HYBIM-1011 Upgrade openai to 3.x.x(BLOCKED)

pyproject.toml

@@ -27,10 +27,10 @@ packages = [
 
 [project.optional-dependencies]
 langchain = ["langchain-core", "langchain", "langsmith (>=0.8.0)"]
-openai = ["openai (>=2.8.0,<3.0.0)", "packaging (>=24.2,<25.0)", "openai-agents (>=0.4.0,<1.0.0)", "cryptography (>=50.0.0)", "mcp (>=1.27.2,<2)"]
-crewai = ["crewai (>=0.152.0,<2.0.0); python_version < '3.14'", "litellm (>=1.83.14,<2.0.0); python_version < '3.14'", "uv (>=0.9.6); python_version < '3.14'", "aiohttp (>=3.14.1,<4); python_version < '3.14'", "cryptography (>=50.0.0)", "mcp (>=1.27.2,<2)", "pdfminer-six (>=20251107)"]
+openai = ["openai (>=3.0.0,<4.0.0)", "packaging (>=24.2,<25.0)", "openai-agents (>=0.21.0,<1.0.0)", "cryptography (>=50.0.0)", "mcp (>=1.27.2,<2)"]
+crewai = ["crewai (>=0.177.0,<2.0.0); python_version < '3.14'", "uv (>=0.9.6); python_version < '3.14'", "aiohttp (>=3.14.1,<4); python_version < '3.14'", "cryptography (>=50.0.0)", "mcp (>=1.27.2,<2)", "pdfminer-six (>=20251107)"]
 middleware = ["starlette"]
-all = ["langchain-core", "langchain", "langsmith (>=0.8.0)", "openai (>=2.8.0,<3.0.0)", "packaging (>=24.2,<25.0)", "openai-agents (>=0.4.0,<1.0.0)", "crewai (>=0.152.0,<2.0.0); python_version < '3.14'", "starlette", "litellm (>=1.83.14,<2.0.0); python_version < '3.14'", "uv (>=0.9.6); python_version < '3.14'", "aiohttp (>=3.14.1,<4); python_version < '3.14'", "cryptography (>=50.0.0)", "mcp (>=1.27.2,<2)", "pdfminer-six (>=20251107)"]
+all = ["langchain-core", "langchain", "langsmith (>=0.8.0)", "openai (>=3.0.0,<4.0.0)", "packaging (>=24.2,<25.0)", "openai-agents (>=0.21.0,<1.0.0)", "crewai (>=0.177.0,<2.0.0); python_version < '3.14'", "starlette", "uv (>=0.9.6); python_version < '3.14'", "aiohttp (>=3.14.1,<4); python_version < '3.14'", "cryptography (>=50.0.0)", "mcp (>=1.27.2,<2)", "pdfminer-six (>=20251107)"]
 
 
 
@@ -45,14 +45,12 @@ python-dateutil = "^2.8.0"
 langchain-core = { version = ">=1.3.3", optional = true }
 langchain = { version = ">=1.3.9,<2.0.0", optional = true }
 langsmith = { version = ">=0.8.0", optional = true }
-openai = { version = ">=2.8.0,<3.0.0", optional = true }
-openai-agents = { version = ">=0.4.0,<1.0.0", optional = true }
-litellm = { version = ">=1.83.14,<2.0.0", optional = true, python = ">=3.11,<3.14" }
+openai = { version = ">=3.0.0,<4.0.0", optional = true }
+openai-agents = { version = ">=0.21.0,<1.0.0", optional = true }
 galileo-core = "^4.4.0"
 httpx = ">=0.27.0,<0.29.0"
 starlette = { version = ">=1.3.1", optional = true }
 backoff = "^2.2.1"
-crewai = { version = ">=0.152.0,<2.0.0", optional = true, python = ">=3.11,<3.14" }
 tqdm = { version = ">=4.0.0" }
 typing-extensions = { version = ">=4.5.0" }
 opentelemetry-sdk = "^1.38.0"
@@ -93,7 +91,7 @@ time-machine = "^2.17.0"  # freezegun causes problems with pydantic model valida
 pre-commit = "^4.0.1"
 mypy = "^1.16.0"
 invoke = "^2.2.0"
-openai = ">=2.8.0,<3.0.0"
+openai = ">=3.0.0,<4.0.0"
 fastapi = "^0.135.0"
 ruff = "^0.15.22"
 openapi-python-client = "^0.29.0"

src/splunk_ao/handlers/crewai/handler.py

@@ -1,7 +1,6 @@
 import hashlib
 import json
 import logging
-from datetime import datetime
 from typing import Any
 from uuid import UUID
 
@@ -15,19 +14,16 @@
 _logger = logging.getLogger(__name__)
 
 # Lazy import state — crewai wraps sys.stdout/sys.stderr at import time, which
-# causes pytest-xdist worker hangs.  All crewai/litellm imports are deferred to
+# causes pytest-xdist worker hangs.  All crewai imports are deferred to
 # _resolve_crewai_imports(), called once from CrewAIEventListener.__init__.
 CREWAI_AVAILABLE = False
 CREWAI_EVENTS_MODULE_AVAILABLE = False
-LITE_LLM_AVAILABLE = False
-litellm = None
 _crewai_imports_resolved = False
 
 
 def _resolve_crewai_imports() -> None:
-    """Import crewai and litellm on first use, populating module-level globals."""
-    global CREWAI_AVAILABLE, CREWAI_EVENTS_MODULE_AVAILABLE
-    global LITE_LLM_AVAILABLE, litellm, _crewai_imports_resolved
+    """Import crewai on first use, populating module-level globals."""
+    global CREWAI_AVAILABLE, CREWAI_EVENTS_MODULE_AVAILABLE, _crewai_imports_resolved
 
     if _crewai_imports_resolved:
         return
@@ -43,15 +39,6 @@ def _resolve_crewai_imports() -> None:
     except ImportError:
         _logger.warning("CrewAI not available, using stubs")
 
-    try:
-        import litellm as _litellm
-
-        litellm = _litellm
-        LITE_LLM_AVAILABLE = True
-    except ImportError:
-        _logger.warning("LiteLLM not available, using stubs")
-        litellm = None
-
 
 class CrewAIEventListener:
     """
@@ -92,11 +79,6 @@ def __init__(
             except ImportError:
                 _logger.warning("Could not import crewai event bus, skipping listener setup")
 
-        if LITE_LLM_AVAILABLE and litellm is not None:
-            if not litellm.success_callback:
-                litellm.success_callback = []
-            litellm.success_callback.append(self.lite_llm_usage_callback)
-
     def setup_listeners(self, crewai_event_bus: Any) -> None:
         """Setup event listeners for CrewAI events."""
         if not CREWAI_AVAILABLE:
@@ -345,12 +327,7 @@ def _generate_run_id(self, source: Any, event: Any) -> UUID:
             task_id = getattr(getattr(event, "task", None), "id", "")
             return self._hash_to_uuid(f"{event_agent.id}_{task_id}")
 
-        # 7. dict messages — lite_llm callback
-        if isinstance(event, dict) and "messages" in event:
-            messages = json.dumps(event["messages"])
-            return self._hash_to_uuid(messages)
-
-        # 8. Generic fallback
+        # 7. Generic fallback
         return self._hash_to_uuid(
             f"{getattr(event, 'crew_name', '')}_{getattr(event, 'agent', '')}_{getattr(event, 'task', '')}"
         )
@@ -770,21 +747,3 @@ def _handle_memory_retrieval_failed(self, source: Any, event: Any) -> None:
 
         self._handler.end_node(run_id=run_id, output=f"Memory retrieval failed: {metadata['error']}", metadata=metadata)
 
-    def lite_llm_usage_callback(
-        self,
-        kwargs: dict,  # kwargs to completion
-        completion_response: Any,  # response from completion
-        start_time: datetime,
-        end_time: datetime,
-    ) -> None:
-        node_id = self._generate_run_id(kwargs, kwargs)
-
-        node = self._handler.get_node(node_id)
-        if not node:
-            _logger.debug(f"No node exists for run_id {node_id}")
-            return
-        usage = completion_response.model_extra["usage"]
-        node.span_params["usage"] = usage.model_dump() if hasattr(usage, "model_dump") else usage
-        node.span_params["num_input_tokens"] = getattr(usage, "prompt_tokens", 0)
-        node.span_params["num_output_tokens"] = getattr(usage, "completion_tokens", 0)
-        node.span_params["total_tokens"] = getattr(usage, "total_tokens", 0)

tests/test_crewai_handler.py

@@ -99,7 +99,6 @@ def crewai_callback(mock_splunk_ao_logger):
     with (
         patch("splunk_ao.handlers.crewai.handler._crewai_imports_resolved", True),
         patch("splunk_ao.handlers.crewai.handler.CREWAI_AVAILABLE", False),
-        patch("splunk_ao.handlers.crewai.handler.LITE_LLM_AVAILABLE", False),
     ):
         from splunk_ao.handlers.crewai.handler import CrewAIEventListener
 
@@ -113,7 +112,6 @@ def test_initialization_with_crewai_available(mock_splunk_ao_logger) -> None:
     with (
         patch("splunk_ao.handlers.crewai.handler._crewai_imports_resolved", True),
         patch("splunk_ao.handlers.crewai.handler.CREWAI_AVAILABLE", True),
-        patch("splunk_ao.handlers.crewai.handler.LITE_LLM_AVAILABLE", True),
     ):
         from splunk_ao.handlers.crewai.handler import CrewAIEventListener
 
@@ -131,7 +129,6 @@ def test_initialization_with_crewai_unavailable(mock_splunk_ao_logger) -> None:
     with (
         patch("splunk_ao.handlers.crewai.handler._crewai_imports_resolved", True),
         patch("splunk_ao.handlers.crewai.handler.CREWAI_AVAILABLE", False),
-        patch("splunk_ao.handlers.crewai.handler.LITE_LLM_AVAILABLE", False),
     ):
         from splunk_ao.handlers.crewai.handler import CrewAIEventListener
 
@@ -693,55 +690,6 @@ def test_update_crew_input(crewai_callback: CrewAIEventListener) -> None:
         assert "Research market trends" in mock_root_node.span_params["input"]
 
 
-def test_lite_llm_usage_callback(crewai_callback: CrewAIEventListener) -> None:
-    """Test LiteLLM usage callback."""
-    node_id = uuid.uuid4()
-
-    # Mock node
-    mock_node = Mock()
-    mock_node.span_params = {}
-
-    # Mock completion response with usage
-    mock_usage = Mock()
-    mock_usage.model_dump.return_value = {"prompt_tokens": 100, "completion_tokens": 50, "total_tokens": 150}
-    mock_usage.prompt_tokens = 100
-    mock_usage.completion_tokens = 50
-    mock_usage.total_tokens = 150
-
-    mock_response = Mock()
-    mock_response.model_extra = {"usage": mock_usage}
-
-    kwargs = {"messages": [{"role": "user", "content": "test"}]}
-
-    with (
-        patch.object(crewai_callback._handler, "get_node", return_value=mock_node),
-        patch.object(crewai_callback, "_generate_run_id", return_value=node_id),
-    ):
-        crewai_callback.lite_llm_usage_callback(
-            kwargs=kwargs, completion_response=mock_response, start_time=datetime.now(), end_time=datetime.now()
-        )
-
-        # Verify usage was recorded
-        assert mock_node.span_params["num_input_tokens"] == 100
-        assert mock_node.span_params["num_output_tokens"] == 50
-        assert mock_node.span_params["total_tokens"] == 150
-
-
-def test_lite_llm_usage_callback_no_node(crewai_callback) -> None:
-    """Test LiteLLM usage callback when node doesn't exist."""
-    kwargs = {"messages": [{"role": "user", "content": "test"}]}
-    mock_response = Mock()
-
-    with (
-        patch.object(crewai_callback._handler, "get_node", return_value=None),
-        patch.object(crewai_callback, "_generate_run_id", return_value=uuid.uuid4()),
-    ):
-        # Should not raise an exception
-        crewai_callback.lite_llm_usage_callback(
-            kwargs=kwargs, completion_response=mock_response, start_time=datetime.now(), end_time=datetime.now()
-        )
-
-
 # Memory event tests (for CrewAI >= 0.177.0)
 
 

tests/test_openai_agents.py

@@ -2,20 +2,22 @@
 import uuid
 from unittest.mock import MagicMock, Mock, patch
 
-import httpx
+import httpx2
 import pytest
-import respx
 import vcr
 from agents import (
     Agent,
     CodeInterpreterTool,
     GuardrailFunctionOutput,
     InputGuardrail,
     InputGuardrailTripwireTriggered,
+    OpenAIProvider,
+    RunConfig,
     Runner,
     set_trace_processors,
 )
 from agents.tracing import ResponseSpanData
+from openai import AsyncOpenAI
 from pydantic import BaseModel
 from pytest import MonkeyPatch, mark
 
@@ -333,16 +335,22 @@ async def capture_payload(payload):
 
     mock_response_data = _create_mock_response_with_tools(tool_calls)
 
-    with respx.mock(base_url="https://api.openai.com") as respx_mock:
-        respx_mock.post("/v1/responses").mock(return_value=httpx.Response(200, json=mock_response_data))
-        agent = Agent(
-            name="Assistant",
-            instructions="You are a helpful assistant.",
-            tools=[CodeInterpreterTool(tool_config={"type": "code_interpreter", "container": {"type": "auto"}})],
-        )
+    def mock_handler(request: httpx2.Request) -> httpx2.Response:
+        return httpx2.Response(200, json=mock_response_data)
 
-        result = await Runner.run(agent, "Test multiple tools.")
-        assert result
+    mock_openai = AsyncOpenAI(
+        api_key="sk-test",
+        http_client=httpx2.AsyncClient(transport=httpx2.MockTransport(mock_handler)),
+    )
+    provider = OpenAIProvider(openai_client=mock_openai)
+    agent = Agent(
+        name="Assistant",
+        instructions="You are a helpful assistant.",
+        tools=[CodeInterpreterTool(tool_config={"type": "code_interpreter", "container": {"type": "auto"}})],
+    )
+
+    result = await Runner.run(agent, "Test multiple tools.", run_config=RunConfig(model_provider=provider))
+    assert result
 
     traces = splunk_ao_logger.traces
     assert len(traces) == 1