale
+            )
+
+            return {
+                "status": "success",
+                "result": result,
+                "user_locale": user_locale
+            }
+
+        elif task_type == "add_mobile_extras_pack":
+            # Example for Product tools
+            from kernel_tools.product_tools import ProductTools
+
+            new_extras_pack_name = data.get("new_extras_pack_name")
+            start_date = data.get("start_date")
+
+            if not new_extras_pack_name or not start_date:
+                raise HTTPException(status_code=400, detail="new_extras_pack_name and start_date are required")
+
+            result = await ProductTools.add_mobile_extras_pack(
+                new_extras_pack_name=new_extras_pack_name,
+                start_date=start_date,
+                user_locale=user_locale
+            )
+
+            return {
+                "status": "success",
+                "result": result,
+                "user_locale": user_locale
+            }
+
+        else:
+            raise HTTPException(status_code=400, detail=f"Unsupported task_type: {task_type}")
+
+    except Exception as e:
+        track_event_if_configured(
+            "TaskError",
+            {
+                "error": str(e),
+                "task_type": data.get("task_type") if 'data' in locals() else "unknown",
+            },
+        )
+        raise HTTPException(status_code=400, detail=f"Error executing task: {e}")
+
+
+@app.get("/api/tasks/examples")
+async def get_task_examples():
+    """
+    Get examples of how to use the /api/tasks endpoint with different locales.
+    
+    Returns example request payloads for testing the user_locale functionality.
+    """
+    examples = {
+        "schedule_orientation_us": {
+            "task_type": "schedule_orientation",
+            "employee_name": "John Doe",
+            "date": "2025-07-15",
+            "user_locale": "en_US"
+        },
+        "schedule_orientation_uk": {
+            "task_type": "schedule_orientation", 
+            "employee_name": "Jane Smith",
+            "date": "2025-07-15",
+            "user_locale": "en_GB"
+        },
+        "process_leave_request": {
+            "task_type": "process_leave_request",
+            "employee_name": "Alice Johnson", 
+            "leave_type": "Annual Leave",
+            "start_date": "2025-08-01",
+            "end_date": "2025-08-15",
+            "user_locale": "en_GB"
+        },
+        "add_mobile_extras": {
+            "task_type": "add_mobile_extras_pack",
+            "new_extras_pack_name": "International Roaming Pack",
+            "start_date": "2025-07-20",
+            "user_locale": "en_US"
+        }
+    }
+    
+    return {
+        "message": "Example requests for /api/tasks endpoint",
+        "examples": examples,
+        "usage": "POST /api/tasks with any of the example payloads to see locale-specific date formatting"
+    }
+
+
 # Run the app
 if __name__ == "__main__":
     import uvicorn

src/backend/kernel_agents/agent_base.py

@@ -86,11 +86,12 @@ def default_system_message(agent_name=None) -> str:
         name = agent_name
         return f"You are an AI assistant named {name}. Help the user by providing accurate and helpful information."
 
-    async def handle_action_request(self, action_request: ActionRequest) -> str:
+    async def handle_action_request(self, action_request: ActionRequest, **kwargs) -> str:
         """Handle an action request from another agent or the system.
 
         Args:
-            action_request_json: The action request as a JSON string
+            action_request: The action request object
+            **kwargs: Additional keyword arguments (e.g., user_locale)
 
         Returns:
             A JSON string containing the action response

src/backend/kernel_agents/agent_factory.py

@@ -21,6 +21,13 @@
 from kernel_agents.product_agent import ProductAgent
 from kernel_agents.tech_support_agent import TechSupportAgent
 from models.messages_kernel import AgentType, PlannerResponsePlan
+# Import all tool classes to generate agent_tools_list
+from kernel_tools.hr_tools import HrTools
+from kernel_tools.marketing_tools import MarketingTools
+from kernel_tools.product_tools import ProductTools
+from kernel_tools.procurement_tools import ProcurementTools
+from kernel_tools.tech_support_tools import TechSupportTools
+from kernel_tools.generic_tools import GenericTools
 # pylint:disable=E0611
 from semantic_kernel.agents.azure_ai.azure_ai_agent import AzureAIAgent
 
@@ -190,6 +197,7 @@ async def create_all_agents(
         temperature: float = 0.0,
         memory_store: Optional[CosmosMemoryContext] = None,
         client: Optional[Any] = None,
+        user_locale: str = "en_GB",  # ✅ Add this
     ) -> Dict[AgentType, BaseAgent]:
         """Create all agent types for a session in a specific order.
 
@@ -279,13 +287,18 @@ async def create_all_agents(
         agents[planner_agent_type] = planner_agent
 
         # Phase 3: Create group chat manager with all agents including the planner
+        # Generate the agent tools list for the GroupChatManager
+        agent_tools_list = cls._generate_agent_tools_list()
+        
         group_chat_manager = await cls.create_agent(
             agent_type=AgentType.GROUP_CHAT_MANAGER,
             session_id=session_id,
             user_id=user_id,
             temperature=temperature,
             client=client,
             agent_instances=agent_instances,  # Pass agent instances to the planner
+            agent_tools_list=agent_tools_list,  # ✅ Add this
+            user_locale=user_locale,  # ✅ Add this
         )
         agents[group_chat_manager_type] = group_chat_manager
 
@@ -327,3 +340,34 @@ def clear_cache(cls, session_id: Optional[str] = None) -> None:
             cls._agent_cache.clear()
             cls._azure_ai_agent_cache.clear()
             logger.info("Cleared all agent caches")
+
+    @classmethod
+    def _generate_agent_tools_list(cls) -> list[str]:
+        """Generate a list of all available tool names across all agents.
+        
+        Returns:
+            List of tool names that can be used by agents
+        """
+        tool_classes = [
+            HrTools,
+            MarketingTools, 
+            ProductTools,
+            ProcurementTools,
+            TechSupportTools,
+            GenericTools,
+        ]
+        
+        all_tools = []
+        for tool_class in tool_classes:
+            # Get all methods from the tool class
+            for name in dir(tool_class):
+                # Skip private methods and special methods
+                if not name.startswith("_") and name != "generate_tools_json_doc":
+                    # Check if it's a callable method
+                    method = getattr(tool_class, name)
+                    if callable(method):
+                        all_tools.append(f"{tool_class.__name__}.{name}")
+        
+        return all_tools
+
+