diff --git a/sdk/ai/azure-ai-projects/assets.json b/sdk/ai/azure-ai-projects/assets.json
index 4eaa940a0b11..c5814cde3f02 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_314598932e"
+  "Tag": "python/ai/azure-ai-projects_93d1dc0fe7"
 }
diff --git a/sdk/ai/azure-ai-projects/samples/agents/tools/computer_use_util.py b/sdk/ai/azure-ai-projects/samples/agents/tools/computer_use_util.py
index 46e3cb6803e4..c1793779b0fe 100644
--- a/sdk/ai/azure-ai-projects/samples/agents/tools/computer_use_util.py
+++ b/sdk/ai/azure-ai-projects/samples/agents/tools/computer_use_util.py
@@ -159,4 +159,4 @@ def print_final_output(response):
                 final_output += getattr(part, "text", None) or getattr(part, "refusal", None) or "" + "\n"
 
     print(f"Final status: {response.status}")
-    print(f"==> Result: {final_output.strip()}")
+    print(f"Final result: {final_output.strip()}")
diff --git a/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_ai_search.py b/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_ai_search.py
index 0b227f224754..263aee360715 100644
--- a/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_ai_search.py
+++ b/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_ai_search.py
@@ -108,7 +108,7 @@
                             )
         elif event.type == "response.completed":
             print(f"\nFollow-up completed!")
-            print(f"==> Result: {event.response.output_text}")
+            print(f"Agent response: {event.response.output_text}")
 
     print("\nCleaning up...")
     project_client.agents.delete_version(agent_name=agent.name, agent_version=agent.version)
diff --git a/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_code_interpreter.py b/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_code_interpreter.py
index 9cd10f63af2d..db84258efae8 100644
--- a/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_code_interpreter.py
+++ b/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_code_interpreter.py
@@ -108,7 +108,6 @@
         file_path = os.path.join(tempfile.gettempdir(), filename)
         with open(file_path, "wb") as f:
             f.write(file_content.read())
-        # Print result (should contain "file")
-        print(f"==> Result: file, {file_path} downloaded successfully.")
+        print(f"File downloaded successfully: {file_path}")
     else:
         print("No file generated in response")
diff --git a/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_code_interpreter_async.py b/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_code_interpreter_async.py
index be88a1f9555c..b794599b70af 100644
--- a/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_code_interpreter_async.py
+++ b/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_code_interpreter_async.py
@@ -112,8 +112,7 @@ async def main() -> None:
                 file_path = os.path.join(tempfile.gettempdir(), filename)
                 with open(file_path, "wb") as f:
                     f.write(file_content.read())
-                # Print result (should contain "file")
-                print(f"==> Result: file, {file_path} downloaded successfully.")
+                print(f"File downloaded successfully: {file_path}")
             else:
                 print("No file generated in response")
 
diff --git a/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_file_search.py b/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_file_search.py
index 78b69f7e0e33..9351a29a8977 100644
--- a/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_file_search.py
+++ b/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_file_search.py
@@ -78,7 +78,7 @@
         input="Tell me about Contoso products",
         extra_body={"agent": {"name": agent.name, "type": "agent_reference"}},
     )
-    print(f"==> Result: {response.output_text}")
+    print(f"Agent response: {response.output_text}")
     print("\nCleaning up...")
     project_client.agents.delete_version(agent_name=agent.name, agent_version=agent.version)
     print("Agent deleted")
diff --git a/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_file_search_in_stream.py b/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_file_search_in_stream.py
index 48728d150870..21c04ba22bf6 100644
--- a/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_file_search_in_stream.py
+++ b/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_file_search_in_stream.py
@@ -140,7 +140,7 @@
                             print(f"File Citation - Filename: {annotation.filename}, File ID: {annotation.file_id}")
         elif event.type == "response.completed":
             print(f"\nFollow-up completed!")
-            print(f"==> Result: {event.response.output_text}")
+            print(f"Agent response: {event.response.output_text}")
 
     # Clean up resources
     print("\n" + "=" * 60)
diff --git a/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_file_search_in_stream_async.py b/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_file_search_in_stream_async.py
index 3c5e22d1b440..31280352ac7c 100644
--- a/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_file_search_in_stream_async.py
+++ b/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_file_search_in_stream_async.py
@@ -145,7 +145,7 @@ async def main() -> None:
                                 print(f"File Citation - Filename: {annotation.filename}, File ID: {annotation.file_id}")
             elif event.type == "response.completed":
                 print(f"\nFollow-up completed!")
-                print(f"==> Result: {event.response.output_text}")
+                print(f"Agent response: {event.response.output_text}")
 
         # Clean up resources
         print("\n" + "=" * 60)
diff --git a/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_function_tool.py b/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_function_tool.py
index 97d349bec751..cfa1dad9ba7f 100644
--- a/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_function_tool.py
+++ b/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_function_tool.py
@@ -108,8 +108,7 @@ def get_horoscope(sign: str) -> str:
         extra_body={"agent": {"name": agent.name, "type": "agent_reference"}},
     )
 
-    # Print result (should contain "Tuesday")
-    print(f"==> Result: {response.output_text}")
+    print(f"Agent response: {response.output_text}")
 
     print("\nCleaning up...")
     project_client.agents.delete_version(agent_name=agent.name, agent_version=agent.version)
diff --git a/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_function_tool_async.py b/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_function_tool_async.py
index 2ff5c44d108f..9a70726fb2cf 100644
--- a/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_function_tool_async.py
+++ b/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_function_tool_async.py
@@ -109,8 +109,7 @@ async def main():
             extra_body={"agent": {"name": agent.name, "type": "agent_reference"}},
         )
 
-        # Print result (should contain "Tuesday")
-        print(f"==> Result: {response.output_text}")
+        print(f"Agent response: {response.output_text}")
 
 
 if __name__ == "__main__":
diff --git a/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_image_generation.py b/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_image_generation.py
index 2f9764f2ba95..d901af701df4 100644
--- a/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_image_generation.py
+++ b/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_image_generation.py
@@ -96,5 +96,4 @@
             f.write(base64.b64decode(image_data[0]))
 
         # [END download_image]
-        # Print result (should contain "file")
-        print(f"==> Result: Image downloaded and saved to file: {file_path}")
+        print(f"Image saved to: {file_path}")
diff --git a/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_image_generation_async.py b/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_image_generation_async.py
index 4202bee8a8b1..aeb2e099695c 100644
--- a/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_image_generation_async.py
+++ b/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_image_generation_async.py
@@ -91,8 +91,7 @@ async def main():
             with open(file_path, "wb") as f:
                 f.write(base64.b64decode(image_data[0]))
 
-            # Print result (should contain "file")
-            print(f"==> Result: Image downloaded and saved to file: {file_path}")
+            print(f"Image saved to: {file_path}")
 
 
 if __name__ == "__main__":
diff --git a/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_mcp.py b/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_mcp.py
index 4be58ad9c32a..17fcbe97df6f 100644
--- a/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_mcp.py
+++ b/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_mcp.py
@@ -95,8 +95,7 @@
         extra_body={"agent": {"name": agent.name, "type": "agent_reference"}},
     )
 
-    # Print result (should contain "Azure")
-    print(f"==> Result: {response.output_text}")
+    print(f"Agent response: {response.output_text}")
 
     # Clean up resources by deleting the agent version
     # This prevents accumulation of unused agent versions in your project
diff --git a/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_mcp_async.py b/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_mcp_async.py
index 9c1ae981c5b1..ddcbc3e4614c 100644
--- a/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_mcp_async.py
+++ b/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_mcp_async.py
@@ -99,8 +99,7 @@ async def main():
             extra_body={"agent": {"name": agent.name, "type": "agent_reference"}},
         )
 
-        # Print result (should contain "Azure")
-        print(f"==> Result: {response.output_text}")
+        print(f"Agent response: {response.output_text}")
 
         # Clean up resources by deleting the agent version
         # This prevents accumulation of unused agent versions in your project
diff --git a/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_openapi.py b/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_openapi.py
index 781056555036..a95740e2e53f 100644
--- a/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_openapi.py
+++ b/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_openapi.py
@@ -76,8 +76,7 @@
         input="Use the OpenAPI tool to print out, what is the weather in Seattle, WA today.",
         extra_body={"agent": {"name": agent.name, "type": "agent_reference"}},
     )
-    # Print result (should contain "\u00b0F")
-    print(f"==> Result: {response.output_text}")
+    print(f"Agent response: {response.output_text}")
 
     print("\nCleaning up...")
     project_client.agents.delete_version(agent_name=agent.name, agent_version=agent.version)
diff --git a/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_sharepoint.py b/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_sharepoint.py
index 71b6ae49168a..c41be0e489d2 100644
--- a/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_sharepoint.py
+++ b/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_sharepoint.py
@@ -101,7 +101,7 @@
                             )
         elif event.type == "response.completed":
             print(f"\nFollow-up completed!")
-            print(f"==> Result: {event.response.output_text}")
+            print(f"Agent response: {event.response.output_text}")
 
     print("Cleaning up...")
     project_client.agents.delete_version(agent_name=agent.name, agent_version=agent.version)
diff --git a/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_web_search.py b/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_web_search.py
index 3a08a869c62a..7ef45abec3dd 100644
--- a/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_web_search.py
+++ b/sdk/ai/azure-ai-projects/samples/agents/tools/sample_agent_web_search.py
@@ -65,7 +65,7 @@
         input="Show me the latest London Underground service updates",
         extra_body={"agent": {"name": agent.name, "type": "agent_reference"}},
     )
-    print(f"==> Result: {response.output_text}")
+    print(f"Agent response: {response.output_text}")
 
     print("\nCleaning up...")
     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_code_interpreter_and_function.py b/sdk/ai/azure-ai-projects/tests/agents/tools/multitool/test_agent_code_interpreter_and_function.py
index a51bc7b5514e..e4e42fe991f5 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
@@ -18,6 +18,7 @@
 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."""
 
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 ae6d9ad86c70..f95f7f21af1a 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
@@ -328,7 +328,8 @@ def calculate_sum(numbers):
                 # 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"]
+                    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]}"
 
                 input_list.append(
diff --git a/sdk/ai/azure-ai-projects/tests/agents/tools/test_agent_ai_search.py b/sdk/ai/azure-ai-projects/tests/agents/tools/test_agent_ai_search.py
index 5ee41aea715f..188b24806409 100644
--- a/sdk/ai/azure-ai-projects/tests/agents/tools/test_agent_ai_search.py
+++ b/sdk/ai/azure-ai-projects/tests/agents/tools/test_agent_ai_search.py
@@ -19,6 +19,7 @@
 # The tests in this file rely on an existing Azure AI Search project connection that has been populated with the following document:
 # https://arxiv.org/pdf/2508.03680
 
+
 class TestAgentAISearch(TestBase):
 
     # Test questions with expected answers
diff --git a/sdk/ai/azure-ai-projects/tests/agents/tools/test_agent_ai_search_async.py b/sdk/ai/azure-ai-projects/tests/agents/tools/test_agent_ai_search_async.py
index 71525b089398..6f050e553de1 100644
--- a/sdk/ai/azure-ai-projects/tests/agents/tools/test_agent_ai_search_async.py
+++ b/sdk/ai/azure-ai-projects/tests/agents/tools/test_agent_ai_search_async.py
@@ -20,6 +20,7 @@
 # The tests in this file rely on an existing Azure AI Search project connection that has been populated with the following document:
 # https://arxiv.org/pdf/2508.03680
 
+
 class TestAgentAISearchAsync(TestBase):
 
     # Test questions with expected answers
diff --git a/sdk/ai/azure-ai-projects/tests/conftest.py b/sdk/ai/azure-ai-projects/tests/conftest.py
index ad54a729211e..a074884c8063 100644
--- a/sdk/ai/azure-ai-projects/tests/conftest.py
+++ b/sdk/ai/azure-ai-projects/tests/conftest.py
@@ -19,7 +19,7 @@
     add_general_regex_sanitizer,
     add_body_key_sanitizer,
     add_remove_header_sanitizer,
-    add_general_string_sanitizer,
+    add_body_regex_sanitizer,
 )
 
 if not load_dotenv(find_dotenv(), override=True):
@@ -120,6 +120,14 @@ def sanitize_url_paths():
         value="/subscriptions/00000000-0000-0000-0000-000000000000/resourceGroups/00000/providers/Microsoft.MachineLearningServices/workspaces/00000/connections/connector-name",
     )
 
+    # Sanitize print output from sample validation to prevent replay failures when print statements change
+    # Only targets the validation Responses API call by matching the unique input prefix
+    add_body_key_sanitizer(
+        json_path="$.input",
+        value="sanitized-print-output",
+        regex=r"print contents array = .*",
+    )
+
     # Remove Stainless headers from OpenAI client requests, since they include platform and OS specific info, which we can't have in recorded requests.
     # Here is an example of all the `x-stainless` headers from a Responses call:
     #   x-stainless-arch: other:amd64
diff --git a/sdk/ai/azure-ai-projects/tests/samples/test_samples.py b/sdk/ai/azure-ai-projects/tests/samples/test_samples.py
index a60773a0b691..edfa548ab38d 100644
--- a/sdk/ai/azure-ai-projects/tests/samples/test_samples.py
+++ b/sdk/ai/azure-ai-projects/tests/samples/test_samples.py
@@ -3,20 +3,34 @@
 # Copyright (c) Microsoft Corporation.
 # Licensed under the MIT License.
 # ------------------------------------
-import csv, os, pytest, re, inspect, sys
+import csv, os, pytest, re, inspect, sys, json
 import importlib.util
 import unittest.mock as mock
+from typing import Union, cast, overload
+from azure.core.credentials import TokenCredential
+from azure.core.credentials_async import AsyncTokenCredential
 from azure.core.exceptions import HttpResponseError
 from devtools_testutils.aio import recorded_by_proxy_async
 from devtools_testutils import AzureRecordedTestCase, recorded_by_proxy, RecordedTransport
 from test_base import servicePreparer, patched_open_crlf_to_lf
 from pytest import MonkeyPatch
 from azure.ai.projects import AIProjectClient
+from azure.ai.projects.aio import AIProjectClient as AsyncAIProjectClient
+from azure.identity import DefaultAzureCredential
+from azure.identity.aio import DefaultAzureCredential as AsyncDefaultAzureCredential
+from pydantic import BaseModel
 
 
 class SampleExecutor:
     """Helper class for executing sample files with proper environment setup and credential mocking."""
 
+    class TestReport(BaseModel):
+        """Schema for validation test report."""
+
+        model_config = {"extra": "forbid"}
+        correct: bool
+        reason: str
+
     def __init__(
         self, test_instance: "AzureRecordedTestCase", sample_path: str, env_var_mapping: dict[str, str], **kwargs
     ):
@@ -52,92 +66,147 @@ def _capture_print(self, *args, **kwargs):
         self.print_calls.append(" ".join(str(arg) for arg in args))
         self._original_print(*args, **kwargs)
 
-    def execute(self):
+    def _get_mock_credential(self, is_async: bool):
+        """Get a mock credential that supports context manager protocol."""
+        credential_instance = self.test_instance.get_credential(AIProjectClient, is_async=is_async)
+        if is_async:
+            patch_target = "azure.identity.aio.DefaultAzureCredential"
+        else:
+            patch_target = "azure.identity.DefaultAzureCredential"
+
+        # Create a mock that returns a context manager wrapping the credential
+        mock_credential_class = mock.MagicMock()
+        mock_credential_class.return_value.__enter__ = mock.MagicMock(return_value=credential_instance)
+        mock_credential_class.return_value.__exit__ = mock.MagicMock(return_value=None)
+        mock_credential_class.return_value.__aenter__ = mock.AsyncMock(return_value=credential_instance)
+        mock_credential_class.return_value.__aexit__ = mock.AsyncMock(return_value=None)
+
+        return mock.patch(patch_target, new=mock_credential_class)
+
+    def execute(self, enable_llm_validation: bool = True):
         """Execute a synchronous sample with proper mocking and environment setup."""
 
         with (
             MonkeyPatch.context() as mp,
-            mock.patch("builtins.print", side_effect=self._capture_print),
-            mock.patch("builtins.open", side_effect=patched_open_crlf_to_lf),
-            mock.patch("azure.identity.DefaultAzureCredential") as mock_credential,
+            self._get_mock_credential(is_async=False),
         ):
             for var_name, var_value in self.env_vars.items():
                 mp.setenv(var_name, var_value)
-            credential_instance = self.test_instance.get_credential(AIProjectClient, is_async=False)
-            credential_mock = mock.MagicMock()
-            credential_mock.__enter__.return_value = credential_instance
-            credential_mock.__exit__.return_value = False
-            mock_credential.return_value = credential_mock
-            if self.spec.loader is None:
-                raise ImportError(f"Could not load module {self.spec.name} from {self.sample_path}")
-            self.spec.loader.exec_module(self.module)
-
-        self._validate_output()
-
-    async def execute_async(self):
+
+            with (
+                mock.patch("builtins.print", side_effect=self._capture_print),
+                mock.patch("builtins.open", side_effect=patched_open_crlf_to_lf),
+            ):
+                if self.spec.loader is None:
+                    raise ImportError(f"Could not load module {self.spec.name} from {self.sample_path}")
+                self.spec.loader.exec_module(self.module)
+
+            if enable_llm_validation:
+                self._validate_output()
+
+    async def execute_async(self, enable_llm_validation: bool = True):
         """Execute an asynchronous sample with proper mocking and environment setup."""
+
         with (
             MonkeyPatch.context() as mp,
-            mock.patch("builtins.print", side_effect=self._capture_print),
-            mock.patch("builtins.open", side_effect=patched_open_crlf_to_lf),
-            mock.patch("azure.identity.aio.DefaultAzureCredential") as mock_credential,
+            self._get_mock_credential(is_async=True),
         ):
             for var_name, var_value in self.env_vars.items():
                 mp.setenv(var_name, var_value)
 
-            # Create a mock credential that supports async context manager protocol
-            credential_instance = self.test_instance.get_credential(AIProjectClient, is_async=True)
-            credential_mock = mock.AsyncMock()
-            credential_mock.__aenter__.return_value = credential_instance
-            credential_mock.__aexit__.return_value = False
-            mock_credential.return_value = credential_mock
-            if self.spec.loader is None:
-                raise ImportError(f"Could not load module {self.spec.name} from {self.sample_path}")
-            self.spec.loader.exec_module(self.module)
-            await self.module.main()
+            with (
+                mock.patch("builtins.open", side_effect=patched_open_crlf_to_lf),
+                mock.patch("builtins.print", side_effect=self._capture_print),
+            ):
+                if self.spec.loader is None:
+                    raise ImportError(f"Could not load module {self.spec.name} from {self.sample_path}")
+                self.spec.loader.exec_module(self.module)
+                await self.module.main()
 
-        self._validate_output()
+            if enable_llm_validation:
+                await self._validate_output_async()
+
+    def _get_validation_request_params(self) -> dict:
+        """Get common parameters for validation request."""
+        return {
+            "model": "gpt-4o",
+            "instructions": """We just run Python code and captured a Python array of print statements.
+Validating the printed content to determine if correct or not:
+Respond false if any entries show:
+- Error messages or exception text
+- Empty or null results where data is expected
+- Malformed or corrupted data
+- Timeout or connection errors
+- Warning messages indicating failures
+- Failure to retrieve or process data
+- Statements saying documents/information didn't provide relevant data
+- Statements saying unable to find/retrieve information
+- Asking the user to specify, clarify, or provide more details
+- Suggesting to use other tools or sources
+- Asking follow-up questions to complete the task
+- Indicating lack of knowledge or missing information
+- Responses that defer answering or redirect the question
+Respond with true only if the result provides a complete, substantive answer with actual data/information.
+Always respond with `reason` indicating the reason for the response.""",
+            "text": {
+                "format": {
+                    "type": "json_schema",
+                    "name": "TestReport",
+                    "schema": self.TestReport.model_json_schema(),
+                }
+            },
+            # The input field is sanitized in recordings (see conftest.py) by matching the unique prefix
+            # "print contents array = ". This allows sample print statements to change without breaking playback.
+            # The instructions field is preserved as-is in recordings. If you modify the instructions,
+            # you must re-record the tests.
+            "input": f"print contents array = {self.print_calls}",
+        }
+
+    def _assert_validation_result(self, test_report: dict) -> None:
+        """Assert validation result and print reason."""
+        if not test_report["correct"]:
+            # Write print statements to log file in temp folder for debugging
+            import tempfile
+            from datetime import datetime
+
+            timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
+            log_file = os.path.join(tempfile.gettempdir(), f"sample_validation_error_{timestamp}.log")
+            with open(log_file, "w") as f:
+                f.write(f"Sample: {self.sample_path}\n")
+                f.write(f"Validation Error: {test_report['reason']}\n\n")
+                f.write("Print Statements:\n")
+                f.write("=" * 80 + "\n")
+                for i, print_call in enumerate(self.print_calls, 1):
+                    f.write(f"{i}. {print_call}\n")
+            print(f"\nValidation failed! Print statements logged to: {log_file}")
+        assert test_report["correct"], f"Error is identified: {test_report['reason']}"
+        print(f"Reason: {test_report['reason']}")
 
     def _validate_output(self):
-        """
-        Validates:
-        * Sample output contains the marker '==> Result: <content>'
-        * If the sample includes a comment '# Print result (should contain "<keyword>")',
-          the result must include that keyword (case-insensitive)
-        * If no keyword is specified, the result must be at least 200 characters long
-        """
-        # Find content after ==> Result: marker in print_calls array
-        result_line = None
-        for call in self.print_calls:
-            if call.startswith("==> Result:"):
-                result_line = call
-                break
-
-        if not result_line:
-            assert False, "Expected to find '==> Result:' in print calls."
-
-        # Extract content after ==> Result:
-        arrow_match = re.search(r"==> Result:(.*)", result_line, re.IGNORECASE | re.DOTALL)
-        if not arrow_match:
-            assert False, f"Expected to find '==> Result:' in line: {result_line}"
-
-        content_after_arrow = arrow_match.group(1).strip()
-
-        # Read the sample file to check for expected output comment
-        with open(self.sample_path) as f:
-            sample_code = f.read()
-
-        # Verify pattern: # Print result (should contain '...') if exist
-        match = re.search(r"# Print result \(should contain ['\"](.+?)['\"]\)", sample_code)
-        if match:
-            # Decode Unicode escape sequences like \u00b0F to actual characters
-            expected_contain = match.group(1).encode().decode("unicode_escape")
-            assert (
-                expected_contain.lower() in content_after_arrow.lower()
-            ), f"Expected to find '{expected_contain}' after '==> Result:', but got: {content_after_arrow}"
-        else:
-            result_len = len(content_after_arrow)
-            assert result_len > 200, f"Expected 200 characters after '==> Result:', but got {result_len} characters"
+        """Validate sample output using synchronous OpenAI client."""
+        credential = self.test_instance.get_credential(AIProjectClient, is_async=False)
+        with (
+            AIProjectClient(
+                endpoint=os.environ["AZURE_AI_PROJECT_ENDPOINT"], credential=cast(TokenCredential, credential)
+            ) as project_client,
+            project_client.get_openai_client() as openai_client,
+        ):
+            response = openai_client.responses.create(**self._get_validation_request_params())
+            test_report = json.loads(response.output_text)
+            self._assert_validation_result(test_report)
+
+    async def _validate_output_async(self):
+        """Validate sample output using asynchronous OpenAI client."""
+        credential = self.test_instance.get_credential(AIProjectClient, is_async=True)
+        async with (
+            AsyncAIProjectClient(
+                endpoint=os.environ["AZURE_AI_PROJECT_ENDPOINT"], credential=cast(AsyncTokenCredential, credential)
+            ) as project_client,
+        ):
+            async with project_client.get_openai_client() as openai_client:
+                response = await openai_client.responses.create(**self._get_validation_request_params())
+                test_report = json.loads(response.output_text)
+                self._assert_validation_result(test_report)
 
 
 class SamplePathPasser:
@@ -156,367 +225,119 @@ def _wrapper_sync(test_class, sample_path, **kwargs):
             return _wrapper_sync
 
 
-def _get_tools_sample_paths():
-    # Get the path to the samples folder
-    current_dir = os.path.dirname(os.path.abspath(__file__))
-    samples_folder_path = os.path.normpath(os.path.join(current_dir, os.pardir, os.pardir))
-    tools_folder = os.path.join(samples_folder_path, "samples", "agents", "tools")
-
-    tools_samples_to_skip = [
-        "sample_agent_bing_custom_search.py",
-        "sample_agent_bing_grounding.py",
-        "sample_agent_browser_automation.py",
-        "sample_agent_fabric.py",
-        "sample_agent_mcp_with_project_connection.py",
-        "sample_agent_memory_search.py",
-        "sample_agent_openapi_with_project_connection.py",
-        "sample_agent_to_agent.py",
-    ]
-    samples = []
+@overload
+def _get_sample_paths(sub_folder: str, *, samples_to_test: list[str]) -> list:
+    """Get sample paths for testing (whitelist mode).
 
-    for filename in sorted(os.listdir(tools_folder)):
-        # Only include .py files, exclude __pycache__ and utility files
-        if "sample_" in filename and "_async" not in filename and filename not in tools_samples_to_skip:
-            sample_path = os.path.join(tools_folder, filename)
-            samples.append(pytest.param(sample_path, id=filename.replace(".py", "")))
+    Args:
+        sub_folder: Relative path to the samples subfolder (e.g., "agents/tools")
+        samples_to_test: Whitelist of sample filenames to include
 
-    return samples
+    Returns:
+        List of pytest.param objects with sample paths and test IDs
+    """
+    ...
 
 
-def _get_tools_sample_paths_async():
+@overload
+def _get_sample_paths(sub_folder: str, *, samples_to_skip: list[str], is_async: bool) -> list:
+    """Get sample paths for testing (blacklist mode).
+
+    Args:
+        sub_folder: Relative path to the samples subfolder (e.g., "agents/tools")
+        samples_to_skip: Blacklist of sample filenames to exclude (auto-discovers all samples)
+        is_async: Whether to filter for async samples (_async.py suffix)
+
+    Returns:
+        List of pytest.param objects with sample paths and test IDs
+    """
+    ...
+
+
+def _get_sample_paths(
+    sub_folder: str,
+    *,
+    samples_to_skip: Union[list[str], None] = None,
+    is_async: Union[bool, None] = None,
+    samples_to_test: Union[list[str], None] = None,
+) -> list:
     # Get the path to the samples folder
     current_dir = os.path.dirname(os.path.abspath(__file__))
     samples_folder_path = os.path.normpath(os.path.join(current_dir, os.pardir, os.pardir))
-    tools_folder = os.path.join(samples_folder_path, "samples", "agents", "tools")
+    target_folder = os.path.join(samples_folder_path, "samples", *sub_folder.split("/"))
 
-    # Skip async samples that are not yet ready for testing
-    tools_samples_to_skip = [
-        "sample_agent_mcp_with_project_connection_async.py",
-        "sample_agent_memory_search_async.py",
-    ]
-    samples = []
+    if not os.path.exists(target_folder):
+        raise ValueError(f"Target folder does not exist: {target_folder}")
 
-    for filename in sorted(os.listdir(tools_folder)):
-        # Only include async .py files, exclude __pycache__ and utility files
-        if "sample_" in filename and "_async" in filename and filename not in tools_samples_to_skip:
-            sample_path = os.path.join(tools_folder, filename)
-            samples.append(pytest.param(sample_path, id=filename.replace(".py", "")))
+    # Discover all sample files in the folder
+    all_files = [f for f in os.listdir(target_folder) if f.startswith("sample_") and f.endswith(".py")]
 
-    return samples
+    # Filter by async suffix only when using samples_to_skip
+    if samples_to_skip is not None and is_async is not None:
+        if is_async:
+            all_files = [f for f in all_files if f.endswith("_async.py")]
+        else:
+            all_files = [f for f in all_files if not f.endswith("_async.py")]
 
+    # Apply whitelist or blacklist
+    if samples_to_test is not None:
+        files_to_test = [f for f in all_files if f in samples_to_test]
+    else:  # samples_to_skip is not None
+        assert samples_to_skip is not None
+        files_to_test = [f for f in all_files if f not in samples_to_skip]
 
-class TestSamples(AzureRecordedTestCase):
-    _samples_folder_path: str
-    _results: dict[str, tuple[bool, str]]
+    # Create pytest.param objects
+    samples = []
+    for filename in sorted(files_to_test):
+        sample_path = os.path.join(target_folder, filename)
+        test_id = filename.replace(".py", "")
+        samples.append(pytest.param(sample_path, id=test_id))
 
-    """
-    Test class for running all samples in the `/sdk/ai/azure-ai-projects/samples` folder.
-
-    To run this test:
-    * 'cd' to the folder '/sdk/ai/azure-ai-projects' in your azure-sdk-for-python repo.
-    * set AZURE_AI_PROJECT_ENDPOINT=<your-project-endpoint> - Define your Microsoft Foundry project endpoint used by the test.
-    * set AZURE_AI_PROJECTS_CONSOLE_LOGGING=false - to make sure logging is not enabled in the test, to reduce console spew.
-    * Uncomment the two lines that start with "@pytest.mark.skip" below.
-    * Run:  pytest tests\\samples\\test_samples.py::TestSamples
-    * Load the resulting report in Excel: tests\\samples\\samples_report.csv
-    """
+    return samples
 
-    @classmethod
-    def setup_class(cls):
-        current_path = os.path.abspath(__file__)
-        cls._samples_folder_path = os.path.join(current_path, os.pardir, os.pardir, os.pardir)
-        cls._results: dict[str, tuple[bool, str]] = {}
-
-    @classmethod
-    def teardown_class(cls):
-        """
-        Class-level teardown method that generates a report file named "samples_report.csv" after all tests have run.
-
-        The report contains one line per sample run, with three columns:
-            1. PASS or FAIL indicating the sample result.
-            2. The name of the sample.
-            3. The exception string summary if the sample failed, otherwise empty.
-
-        The report is written to the same directory as this test file.
-        """
-        report_path = os.path.join(os.path.dirname(os.path.abspath(__file__)), "samples_report.csv")
-        with open(report_path, mode="w", newline="") as file:
-            writer = csv.writer(file, quotechar='"', quoting=csv.QUOTE_ALL)  # Ensures proper quoting
-            for test_name, (passed, exception_string) in cls._results.items():
-                exception_message = f'"{exception_string.splitlines()[0]}"' if exception_string else ""
-                writer.writerow([f"{'PASS' if passed else 'FAIL'}", test_name, exception_message])
-
-    @classmethod
-    def _set_env_vars(cls, sample_name: str, **kwargs):
-        """
-        Sets environment variables for a given sample run and prints them.
-
-        Args:
-            sample_name (str): The name of the sample being executed.
-            **kwargs: Arbitrary keyword arguments representing environment variable names and their values.
-        """
-
-        print(f"\nRunning {sample_name} with environment variables: ", end="")
-        for key, value in kwargs.items():
-            if value:
-                env_key = key.upper()
-                os.environ[env_key] = value
-                print(f"{env_key}={value} ", end="")
-        print("\n")
-
-    @classmethod
-    def _run_sample(cls, sample_name: str) -> None:
-        """
-        Executes a synchronous sample file and records the result.
-
-        Args:
-            sample_name (str): The name of the sample file to execute.
-
-        Raises:
-            Exception: Re-raises any exception encountered during execution of the sample file.
-
-        Side Effects:
-            Updates the class-level _results dictionary with the execution status and error message (if any)
-            for the given sample.
-            Prints an error message to stdout if execution fails.
-        """
-
-        sample_path = os.path.normpath(os.path.join(TestSamples._samples_folder_path, sample_name))
-        with open(sample_path) as f:
-            code = f.read()
-            try:
-                exec(code)
-            except HttpResponseError as exc:
-                exception_message = f"{exc.status_code}, {exc.reason}, {str(exc)}"
-                TestSamples._results[sample_name] = (False, exception_message)
-                print(f"=================> Error running sample {sample_path}: {exception_message}")
-                raise Exception from exc
-            except Exception as exc:
-                TestSamples._results[sample_name] = (False, str(exc))
-                print(f"=================> Error running sample {sample_path}: {exc}")
-                raise Exception from exc
-            TestSamples._results[sample_name] = (True, "")
-
-    @classmethod
-    async def _run_sample_async(cls, sample_name: str) -> None:
-        """
-        Asynchronously runs a sample Python script specified by its file name.
-
-        This method dynamically imports the sample module from the given file path,
-        executes its `main()` coroutine, and records the result. If an exception occurs
-        during execution, the error is logged and re-raised.
-
-        Args:
-            sample_name (str): The name of the sample Python file to run (relative to the samples folder).
-
-        Raises:
-            ImportError: If the sample module cannot be loaded.
-            Exception: If an error occurs during the execution of the sample's `main()` coroutine.
-
-        Side Effects:
-            Updates the `_results` dictionary with the execution status and error message (if any).
-            Prints error messages to the console if execution fails.
-        """
-
-        sample_path = os.path.normpath(os.path.join(TestSamples._samples_folder_path, sample_name))
-        # Dynamically import the module from the given path
-        module_name = os.path.splitext(os.path.basename(sample_path))[0]
-        spec = importlib.util.spec_from_file_location(module_name, sample_path)
-        if spec is None or spec.loader is None:
-            raise ImportError(f"Could not load module {module_name} from {sample_path}")
-        module = importlib.util.module_from_spec(spec)
-        spec.loader.exec_module(module)
-        # Await the main() coroutine defined in the sample
-        try:
-            await module.main()
-        except HttpResponseError as exc:
-            exception_message = f"{exc.status_code}, {exc.reason}, {str(exc)}"
-            TestSamples._results[sample_name] = (False, exception_message)
-            print(f"=================> Error running sample {sample_path}: {exception_message}")
-            raise Exception from exc
-        except Exception as exc:
-            TestSamples._results[sample_name] = (False, str(exc))
-            print(f"=================> Error running sample {sample_path}: {exc}")
-            raise Exception from exc
-        TestSamples._results[sample_name] = (True, "")
 
-    @pytest.mark.parametrize(
-        "sample_name, model_deployment_name, connection_name, data_folder",
-        [
-            ("samples\\agents\\sample_agents.py", "gpt-4o", "", ""),
-            ("samples\\connections\\sample_connections.py", "", "connection1", ""),
-            ("samples\\deployments\\sample_deployments.py", "DeepSeek-V3", "", ""),
-            ("samples\\datasets\\sample_datasets.py", "", "balapvbyostoragecanary", "samples\\datasets\\data_folder"),
-            (
-                "samples\\datasets\\sample_datasets_download.py",
-                "",
-                "balapvbyostoragecanary",
-                "samples\\datasets\\data_folder",
-            ),
-            ("samples\\indexes\\sample_indexes.py", "", "", ""),
-            (
-                "samples\\inference\\azure-ai-inference\\sample_chat_completions_with_azure_ai_inference_client.py",
-                "Phi-4",
-                "",
-                "",
-            ),
-            (
-                "samples\\inference\\azure-ai-inference\\sample_chat_completions_with_azure_ai_inference_client_and_azure_monitor_tracing.py",
-                "Phi-4",
-                "",
-                "",
-            ),
-            (
-                "samples\\inference\\azure-ai-inference\\sample_chat_completions_with_azure_ai_inference_client_and_console_tracing.py",
-                "Phi-4",
-                "",
-                "",
-            ),
-            (
-                "samples\\inference\\azure-openai\\sample_chat_completions_with_azure_openai_client.py",
-                "gpt-4o",
-                "connection1",
-                "",
-            ),
-            (
-                "samples\\inference\\azure-openai\\sample_responses_with_azure_openai_client.py",
-                "gpt-4o",
-                "connection1",
-                "",
-            ),
-            (
-                "samples\\inference\\azure-openai\\sample_chat_completions_with_azure_openai_client_and_azure_monitor_tracing.py",
-                "gpt-4o",
-                "",
-                "",
-            ),
-            (
-                "samples\\inference\\azure-openai\\sample_chat_completions_with_azure_openai_client_and_console_tracing.py",
-                "gpt-4o",
-                "",
-                "",
-            ),
-            (
-                "samples\\inference\\azure-ai-inference\\sample_image_embeddings_with_azure_ai_inference_client.py",
-                "Cohere-embed-v3-english",
-                "",
-                "samples\\inference\\azure-ai-inference",
-            ),
-            (
-                "samples\\inference\\azure-ai-inference\\sample_text_embeddings_with_azure_ai_inference_client.py",
-                "text-embedding-3-large",
-                "",
-                "",
-            ),
-            ("samples\\telemetry\\sample_telemetry.py", "", "", ""),
-        ],
-    )
-    @pytest.mark.skip(reason="This test should only run manually on your local machine, with live service calls.")
-    def test_samples(
-        self, sample_name: str, model_deployment_name: str, connection_name: str, data_folder: str
-    ) -> None:
-        """
-        Run all the synchronous sample code in the samples folder. If a sample throws an exception, which for example
-        happens when the service responds with an error, the test will fail.
-
-        Before running this test, you need to define the following environment variables:
-        1) AZURE_AI_PROJECT_ENDPOINT - The Azure AI Project endpoint, as found in the overview page of your
-           Microsoft Foundry project.
-        """
-
-        self._set_env_vars(
-            sample_name,
-            **{
-                "model_deployment_name": model_deployment_name,
-                "connection_name": connection_name,
-                "data_folder": data_folder,
-            },
-        )
-        TestSamples._run_sample(sample_name)
+class TestSamples(AzureRecordedTestCase):
 
+    @servicePreparer()
     @pytest.mark.parametrize(
-        "sample_name, model_deployment_name, connection_name, data_folder",
-        [
-            ("samples\\agents\\sample_agents_async.py", "gpt-4o", "", ""),
-            ("samples\\connections\\sample_connections_async.py", "", "connection1", ""),
-            (
-                "samples\\datasets\\sample_datasets_async.py",
-                "",
-                "balapvbyostoragecanary",
-                "samples\\datasets\\data_folder",
-            ),
-            ("samples\\deployments\\sample_deployments_async.py", "DeepSeek-V3", "", ""),
-            ("samples\\indexes\\sample_indexes_async.py", "", "", ""),
-            (
-                "samples\\inference\\azure-ai-inference\\sample_chat_completions_with_azure_ai_inference_client_async.py",
-                "Phi-4",
-                "",
-                "",
-            ),
-            (
-                "samples\\inference\\azure-openai\\sample_chat_completions_with_azure_openai_client_async.py",
-                "gpt-4o",
-                "connection1",
-                "",
-            ),
-            (
-                "samples\\inference\\azure-openai\\sample_responses_with_azure_openai_client_async.py",
-                "gpt-4o",
-                "connection1",
-                "",
-            ),
-            (
-                "samples\\inference\\azure-ai-inference\\sample_image_embeddings_with_azure_ai_inference_client_async.py",
-                "Cohere-embed-v3-english",
-                "",
-                "samples\\inference\\azure-ai-inference",
-            ),
-            (
-                "samples\\inference\\azure-ai-inference\\sample_text_embeddings_with_azure_ai_inference_client_async.py",
-                "text-embedding-3-large",
-                "",
-                "",
-            ),
-            ("samples\\telemetry\\sample_telemetry_async.py", "", "", ""),
-        ],
+        "sample_path",
+        _get_sample_paths(
+            "agents/tools",
+            samples_to_skip=[
+                "sample_agent_bing_custom_search.py",
+                "sample_agent_bing_grounding.py",
+                "sample_agent_browser_automation.py",
+                "sample_agent_fabric.py",
+                "sample_agent_mcp_with_project_connection.py",
+                "sample_agent_memory_search.py",
+                "sample_agent_openapi_with_project_connection.py",
+                "sample_agent_to_agent.py",
+            ],
+            is_async=False,
+        ),
     )
-    @pytest.mark.skip(reason="This test should only run manually on your local machine, with live service calls.")
-    async def test_samples_async(
-        self, sample_name: str, model_deployment_name: str, connection_name: str, data_folder: str
-    ) -> None:
-        """
-        Run all the asynchronous sample code in the samples folder. If a sample throws an exception, which for example
-        happens when the service responds with an error, the test will fail.
-
-        Before running this test, you need to define the following environment variables:
-        1) AZURE_AI_PROJECT_ENDPOINT - The Azure AI Project endpoint, as found in the overview page of your
-           Microsoft Foundry project.
-        """
-
-        self._set_env_vars(
-            sample_name,
-            **{
-                "model_deployment_name": model_deployment_name,
-                "connection_name": connection_name,
-                "data_folder": data_folder,
-            },
-        )
-        await TestSamples._run_sample_async(sample_name)
-
-    @servicePreparer()
-    @pytest.mark.parametrize("sample_path", _get_tools_sample_paths())
     @SamplePathPasser()
     @recorded_by_proxy(RecordedTransport.AZURE_CORE, RecordedTransport.HTTPX)
-    def test_tools_samples(self, sample_path: str, **kwargs) -> None:
+    def test_agent_tools_samples(self, sample_path: str, **kwargs) -> None:
         env_var_mapping = self._get_sample_environment_variables_map()
         executor = SampleExecutor(self, sample_path, env_var_mapping, **kwargs)
         executor.execute()
 
     @servicePreparer()
-    @pytest.mark.parametrize("sample_path", _get_tools_sample_paths_async())
+    @pytest.mark.parametrize(
+        "sample_path",
+        _get_sample_paths(
+            "agents/tools",
+            samples_to_skip=[
+                "sample_agent_mcp_with_project_connection_async.py",
+                "sample_agent_memory_search_async.py",
+            ],
+            is_async=True,
+        ),
+    )
     @SamplePathPasser()
     @recorded_by_proxy_async(RecordedTransport.AZURE_CORE, RecordedTransport.HTTPX)
-    async def test_tools_samples_async(self, sample_path: str, **kwargs) -> None:
+    async def test_agent_tools_samples_async(self, sample_path: str, **kwargs) -> None:
         env_var_mapping = self._get_sample_environment_variables_map()
         executor = SampleExecutor(self, sample_path, env_var_mapping, **kwargs)
         await executor.execute_async()