diff --git a/sdk/ai/azure-ai-projects/assets.json b/sdk/ai/azure-ai-projects/assets.json
index d620327691f2..219af81a83e0 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_4160165ae8"
+  "Tag": "python/ai/azure-ai-projects_0c9b00044b"
 }
diff --git a/sdk/ai/azure-ai-projects/samples/agents/assets/.gitattributes b/sdk/ai/azure-ai-projects/samples/agents/assets/.gitattributes
new file mode 100644
index 000000000000..878545b8b1cb
--- /dev/null
+++ b/sdk/ai/azure-ai-projects/samples/agents/assets/.gitattributes
@@ -0,0 +1,13 @@
+# Force LF line endings for test files to ensure consistent binary representation
+# across Windows and Linux platforms.
+#
+# These files are read and sent as binary REST API request payloads in tests.
+# Without consistent line endings, the binary content differs between platforms,
+# causing recorded test traffic to be inconsistent and test playback to fail.
+#
+# By enforcing LF endings, the binary representation remains identical regardless
+# of the platform where tests are executed.
+
+*.md text eol=lf
+*.csv text eol=lf
+*.jsonl text eol=lf
diff --git a/sdk/ai/azure-ai-projects/tests/conftest.py b/sdk/ai/azure-ai-projects/tests/conftest.py
index 5a009ec7c80d..ad54a729211e 100644
--- a/sdk/ai/azure-ai-projects/tests/conftest.py
+++ b/sdk/ai/azure-ai-projects/tests/conftest.py
@@ -3,6 +3,14 @@
 # Copyright (c) Microsoft Corporation.
 # Licensed under the MIT License.
 # ------------------------------------
+
+# Register MIME types before any other imports to ensure consistent Content-Type detection
+# across Windows, macOS, and Linux when uploading files in tests
+import mimetypes
+
+mimetypes.add_type("text/csv", ".csv")
+mimetypes.add_type("text/markdown", ".md")
+
 import os
 import pytest
 from dotenv import load_dotenv, find_dotenv
@@ -100,9 +108,6 @@ def sanitize_url_paths():
 
     sanitize_url_paths()
 
-    # Normalize Content-Type for CSV files in multipart form-data (varies by OS/Python version)
-    add_general_string_sanitizer(target="Content-Type: text/csv", value="Content-Type: application/vnd.ms-excel")
-
     # Sanitize API key from service response (this includes Application Insights connection string)
     add_body_key_sanitizer(json_path="credentials.key", value="sanitized-api-key")
 
diff --git a/sdk/ai/azure-ai-projects/tests/samples/samples_report.csv b/sdk/ai/azure-ai-projects/tests/samples/samples_report.csv
new file mode 100644
index 000000000000..e69de29bb2d1
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 d0c4db50142c..9de90603c0b6 100644
--- a/sdk/ai/azure-ai-projects/tests/samples/test_samples.py
+++ b/sdk/ai/azure-ai-projects/tests/samples/test_samples.py
@@ -165,8 +165,6 @@ def _get_tools_sample_paths():
         "sample_agent_browser_automation.py",
         "sample_agent_code_interpreter.py",
         "sample_agent_fabric.py",
-        "sample_agent_file_search.py",
-        "sample_agent_file_search_in_stream.py",
         "sample_agent_mcp_with_project_connection.py",
         "sample_agent_memory_search.py",
         "sample_agent_openapi_with_project_connection.py",
@@ -192,8 +190,6 @@ def _get_tools_sample_paths_async():
     # Skip async samples that are not yet ready for testing
     tools_samples_to_skip = [
         "sample_agent_code_interpreter_async.py",
-        "sample_agent_file_search_async.py",
-        "sample_agent_file_search_in_stream_async.py",
         "sample_agent_mcp_with_project_connection_async.py",
         "sample_agent_memory_search_async.py",
     ]