Resolved merge conflicts

tests/integration/agents/test_agents.py

@@ -10,8 +10,7 @@ from uuid import uuid4
 import pytest
 from llama_stack_client.lib.agents.agent import Agent
 from llama_stack_client.lib.agents.event_logger import EventLogger
-from llama_stack_client.types.agents.turn_create_params import Document as AgentDocument
-from llama_stack_client.types.memory_insert_params import Document
+from llama_stack_client.types.agents.turn_create_params import Document
 from llama_stack_client.types.shared_params.agent_config import AgentConfig, ToolConfig
 
 from llama_stack.apis.agents.agents import (
@@ -242,7 +241,7 @@ def test_code_interpreter_for_attachments(llama_stack_client_with_mocked_inferen
 
     codex_agent = Agent(llama_stack_client_with_mocked_inference, **agent_config)
     session_id = codex_agent.create_session(f"test-session-{uuid4()}")
-    inflation_doc = AgentDocument(
+    inflation_doc = Document(
         content="https://raw.githubusercontent.com/meta-llama/llama-stack-apps/main/examples/resources/inflation.csv",
         mime_type="text/csv",
     )

tests/integration/datasetio/test_datasetio.py

@@ -9,11 +9,31 @@ import mimetypes
 import os
 from pathlib import Path
 
+<<<<<<< HEAD
+=======
+import pytest
+
+>>>>>>> upstream/main
 # How to run this test:
 #
 # LLAMA_STACK_CONFIG="template-name" pytest -v tests/integration/datasetio
 
 
+<<<<<<< HEAD
+=======
+@pytest.fixture
+def dataset_for_test(llama_stack_client):
+    dataset_id = "test_dataset"
+    register_dataset(llama_stack_client, dataset_id=dataset_id)
+    yield
+    # Teardown - this always runs, even if the test fails
+    try:
+        llama_stack_client.datasets.unregister(dataset_id)
+    except Exception as e:
+        print(f"Warning: Failed to unregister test_dataset: {e}")
+
+
+>>>>>>> upstream/main
 def data_url_from_file(file_path: str) -> str:
     if not os.path.exists(file_path):
         raise FileNotFoundError(f"File not found: {file_path}")
@@ -80,8 +100,12 @@ def test_register_unregister_dataset(llama_stack_client):
     assert len(response) == 0
 
 
+<<<<<<< HEAD
 def test_get_rows_paginated(llama_stack_client):
     register_dataset(llama_stack_client)
+=======
+def test_get_rows_paginated(llama_stack_client, dataset_for_test):
+>>>>>>> upstream/main
     response = llama_stack_client.datasetio.get_rows_paginated(
         dataset_id="test_dataset",
         rows_in_page=3,

tests/integration/fixtures/common.py

@@ -52,6 +52,8 @@ def llama_stack_client_with_mocked_inference(llama_stack_client, request):
 
     If --record-responses is passed, it will call the real APIs and record the responses.
     """
+    # TODO: will rework this to be more stable
+    return llama_stack_client
     if not isinstance(llama_stack_client, LlamaStackAsLibraryClient):
         logging.warning(
             "llama_stack_client_with_mocked_inference is not supported for this client, returning original client without mocking"

tests/integration/inference/test_vision_inference.py

@@ -36,7 +36,7 @@ def test_image_chat_completion_non_streaming(client_with_models, vision_model_id
                 "type": "image",
                 "image": {
                     "url": {
-                        "uri": "https://raw.githubusercontent.com/meta-llama/llama-stack/main/tests/api/inference/dog.png"
+                        "uri": "https://raw.githubusercontent.com/meta-llama/llama-stack/main/tests/integration/inference/dog.png"
                     },
                 },
             },
@@ -65,7 +65,7 @@ def test_image_chat_completion_streaming(client_with_models, vision_model_id):
                 "type": "image",
                 "image": {
                     "url": {
-                        "uri": "https://raw.githubusercontent.com/meta-llama/llama-stack/main/tests/api/inference/dog.png"
+                        "uri": "https://raw.githubusercontent.com/meta-llama/llama-stack/main/tests/integration/inference/dog.png"
                     },
                 },
             },

tests/integration/providers/__init__.py

@@ -0,0 +1,5 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the terms described in the LICENSE file in
+# the root directory of this source tree.

tests/integration/providers/test_providers.py

@@ -0,0 +1,17 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the terms described in the LICENSE file in
+# the root directory of this source tree.
+
+import pytest
+from llama_stack_client import LlamaStackClient
+
+from llama_stack import LlamaStackAsLibraryClient
+
+
+class TestProviders:
+    @pytest.mark.asyncio
+    def test_list(self, llama_stack_client: LlamaStackAsLibraryClient | LlamaStackClient):
+        provider_list = llama_stack_client.providers.list()
+        assert provider_list is not None

tests/integration/scoring/test_scoring.py

@@ -10,6 +10,19 @@ import pytest
 from ..datasetio.test_datasetio import register_dataset
 
 
+@pytest.fixture
+def rag_dataset_for_test(llama_stack_client):
+    dataset_id = "test_dataset"
+    register_dataset(llama_stack_client, for_rag=True, dataset_id=dataset_id)
+    yield  # This is where the test function will run
+
+    # Teardown - this always runs, even if the test fails
+    try:
+        llama_stack_client.datasets.unregister(dataset_id)
+    except Exception as e:
+        print(f"Warning: Failed to unregister test_dataset: {e}")
+
+
 @pytest.fixture
 def sample_judge_prompt_template():
     return "Output a number response in the following format: Score: <number>, where <number> is the number between 0 and 9."
@@ -79,9 +92,7 @@ def test_scoring_functions_register(
     # TODO: add unregister api for scoring functions
 
 
-def test_scoring_score(llama_stack_client):
-    register_dataset(llama_stack_client, for_rag=True)
-
+def test_scoring_score(llama_stack_client, rag_dataset_for_test):
     # scoring individual rows
     rows = llama_stack_client.datasetio.get_rows_paginated(
         dataset_id="test_dataset",
@@ -115,9 +126,9 @@ def test_scoring_score(llama_stack_client):
         assert len(response.results[x].score_rows) == 5
 
 
-def test_scoring_score_with_params_llm_as_judge(llama_stack_client, sample_judge_prompt_template, judge_model_id):
-    register_dataset(llama_stack_client, for_rag=True)
-
+def test_scoring_score_with_params_llm_as_judge(
+    llama_stack_client, sample_judge_prompt_template, judge_model_id, rag_dataset_for_test
+):
     # scoring individual rows
     rows = llama_stack_client.datasetio.get_rows_paginated(
         dataset_id="test_dataset",
@@ -167,9 +178,8 @@ def test_scoring_score_with_params_llm_as_judge(llama_stack_client, sample_judge
     ],
 )
 def test_scoring_score_with_aggregation_functions(
-    llama_stack_client, sample_judge_prompt_template, judge_model_id, provider_id
+    llama_stack_client, sample_judge_prompt_template, judge_model_id, provider_id, rag_dataset_for_test
 ):
-    register_dataset(llama_stack_client, for_rag=True)
     rows = llama_stack_client.datasetio.get_rows_paginated(
         dataset_id="test_dataset",
         rows_in_page=3,