phoenix-oss/llama-stack-mirror
0
0
Fork 1
mirror of https://github.com/meta-llama/llama-stack.git synced 2025-10-11 21:48:36 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

feat: Add support for Conversations in Responses API (#3743)

Browse source 
# What does this PR do?
This PR adds support for Conversations in Responses.

<!-- If resolving an issue, uncomment and update the line below -->
<!-- Closes #[issue-number] -->

## Test Plan
Unit tests
Integration tests

<Details>
<Summary>Manual testing with this script: (click to expand)</Summary>

```python
from openai import OpenAI

client = OpenAI()
client = OpenAI(base_url="http://localhost:8321/v1/", api_key="none")

def test_conversation_create():
    print("Testing conversation create...")
    conversation = client.conversations.create(
        metadata={"topic": "demo"},
        items=[
            {"type": "message", "role": "user", "content": "Hello!"}
        ]
    )
    print(f"Created: {conversation}")
    return conversation

def test_conversation_retrieve(conv_id):
    print(f"Testing conversation retrieve for {conv_id}...")
    retrieved = client.conversations.retrieve(conv_id)
    print(f"Retrieved: {retrieved}")
    return retrieved

def test_conversation_update(conv_id):
    print(f"Testing conversation update for {conv_id}...")
    updated = client.conversations.update(
        conv_id,
        metadata={"topic": "project-x"}
    )
    print(f"Updated: {updated}")
    return updated

def test_conversation_delete(conv_id):
    print(f"Testing conversation delete for {conv_id}...")
    deleted = client.conversations.delete(conv_id)
    print(f"Deleted: {deleted}")
    return deleted

def test_conversation_items_create(conv_id):
    print(f"Testing conversation items create for {conv_id}...")
    items = client.conversations.items.create(
        conv_id,
        items=[
            {
                "type": "message",
                "role": "user",
                "content": [{"type": "input_text", "text": "Hello!"}]
            },
            {
                "type": "message",
                "role": "user",
                "content": [{"type": "input_text", "text": "How are you?"}]
            }
        ]
    )
    print(f"Items created: {items}")
    return items

def test_conversation_items_list(conv_id):
    print(f"Testing conversation items list for {conv_id}...")
    items = client.conversations.items.list(conv_id, limit=10)
    print(f"Items list: {items}")
    return items

def test_conversation_item_retrieve(conv_id, item_id):
    print(f"Testing conversation item retrieve for {conv_id}/{item_id}...")
    item = client.conversations.items.retrieve(conversation_id=conv_id, item_id=item_id)
    print(f"Item retrieved: {item}")
    return item

def test_conversation_item_delete(conv_id, item_id):
    print(f"Testing conversation item delete for {conv_id}/{item_id}...")
    deleted = client.conversations.items.delete(conversation_id=conv_id, item_id=item_id)
    print(f"Item deleted: {deleted}")
    return deleted

def test_conversation_responses_create():
    print("\nTesting conversation create for a responses example...")
    conversation = client.conversations.create()
    print(f"Created: {conversation}")

    response = client.responses.create(
      model="gpt-4.1",
      input=[{"role": "user", "content": "What are the 5 Ds of dodgeball?"}],
      conversation=conversation.id,
    )
    print(f"Created response: {response} for conversation {conversation.id}")

    return response, conversation

def test_conversations_responses_create_followup(
        conversation,
        content="Repeat what you just said but add 'this is my second time saying this'",
    ):
    print(f"Using: {conversation.id}")

    response = client.responses.create(
      model="gpt-4.1",
      input=[{"role": "user", "content": content}],
      conversation=conversation.id,
    )
    print(f"Created response: {response} for conversation {conversation.id}")

    conv_items = client.conversations.items.list(conversation.id)
    print(f"\nRetrieving list of items for conversation {conversation.id}:")
    print(conv_items.model_dump_json(indent=2))

def test_response_with_fake_conv_id():
    fake_conv_id = "conv_zzzzzzzzz5dc81908289d62779d2ac510a2b0b602ef00a44"
    print(f"Using {fake_conv_id}")
    try:
        response = client.responses.create(
          model="gpt-4.1",
          input=[{"role": "user", "content": "say hello"}],
          conversation=fake_conv_id,
        )
        print(f"Created response: {response} for conversation {fake_conv_id}")
    except Exception as e:
        print(f"failed to create response for conversation {fake_conv_id} with error {e}")


def main():
    print("Testing OpenAI Conversations API...")

    # Create conversation
    conversation = test_conversation_create()
    conv_id = conversation.id

    # Retrieve conversation
    test_conversation_retrieve(conv_id)

    # Update conversation
    test_conversation_update(conv_id)

    # Create items
    items = test_conversation_items_create(conv_id)

    # List items
    items_list = test_conversation_items_list(conv_id)

    # Retrieve specific item
    if items_list.data:
        item_id = items_list.data[0].id
        test_conversation_item_retrieve(conv_id, item_id)

        # Delete item
        test_conversation_item_delete(conv_id, item_id)

    # Delete conversation
    test_conversation_delete(conv_id)

    response, conversation2 = test_conversation_responses_create()
    print('\ntesting reseponse retrieval')
    test_conversation_retrieve(conversation2.id)

    print('\ntesting responses follow up')
    test_conversations_responses_create_followup(conversation2)

    print('\ntesting responses follow up x2!')

    test_conversations_responses_create_followup(
        conversation2,
        content="Repeat what you just said but add 'this is my third time saying this'",
    )

    test_response_with_fake_conv_id()

    print("All tests completed!")


if __name__ == "__main__":
    main()
```
</Details>

---------

Signed-off-by: Francisco Javier Arceo <farceo@redhat.com>
Co-authored-by: Ashwin Bharambe <ashwin.bharambe@gmail.com>
This commit is contained in:
Francisco Arceo 2025-10-10 14:57:40 -04:00 • committed by GitHub
parent 932fea813a
commit e7d21e1ee3
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
41 changed files with 6221 additions and 19 deletions
Download patch file Download diff file Expand all files Collapse all files

15
tests/unit/providers/agents/meta_reference/test_openai_responses.py
View file

@ -83,9 +83,21 @@ def mock_vector_io_api():
return vector_io_api
@pytest.fixture
def mock_conversations_api():
"""Mock conversations API for testing."""
mock_api = AsyncMock()
return mock_api
@pytest.fixture
def openai_responses_impl(
mock_inference_api, mock_tool_groups_api, mock_tool_runtime_api, mock_responses_store, mock_vector_io_api
mock_inference_api,
mock_tool_groups_api,
mock_tool_runtime_api,
mock_responses_store,
mock_vector_io_api,
mock_conversations_api,
):
return OpenAIResponsesImpl(
inference_api=mock_inference_api,
@ -93,6 +105,7 @@ def openai_responses_impl(
tool_runtime_api=mock_tool_runtime_api,
responses_store=mock_responses_store,
vector_io_api=mock_vector_io_api,
conversations_api=mock_conversations_api,
)

331
tests/unit/providers/agents/meta_reference/test_openai_responses_conversations.py Normal file
View file

@ -0,0 +1,331 @@
# 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.apis.agents.openai_responses import (
OpenAIResponseMessage,
OpenAIResponseObject,
OpenAIResponseObjectStreamResponseCompleted,
OpenAIResponseOutputMessageContentOutputText,
)
from llama_stack.apis.common.errors import (
ConversationNotFoundError,
InvalidConversationIdError,
)
from llama_stack.apis.conversations.conversations import (
ConversationItemList,
)
# Import existing fixtures from the main responses test file
pytest_plugins = ["tests.unit.providers.agents.meta_reference.test_openai_responses"]
from llama_stack.providers.inline.agents.meta_reference.responses.openai_responses import (
OpenAIResponsesImpl,
)
@pytest.fixture
def responses_impl_with_conversations(
mock_inference_api,
mock_tool_groups_api,
mock_tool_runtime_api,
mock_responses_store,
mock_vector_io_api,
mock_conversations_api,
):
"""Create OpenAIResponsesImpl instance with conversations API."""
return OpenAIResponsesImpl(
inference_api=mock_inference_api,
tool_groups_api=mock_tool_groups_api,
tool_runtime_api=mock_tool_runtime_api,
responses_store=mock_responses_store,
vector_io_api=mock_vector_io_api,
conversations_api=mock_conversations_api,
)
class TestConversationValidation:
"""Test conversation ID validation logic."""
async def test_nonexistent_conversation_raises_error(
self, responses_impl_with_conversations, mock_conversations_api
):
"""Test that ConversationNotFoundError is raised for non-existent conversation."""
conv_id = "conv_nonexistent"
# Mock conversation not found
mock_conversations_api.list.side_effect = ConversationNotFoundError("conv_nonexistent")
with pytest.raises(ConversationNotFoundError):
await responses_impl_with_conversations.create_openai_response(
input="Hello", model="test-model", conversation=conv_id, stream=False
)
class TestConversationContextLoading:
"""Test conversation context loading functionality."""
async def test_load_conversation_context_simple_input(
self, responses_impl_with_conversations, mock_conversations_api
):
"""Test loading conversation context with simple string input."""
conv_id = "conv_test123"
input_text = "Hello, how are you?"
# mock items in chronological order (a consequence of order="asc")
mock_conversation_items = ConversationItemList(
data=[
OpenAIResponseMessage(
id="msg_1",
content=[{"type": "input_text", "text": "Previous user message"}],
role="user",
status="completed",
type="message",
),
OpenAIResponseMessage(
id="msg_2",
content=[{"type": "output_text", "text": "Previous assistant response"}],
role="assistant",
status="completed",
type="message",
),
],
first_id="msg_1",
has_more=False,
last_id="msg_2",
object="list",
)
mock_conversations_api.list.return_value = mock_conversation_items
result = await responses_impl_with_conversations._load_conversation_context(conv_id, input_text)
# should have conversation history + new input
assert len(result) == 3
assert isinstance(result[0], OpenAIResponseMessage)
assert result[0].role == "user"
assert isinstance(result[1], OpenAIResponseMessage)
assert result[1].role == "assistant"
assert isinstance(result[2], OpenAIResponseMessage)
assert result[2].role == "user"
assert result[2].content == input_text
async def test_load_conversation_context_api_error(self, responses_impl_with_conversations, mock_conversations_api):
"""Test loading conversation context when API call fails."""
conv_id = "conv_test123"
input_text = "Hello"
mock_conversations_api.list.side_effect = Exception("API Error")
with pytest.raises(Exception, match="API Error"):
await responses_impl_with_conversations._load_conversation_context(conv_id, input_text)
async def test_load_conversation_context_with_list_input(
self, responses_impl_with_conversations, mock_conversations_api
):
"""Test loading conversation context with list input."""
conv_id = "conv_test123"
input_messages = [
OpenAIResponseMessage(role="user", content="First message"),
OpenAIResponseMessage(role="user", content="Second message"),
]
mock_conversations_api.list.return_value = ConversationItemList(
data=[], first_id=None, has_more=False, last_id=None, object="list"
)
result = await responses_impl_with_conversations._load_conversation_context(conv_id, input_messages)
assert len(result) == 2
assert result == input_messages
async def test_load_conversation_context_empty_conversation(
self, responses_impl_with_conversations, mock_conversations_api
):
"""Test loading context from empty conversation."""
conv_id = "conv_empty"
input_text = "Hello"
mock_conversations_api.list.return_value = ConversationItemList(
data=[], first_id=None, has_more=False, last_id=None, object="list"
)
result = await responses_impl_with_conversations._load_conversation_context(conv_id, input_text)
assert len(result) == 1
assert result[0].role == "user"
assert result[0].content == input_text
class TestMessageSyncing:
"""Test message syncing to conversations."""
async def test_sync_response_to_conversation_simple(
self, responses_impl_with_conversations, mock_conversations_api
):
"""Test syncing simple response to conversation."""
conv_id = "conv_test123"
input_text = "What are the 5 Ds of dodgeball?"
# mock response
mock_response = OpenAIResponseObject(
id="resp_123",
created_at=1234567890,
model="test-model",
object="response",
output=[
OpenAIResponseMessage(
id="msg_response",
content=[
OpenAIResponseOutputMessageContentOutputText(
text="The 5 Ds are: Dodge, Duck, Dip, Dive, and Dodge.", type="output_text", annotations=[]
)
],
role="assistant",
status="completed",
type="message",
)
],
status="completed",
)
await responses_impl_with_conversations._sync_response_to_conversation(conv_id, input_text, mock_response)
# should call add_items with user input and assistant response
mock_conversations_api.add_items.assert_called_once()
call_args = mock_conversations_api.add_items.call_args
assert call_args[0][0] == conv_id # conversation_id
items = call_args[0][1] # conversation_items
assert len(items) == 2
# User message
assert items[0].type == "message"
assert items[0].role == "user"
assert items[0].content[0].type == "input_text"
assert items[0].content[0].text == input_text
# Assistant message
assert items[1].type == "message"
assert items[1].role == "assistant"
async def test_sync_response_to_conversation_api_error(
self, responses_impl_with_conversations, mock_conversations_api
):
mock_conversations_api.add_items.side_effect = Exception("API Error")
mock_response = OpenAIResponseObject(
id="resp_123", created_at=1234567890, model="test-model", object="response", output=[], status="completed"
)
# matching the behavior of OpenAI here
with pytest.raises(Exception, match="API Error"):
await responses_impl_with_conversations._sync_response_to_conversation(
"conv_test123", "Hello", mock_response
)
async def test_sync_unsupported_types(self, responses_impl_with_conversations):
mock_response = OpenAIResponseObject(
id="resp_123", created_at=1234567890, model="test-model", object="response", output=[], status="completed"
)
with pytest.raises(NotImplementedError, match="Unsupported input item type"):
await responses_impl_with_conversations._sync_response_to_conversation(
"conv_123", [{"not": "message"}], mock_response
)
with pytest.raises(NotImplementedError, match="Unsupported message role: system"):
await responses_impl_with_conversations._sync_response_to_conversation(
"conv_123", [OpenAIResponseMessage(role="system", content="test")], mock_response
)
class TestIntegrationWorkflow:
"""Integration tests for the full conversation workflow."""
async def test_create_response_with_valid_conversation(
self, responses_impl_with_conversations, mock_conversations_api
):
"""Test creating a response with a valid conversation parameter."""
mock_conversations_api.list.return_value = ConversationItemList(
data=[], first_id=None, has_more=False, last_id=None, object="list"
)
async def mock_streaming_response(*args, **kwargs):
mock_response = OpenAIResponseObject(
id="resp_test123",
created_at=1234567890,
model="test-model",
object="response",
output=[
OpenAIResponseMessage(
id="msg_response",
content=[
OpenAIResponseOutputMessageContentOutputText(
text="Test response", type="output_text", annotations=[]
)
],
role="assistant",
status="completed",
type="message",
)
],
status="completed",
)
yield OpenAIResponseObjectStreamResponseCompleted(response=mock_response, type="response.completed")
responses_impl_with_conversations._create_streaming_response = mock_streaming_response
input_text = "Hello, how are you?"
conversation_id = "conv_test123"
response = await responses_impl_with_conversations.create_openai_response(
input=input_text, model="test-model", conversation=conversation_id, stream=False
)
assert response is not None
assert response.id == "resp_test123"
mock_conversations_api.list.assert_called_once_with(conversation_id, order="asc")
# Note: conversation sync happens in the streaming response flow,
# which is complex to mock fully in this unit test
async def test_create_response_with_invalid_conversation_id(self, responses_impl_with_conversations):
"""Test creating a response with an invalid conversation ID."""
with pytest.raises(InvalidConversationIdError) as exc_info:
await responses_impl_with_conversations.create_openai_response(
input="Hello", model="test-model", conversation="invalid_id", stream=False
)
assert "Expected an ID that begins with 'conv_'" in str(exc_info.value)
async def test_create_response_with_nonexistent_conversation(
self, responses_impl_with_conversations, mock_conversations_api
):
"""Test creating a response with a non-existent conversation."""
mock_conversations_api.list.side_effect = ConversationNotFoundError("conv_nonexistent")
with pytest.raises(ConversationNotFoundError) as exc_info:
await responses_impl_with_conversations.create_openai_response(
input="Hello", model="test-model", conversation="conv_nonexistent", stream=False
)
assert "not found" in str(exc_info.value)
async def test_conversation_and_previous_response_id(
self, responses_impl_with_conversations, mock_conversations_api, mock_responses_store
):
with pytest.raises(ValueError) as exc_info:
await responses_impl_with_conversations.create_openai_response(
input="test", model="test", conversation="conv_123", previous_response_id="resp_123"
)
assert "Mutually exclusive parameters" in str(exc_info.value)
assert "previous_response_id" in str(exc_info.value)
assert "conversation" in str(exc_info.value)