diff --git a/sdk/ai/azure-ai-projects/assets.json b/sdk/ai/azure-ai-projects/assets.json
index e71a9f48bcac..4eaa940a0b11 100644
--- a/sdk/ai/azure-ai-projects/assets.json
+++ b/sdk/ai/azure-ai-projects/assets.json
@@ -2,5 +2,5 @@
   "AssetsRepo": "Azure/azure-sdk-assets",
   "AssetsRepoPrefixPath": "python",
   "TagPrefix": "python/ai/azure-ai-projects",
-  "Tag": "python/ai/azure-ai-projects_febb246e47"
+  "Tag": "python/ai/azure-ai-projects_314598932e"
 }
diff --git a/sdk/ai/azure-ai-projects/tests/agents/tools/multitool/test_agent_code_interpreter_and_function.py b/sdk/ai/azure-ai-projects/tests/agents/tools/multitool/test_agent_code_interpreter_and_function.py
index d9894ea572a8..a51bc7b5514e 100644
--- a/sdk/ai/azure-ai-projects/tests/agents/tools/multitool/test_agent_code_interpreter_and_function.py
+++ b/sdk/ai/azure-ai-projects/tests/agents/tools/multitool/test_agent_code_interpreter_and_function.py
@@ -12,10 +12,11 @@
 All tests use the same tool combination but different inputs and workflows.
 """
 
+import json
 from test_base import TestBase, servicePreparer
 from devtools_testutils import recorded_by_proxy, RecordedTransport
 from azure.ai.projects.models import PromptAgentDefinition, CodeInterpreterTool, CodeInterpreterToolAuto, FunctionTool
-
+from openai.types.responses.response_input_param import FunctionCallOutput, ResponseInputParam
 
 class TestAgentCodeInterpreterAndFunction(TestBase):
     """Tests for agents using Code Interpreter + Function Tool combination."""
@@ -25,6 +26,10 @@ class TestAgentCodeInterpreterAndFunction(TestBase):
     def test_calculate_and_save(self, **kwargs):
         """
         Test calculation with Code Interpreter and saving with Function Tool.
+
+        This test verifies that both tools are used:
+        1. Code Interpreter: Performs a calculation that requires actual computation
+        2. Function Tool: Saves the computed result
         """
 
         model = self.test_agents_params["model_deployment_name"]
@@ -36,24 +41,25 @@ def test_calculate_and_save(self, **kwargs):
         # Define function tool
         func_tool = FunctionTool(
             name="save_result",
-            description="Save analysis result",
+            description="Save the calculation result. Must be called to persist the result.",
             parameters={
                 "type": "object",
                 "properties": {
-                    "result": {"type": "string", "description": "The result"},
+                    "calculation": {"type": "string", "description": "Description of the calculation"},
+                    "result": {"type": "string", "description": "The numerical result"},
                 },
-                "required": ["result"],
+                "required": ["calculation", "result"],
                 "additionalProperties": False,
             },
             strict=True,
         )
 
-        # Create agent
+        # Create agent with explicit instructions to use both tools
         agent = project_client.agents.create_version(
             agent_name="code-func-agent",
             definition=PromptAgentDefinition(
                 model=model,
-                instructions="Run calculations and save results.",
+                instructions="You are a calculator assistant. Use code interpreter to perform calculations, then ALWAYS save the result using the save_result function.",
                 tools=[
                     CodeInterpreterTool(container=CodeInterpreterToolAuto()),
                     func_tool,
@@ -63,9 +69,10 @@ def test_calculate_and_save(self, **kwargs):
         )
         print(f"Agent created (id: {agent.id})")
 
-        # Use the agent
+        # Request a calculation that requires Code Interpreter (not trivial math)
+        # 17^4 = 83521 - not something easily computed mentally
         response = openai_client.responses.create(
-            input="Calculate 5 + 3 and save the result.",
+            input="Calculate 17 to the power of 4 using code, then save the result.",
             extra_body={"agent": {"name": agent.name, "type": "agent_reference"}},
         )
         self.validate_response(response)
@@ -79,6 +86,10 @@ def test_calculate_and_save(self, **kwargs):
     def test_generate_data_and_report(self, **kwargs):
         """
         Test generating data with Code Interpreter and reporting with Function.
+
+        This test verifies that both tools are used:
+        1. Code Interpreter: Generates random data and calculates statistics
+        2. Function Tool: Creates a report with the computed statistics
         """
 
         model = self.test_agents_params["model_deployment_name"]
@@ -90,25 +101,27 @@ def test_generate_data_and_report(self, **kwargs):
         # Define function tool
         report_function = FunctionTool(
             name="generate_report",
-            description="Generate a report with the provided data",
+            description="Generate and save a report with the analysis results. Must be called to create the report.",
             parameters={
                 "type": "object",
                 "properties": {
                     "title": {"type": "string", "description": "Report title"},
-                    "summary": {"type": "string", "description": "Report summary"},
+                    "data_count": {"type": "integer", "description": "Number of data points analyzed"},
+                    "average": {"type": "number", "description": "Calculated average value"},
+                    "summary": {"type": "string", "description": "Summary of findings"},
                 },
-                "required": ["title", "summary"],
+                "required": ["title", "data_count", "average", "summary"],
                 "additionalProperties": False,
             },
             strict=True,
         )
 
-        # Create agent
+        # Create agent with explicit instructions
         agent = project_client.agents.create_version(
             agent_name="code-func-report-agent",
             definition=PromptAgentDefinition(
                 model=model,
-                instructions="Generate data using code and create reports with the generate_report function.",
+                instructions="You are a data analyst. Use code interpreter to generate and analyze data, then ALWAYS create a report using the generate_report function with the exact statistics you computed.",
                 tools=[
                     CodeInterpreterTool(container=CodeInterpreterToolAuto()),
                     report_function,
@@ -118,9 +131,9 @@ def test_generate_data_and_report(self, **kwargs):
         )
         print(f"Agent created (id: {agent.id})")
 
-        # Request data generation and report
+        # Request data generation and report - use a fixed seed for reproducibility in verification
         response = openai_client.responses.create(
-            input="Generate a list of 10 random numbers between 1 and 100, calculate their average, and create a report.",
+            input="Using Python with random.seed(42), generate exactly 10 random integers between 1 and 100, calculate their average, and create a report with the results.",
             extra_body={"agent": {"name": agent.name, "type": "agent_reference"}},
         )
 
diff --git a/sdk/ai/azure-ai-projects/tests/agents/tools/multitool/test_agent_file_search_and_code_interpreter.py b/sdk/ai/azure-ai-projects/tests/agents/tools/multitool/test_agent_file_search_and_code_interpreter.py
index 953bb43e037c..fadf656ca7f7 100644
--- a/sdk/ai/azure-ai-projects/tests/agents/tools/multitool/test_agent_file_search_and_code_interpreter.py
+++ b/sdk/ai/azure-ai-projects/tests/agents/tools/multitool/test_agent_file_search_and_code_interpreter.py
@@ -26,6 +26,10 @@ class TestAgentFileSearchAndCodeInterpreter(TestBase):
     def test_find_and_analyze_data(self, **kwargs):
         """
         Test finding data with File Search and analyzing with Code Interpreter.
+
+        This test verifies that both tools are used:
+        1. File Search: Agent finds the data file containing numbers
+        2. Code Interpreter: Agent calculates the average of those numbers
         """
 
         model = self.test_agents_params["model_deployment_name"]
@@ -34,12 +38,36 @@ def test_find_and_analyze_data(self, **kwargs):
         project_client = self.create_client(operation_group="agents", **kwargs)
         openai_client = project_client.get_openai_client()
 
-        # Create data file
-        txt_content = "Sample data: 10, 20, 30, 40, 50"
+        # Create data file with numbers that require actual computation
+        # Numbers: 31, 20, 52, 48, 45, 34, 30, 86, 28, 71, 21, 20, 28, 44, 46
+        # Sum: 604, Count: 15, Average: 40.266... ≈ 40.27
+        # This is impossible to calculate mentally - requires Code Interpreter
+        txt_content = """Sensor Readings Log - Experiment #2847
+
+The following temperature readings (Celsius) were recorded over a 15-hour period:
+
+Hour 1: 31
+Hour 2: 20
+Hour 3: 52
+Hour 4: 48
+Hour 5: 45
+Hour 6: 34
+Hour 7: 30
+Hour 8: 86
+Hour 9: 28
+Hour 10: 71
+Hour 11: 21
+Hour 12: 20
+Hour 13: 28
+Hour 14: 44
+Hour 15: 46
+
+End of sensor log.
+"""
         vector_store = openai_client.vector_stores.create(name="DataStore")
 
         txt_file = BytesIO(txt_content.encode("utf-8"))
-        txt_file.name = "data.txt"
+        txt_file.name = "sensor_readings.txt"
 
         file = openai_client.vector_stores.files.upload_and_poll(
             vector_store_id=vector_store.id,
@@ -47,12 +75,12 @@ def test_find_and_analyze_data(self, **kwargs):
         )
         print(f"File uploaded (id: {file.id})")
 
-        # Create agent
+        # Create agent with explicit instructions to use both tools
         agent = project_client.agents.create_version(
             agent_name="file-search-code-agent",
             definition=PromptAgentDefinition(
                 model=model,
-                instructions="Find data and analyze it.",
+                instructions="You are a data analyst. Use file search to find data files, then use code interpreter to perform calculations on the data.",
                 tools=[
                     FileSearchTool(vector_store_ids=[vector_store.id]),
                     CodeInterpreterTool(container=CodeInterpreterToolAuto()),
@@ -62,9 +90,9 @@ def test_find_and_analyze_data(self, **kwargs):
         )
         print(f"Agent created (id: {agent.id})")
 
-        # Use the agent
+        # Request that requires both tools: find data AND calculate
         response = openai_client.responses.create(
-            input="Find the data file and calculate the average.",
+            input="Find the sensor readings file and use code to calculate the average temperature. Show me the result.",
             extra_body={"agent": {"name": agent.name, "type": "agent_reference"}},
         )
         self.validate_response(response)
@@ -79,7 +107,11 @@ def test_find_and_analyze_data(self, **kwargs):
     @recorded_by_proxy(RecordedTransport.AZURE_CORE, RecordedTransport.HTTPX)
     def test_analyze_code_file(self, **kwargs):
         """
-        Test finding code file and analyzing it.
+        Test finding code file and running it with Code Interpreter.
+
+        This test verifies that both tools are used:
+        1. File Search: Agent finds the Python code file
+        2. Code Interpreter: Agent executes the code and returns the computed result
         """
 
         model = self.test_agents_params["model_deployment_name"]
@@ -88,14 +120,18 @@ def test_analyze_code_file(self, **kwargs):
         project_client = self.create_client(operation_group="agents", **kwargs)
         openai_client = project_client.get_openai_client()
 
-        # Create Python code file
-        python_code = """def fibonacci(n):
+        # Create Python code file with a function that computes a specific value
+        # fibonacci(15) = 610 - this is not a commonly memorized value
+        python_code = """# Fibonacci sequence calculator
+
+def fibonacci(n):
+    \"\"\"Calculate the nth Fibonacci number recursively.\"\"\"
     if n <= 1:
         return n
     return fibonacci(n-1) + fibonacci(n-2)
 
-result = fibonacci(10)
-print(f"Fibonacci(10) = {result}")
+# The code needs to be executed to find what fibonacci(15) equals
+# This is not a commonly known value - it requires actual computation
 """
 
         vector_store = openai_client.vector_stores.create(name="CodeAnalysisStore")
@@ -109,37 +145,46 @@ def test_analyze_code_file(self, **kwargs):
         )
         print(f"Code file uploaded (id: {file.id})")
 
-        # Create agent
+        # Create agent with explicit instructions to run code
         agent = project_client.agents.create_version(
             agent_name="file-search-code-analysis-agent",
             definition=PromptAgentDefinition(
                 model=model,
-                instructions="Find code files and analyze them. You can run code to test it.",
+                instructions="You are a code analyst. Use file search to find code files, then use code interpreter to execute and test the code.",
                 tools=[
                     FileSearchTool(vector_store_ids=[vector_store.id]),
                     CodeInterpreterTool(container=CodeInterpreterToolAuto()),
                 ],
             ),
-            description="Agent for code analysis.",
+            description="Agent for code analysis and execution.",
         )
         print(f"Agent created (id: {agent.id})")
 
-        # Request analysis
+        # Request that requires both tools: find code AND execute it
         response = openai_client.responses.create(
-            input="Find the fibonacci code and explain what it does. What is the computational complexity?",
+            input="Find the fibonacci code file and run it to calculate fibonacci(15). What is the result?",
             extra_body={"agent": {"name": agent.name, "type": "agent_reference"}},
         )
 
         response_text = response.output_text
-        print(f"Response: {response_text[:300]}...")
+        print(f"Response: {response_text[:400]}...")
+
+        # Verify response is meaningful
+        assert len(response_text) > 30, "Expected detailed response"
 
-        assert len(response_text) > 50
+        # Verify File Search was used - response should reference the fibonacci code
         response_lower = response_text.lower()
         assert any(
-            keyword in response_lower for keyword in ["fibonacci", "recursive", "complexity", "exponential"]
-        ), "Expected analysis of fibonacci algorithm"
+            keyword in response_lower for keyword in ["fibonacci", "function", "recursive", "code"]
+        ), f"Expected response to reference the fibonacci code. Got: {response_text[:200]}"
+
+        # Verify Code Interpreter executed the code and got the correct result
+        # fibonacci(15) = 610 - this requires actual execution
+        assert "610" in response_text, f"Expected fibonacci(15) = 610 in response. Got: {response_text[:300]}"
 
-        print("✓ Code file analysis completed")
+        print("[PASS] File Search + Code Interpreter both verified!")
+        print("  - File Search: Found the fibonacci code file")
+        print("  - Code Interpreter: Executed code and computed fibonacci(15) = 610")
 
         # Cleanup
         project_client.agents.delete_version(agent_name=agent.name, agent_version=agent.version)
diff --git a/sdk/ai/azure-ai-projects/tests/agents/tools/multitool/test_agent_file_search_and_function.py b/sdk/ai/azure-ai-projects/tests/agents/tools/multitool/test_agent_file_search_and_function.py
index c2373f750221..ae6d9ad86c70 100644
--- a/sdk/ai/azure-ai-projects/tests/agents/tools/multitool/test_agent_file_search_and_function.py
+++ b/sdk/ai/azure-ai-projects/tests/agents/tools/multitool/test_agent_file_search_and_function.py
@@ -147,7 +147,7 @@ def test_data_analysis_workflow(self, **kwargs):
             )
             print(f"Final response: {response.output_text[:200]}...")
 
-        print("\n✓ Workflow completed successfully")
+        print("\n[PASS] Workflow completed successfully")
 
         # Teardown
         project_client.agents.delete_version(agent_name=agent.name, agent_version=agent.version)
@@ -222,7 +222,7 @@ def test_empty_vector_store_handling(self, **kwargs):
         if response_text:
             assert len(response_text) > 10, "Non-empty response should be meaningful"
 
-        print("\n✓ Agent handled missing data gracefully")
+        print("\n[PASS] Agent handled missing data gracefully")
 
         # Teardown
         project_client.agents.delete_version(agent_name=agent.name, agent_version=agent.version)
@@ -233,7 +233,11 @@ def test_empty_vector_store_handling(self, **kwargs):
     @recorded_by_proxy(RecordedTransport.AZURE_CORE, RecordedTransport.HTTPX)
     def test_python_code_file_search(self, **kwargs):
         """
-        Test searching for Python code files.
+        Test searching for Python code files and saving findings.
+
+        This test verifies that both File Search and Function Tool are used:
+        1. File Search: Agent searches vector store for Python code
+        2. Function Tool: Agent saves the code review findings
         """
 
         model = self.test_agents_params["model_deployment_name"]
@@ -268,7 +272,7 @@ def calculate_sum(numbers):
         # Define function tool
         save_function = FunctionTool(
             name="save_code_review",
-            description="Save code review findings",
+            description="Save code review findings. Must be called to persist the review.",
             parameters={
                 "type": "object",
                 "properties": {
@@ -283,12 +287,12 @@ def calculate_sum(numbers):
             strict=True,
         )
 
-        # Create agent
+        # Create agent with explicit instructions to use both tools
         agent = project_client.agents.create_version(
             agent_name="file-search-function-code-agent",
             definition=PromptAgentDefinition(
                 model=model,
-                instructions="You can search for code files and describe what they do. Save your findings.",
+                instructions="You are a code reviewer. Search for code files, analyze them, and ALWAYS save your findings using the save_code_review function.",
                 tools=[
                     FileSearchTool(vector_store_ids=[vector_store.id]),
                     save_function,
@@ -298,26 +302,55 @@ def calculate_sum(numbers):
         )
         print(f"Agent created (id: {agent.id})")
 
-        # Request code analysis
-        print("\nAsking agent to find and analyze the Python code...")
+        # Request code analysis with explicit save instruction
+        print("\nAsking agent to find, analyze, and save the code review...")
 
         response = openai_client.responses.create(
-            input="Find the Python code file and tell me what the calculate_sum function does.",
+            input="Find the Python code file, explain what the calculate_sum function does, and save your code review findings.",
             extra_body={"agent": {"name": agent.name, "type": "agent_reference"}},
         )
 
-        response_text = response.output_text
-        print(f"Response: {response_text[:300]}...")
+        # Verify function call is made (both tools should be used)
+        function_call_found = False
+        input_list: ResponseInputParam = []
+
+        for item in response.output:
+            if item.type == "function_call":
+                function_call_found = True
+                print(f"Function call detected (name: {item.name})")
+                assert item.name == "save_code_review", f"Expected save_code_review, got {item.name}"
+
+                arguments = json.loads(item.arguments)
+                print(f"Function arguments: {arguments}")
+                assert "findings" in arguments
+                assert len(arguments["findings"]) > 20, "Expected meaningful findings"
 
-        # Verify agent found and analyzed the code
-        assert len(response_text) > 50, "Expected detailed analysis"
+                # Verify findings discuss the code (proves File Search was used)
+                findings_lower = arguments["findings"].lower()
+                assert any(
+                    keyword in findings_lower for keyword in ["sum", "calculate", "function", "numbers", "list", "return"]
+                ), f"Expected findings to discuss the code content. Got: {arguments['findings'][:100]}"
 
-        response_lower = response_text.lower()
-        assert any(
-            keyword in response_lower for keyword in ["sum", "calculate", "function", "numbers", "code", "python"]
-        ), "Expected response to discuss the code"
+                input_list.append(
+                    FunctionCallOutput(
+                        type="function_call_output",
+                        call_id=item.call_id,
+                        output=json.dumps({"status": "saved", "review_id": "review_001"}),
+                    )
+                )
+
+        assert function_call_found, "Expected save_code_review function to be called"
+
+        # Send function results back to get final response
+        response = openai_client.responses.create(
+            input=input_list,
+            previous_response_id=response.id,
+            extra_body={"agent": {"name": agent.name, "type": "agent_reference"}},
+        )
+        response_text = response.output_text
+        print(f"Final response: {response_text[:200] if response_text else '(confirmation)'}...")
 
-        print("\n✓ Agent successfully found code file using File Search")
+        print("\n[PASS] Agent successfully used both File Search and Function Tool")
 
         # Teardown
         project_client.agents.delete_version(agent_name=agent.name, agent_version=agent.version)
@@ -492,7 +525,7 @@ def test_multi_turn_search_and_save_workflow(self, **kwargs):
         response_4_lower = response_4_text.lower()
         assert any(keyword in response_4_lower for keyword in ["bias", "privacy", "ethics", "accountability"])
 
-        print("\n✓ Multi-turn File Search + Function workflow successful!")
+        print("\n[PASS] Multi-turn File Search + Function workflow successful!")
         print("  - Multiple searches across different documents")
         print("  - Function called after context-building searches")
         print("  - Topic switching works correctly")
diff --git a/sdk/ai/azure-ai-projects/tests/agents/tools/multitool/test_agent_file_search_code_interpreter_function.py b/sdk/ai/azure-ai-projects/tests/agents/tools/multitool/test_agent_file_search_code_interpreter_function.py
index ef44a8c644bc..c6df27f34b09 100644
--- a/sdk/ai/azure-ai-projects/tests/agents/tools/multitool/test_agent_file_search_code_interpreter_function.py
+++ b/sdk/ai/azure-ai-projects/tests/agents/tools/multitool/test_agent_file_search_code_interpreter_function.py
@@ -12,6 +12,7 @@
 All tests use the same 3-tool combination but different inputs and workflows.
 """
 
+import json
 from io import BytesIO
 from test_base import TestBase, servicePreparer
 from devtools_testutils import recorded_by_proxy, RecordedTransport
@@ -22,6 +23,7 @@
     CodeInterpreterToolAuto,
     FunctionTool,
 )
+from openai.types.responses.response_input_param import FunctionCallOutput, ResponseInputParam
 
 
 class TestAgentFileSearchCodeInterpreterFunction(TestBase):
@@ -32,6 +34,11 @@ class TestAgentFileSearchCodeInterpreterFunction(TestBase):
     def test_complete_analysis_workflow(self, **kwargs):
         """
         Test complete workflow: find data, analyze it, save results.
+
+        This test verifies that all three tools are used:
+        1. File Search: Agent finds the data file with numerical values
+        2. Code Interpreter: Agent performs statistical calculations on the data
+        3. Function Tool: Agent saves the computed results
         """
 
         model = self.test_agents_params["model_deployment_name"]
@@ -40,12 +47,30 @@ def test_complete_analysis_workflow(self, **kwargs):
         project_client = self.create_client(operation_group="agents", **kwargs)
         openai_client = project_client.get_openai_client()
 
-        # Create data file
-        txt_content = "Sample data for analysis"
+        # Create data file with numbers that require computation
+        # Values: 23, 47, 82, 15, 91, 38, 64, 29, 76, 55
+        # Sum: 520, Count: 10, Average: 52.0, Min: 15, Max: 91
+        txt_content = """Monthly Sales Report - Store #147
+
+The following sales figures (in thousands) were recorded:
+
+January: 23
+February: 47
+March: 82
+April: 15
+May: 91
+June: 38
+July: 64
+August: 29
+September: 76
+October: 55
+
+Please analyze this data for the quarterly review.
+"""
         vector_store = openai_client.vector_stores.create(name="ThreeToolStore")
 
         txt_file = BytesIO(txt_content.encode("utf-8"))
-        txt_file.name = "data.txt"
+        txt_file.name = "sales_report.txt"
 
         file = openai_client.vector_stores.files.upload_and_poll(
             vector_store_id=vector_store.id,
@@ -53,27 +78,30 @@ def test_complete_analysis_workflow(self, **kwargs):
         )
         print(f"File uploaded (id: {file.id})")
 
-        # Define function tool
+        # Define function tool for saving analysis results
         func_tool = FunctionTool(
-            name="save_result",
-            description="Save analysis result",
+            name="save_analysis",
+            description="Save the statistical analysis results. Must be called to persist the analysis.",
             parameters={
                 "type": "object",
                 "properties": {
-                    "result": {"type": "string", "description": "The result"},
+                    "report_name": {"type": "string", "description": "Name of the report analyzed"},
+                    "total": {"type": "number", "description": "Sum of all values"},
+                    "average": {"type": "number", "description": "Average of all values"},
+                    "summary": {"type": "string", "description": "Brief summary of findings"},
                 },
-                "required": ["result"],
+                "required": ["report_name", "total", "average", "summary"],
                 "additionalProperties": False,
             },
             strict=True,
         )
 
-        # Create agent with all three tools
+        # Create agent with all three tools and explicit instructions
         agent = project_client.agents.create_version(
             agent_name="three-tool-agent",
             definition=PromptAgentDefinition(
                 model=model,
-                instructions="Use file search to find data, code interpreter to analyze it, and save_result to save findings.",
+                instructions="You are a data analyst. Use file search to find data files, code interpreter to calculate statistics, and ALWAYS save your analysis using the save_analysis function.",
                 tools=[
                     FileSearchTool(vector_store_ids=[vector_store.id]),
                     CodeInterpreterTool(container=CodeInterpreterToolAuto()),
@@ -84,9 +112,9 @@ def test_complete_analysis_workflow(self, **kwargs):
         )
         print(f"Agent created (id: {agent.id})")
 
-        # Use the agent
+        # Request that requires all three tools
         response = openai_client.responses.create(
-            input="Find the data file, analyze it, and save the results.",
+            input="Find the sales report, use code to calculate the total and average of all monthly sales figures, then save the analysis results.",
             extra_body={"agent": {"name": agent.name, "type": "agent_reference"}},
         )
         self.validate_response(response)
@@ -95,81 +123,3 @@ def test_complete_analysis_workflow(self, **kwargs):
         # Cleanup
         project_client.agents.delete_version(agent_name=agent.name, agent_version=agent.version)
         openai_client.vector_stores.delete(vector_store.id)
-
-    @servicePreparer()
-    @recorded_by_proxy(RecordedTransport.AZURE_CORE, RecordedTransport.HTTPX)
-    def test_four_tools_combination(self, **kwargs):
-        """
-        Test with 4 tools: File Search + Code Interpreter + 2 Functions.
-        """
-
-        model = self.test_agents_params["model_deployment_name"]
-
-        # Setup
-        project_client = self.create_client(operation_group="agents", **kwargs)
-        openai_client = project_client.get_openai_client()
-
-        # Create vector store
-        txt_content = "Test data"
-        vector_store = openai_client.vector_stores.create(name="FourToolStore")
-
-        txt_file = BytesIO(txt_content.encode("utf-8"))
-        txt_file.name = "data.txt"
-
-        file = openai_client.vector_stores.files.upload_and_poll(
-            vector_store_id=vector_store.id,
-            file=txt_file,
-        )
-
-        # Define two function tools
-        func_tool_1 = FunctionTool(
-            name="save_result",
-            description="Save result",
-            parameters={
-                "type": "object",
-                "properties": {
-                    "result": {"type": "string", "description": "The result"},
-                },
-                "required": ["result"],
-                "additionalProperties": False,
-            },
-            strict=True,
-        )
-
-        func_tool_2 = FunctionTool(
-            name="log_action",
-            description="Log an action",
-            parameters={
-                "type": "object",
-                "properties": {
-                    "action": {"type": "string", "description": "Action taken"},
-                },
-                "required": ["action"],
-                "additionalProperties": False,
-            },
-            strict=True,
-        )
-
-        # Create agent with 4 tools
-        agent = project_client.agents.create_version(
-            agent_name="four-tool-agent",
-            definition=PromptAgentDefinition(
-                model=model,
-                instructions="Use all available tools.",
-                tools=[
-                    FileSearchTool(vector_store_ids=[vector_store.id]),
-                    CodeInterpreterTool(container=CodeInterpreterToolAuto()),
-                    func_tool_1,
-                    func_tool_2,
-                ],
-            ),
-            description="Agent with 4 tools.",
-        )
-        print(f"Agent with 4 tools created (id: {agent.id})")
-
-        assert agent.id is not None
-        print("✓ 4 tools works!")
-
-        # Cleanup
-        project_client.agents.delete_version(agent_name=agent.name, agent_version=agent.version)
-        openai_client.vector_stores.delete(vector_store.id)
diff --git a/sdk/ai/azure-ai-projects/tests/agents/tools/multitool/test_multitool_with_conversations.py b/sdk/ai/azure-ai-projects/tests/agents/tools/multitool/test_multitool_with_conversations.py
index 0e3a4cc2e9ee..25a350e35935 100644
--- a/sdk/ai/azure-ai-projects/tests/agents/tools/multitool/test_multitool_with_conversations.py
+++ b/sdk/ai/azure-ai-projects/tests/agents/tools/multitool/test_multitool_with_conversations.py
@@ -149,7 +149,7 @@ def test_file_search_and_function_with_conversation(self, **kwargs):
         )
         print(f"Response 3: {response_3.output_text[:150]}...")
 
-        print("\n✓ Mixed tools with conversation successful!")
+        print("\n[PASS] Mixed tools with conversation successful!")
         print("  - File search (server-side) worked")
         print("  - Function call (client-side) worked")
         print("  - Both tools used in same conversation")
@@ -185,7 +185,7 @@ def test_file_search_and_function_with_conversation(self, **kwargs):
         assert function_calls >= 1, "Expected at least 1 function call (save_report)"
         assert function_outputs >= 1, "Expected at least 1 function output"
 
-        print("\n✓ Multi-tool conversation state verified")
+        print("\n[PASS] Multi-tool conversation state verified")
         print("  - Both server-side (FileSearch) and client-side (Function) tools tracked")
         print("  - All 3 turns preserved in conversation")