From c2463728593b935448bcb09f54ef4fd070e9f8bf Mon Sep 17 00:00:00 2001
From: mateo-berri <277851410+mateo-berri@users.noreply.github.com>
Date: Sun, 13 Sep 2026 02:35:53 -0700
Subject: [PATCH] fix(responses): import BaseLLMException lazily and collect
 stream chunks via anext

Move the BaseLLMException import into _map_error_event_exception so the
module no longer imports it at load time, clearing the module-level cyclic
import CodeQL flagged. The class is used only on the cold error path.

Replace the mutable list-append test collector with aiter/anext so the
regression tests read the stream immutably.
---
 litellm/responses/streaming_iterator.py       |  3 ++-
 .../test_streaming_iterator_error_events.py   | 22 ++++++++-----------
 2 files changed, 11 insertions(+), 14 deletions(-)

diff --git a/litellm/responses/streaming_iterator.py b/litellm/responses/streaming_iterator.py
index b3426fbbfef..a15571acb7c 100644
--- a/litellm/responses/streaming_iterator.py
+++ b/litellm/responses/streaming_iterator.py
@@ -31,7 +31,6 @@ from litellm.litellm_core_utils.llm_response_utils.response_metadata import (
     update_response_metadata,
 )
 from litellm.litellm_core_utils.thread_pool_executor import executor
-from litellm.llms.base_llm.chat.transformation import BaseLLMException
 from litellm.llms.base_llm.responses.transformation import BaseResponsesAPIConfig
 from litellm.responses.utils import ResponseAPILoggingUtils, ResponsesAPIRequestUtils
 from litellm.types.llms.openai import (
@@ -553,6 +552,8 @@ class BaseResponsesAPIStreamingIterator:
         )
 
     def _map_error_event_exception(self, error_obj: object) -> Exception:
+        from litellm.llms.base_llm.chat.transformation import BaseLLMException
+
         error_message, error_type, error_code = _error_event_fields(error_obj)
         status_code: Final = _status_code_for_error_fields(error_type, error_code)
         error_body: Final = {"message": error_message, "type": error_type, "code": error_code}
diff --git a/tests/test_litellm/responses/test_streaming_iterator_error_events.py b/tests/test_litellm/responses/test_streaming_iterator_error_events.py
index 73afbb5e63a..2f4fba45cee 100644
--- a/tests/test_litellm/responses/test_streaming_iterator_error_events.py
+++ b/tests/test_litellm/responses/test_streaming_iterator_error_events.py
@@ -287,15 +287,12 @@ async def test_async_iterator_content_policy_violation_after_first_chunk_carries
         ]
     )
 
-    chunks = []
-
-    async def _drain():
-        async for chunk in iterator:
-            chunks.append(chunk)
+    stream = aiter(iterator)
+    first_chunk = await anext(stream)
+    assert first_chunk is not None
 
     with pytest.raises(MidStreamFallbackError) as exc_info:
-        await _drain()
-    assert len(chunks) == 1
+        await anext(stream)
     assert isinstance(exc_info.value.original_exception, litellm.ContentPolicyViolationError)
     assert exc_info.value.is_pre_first_chunk is False
     assert exc_info.value.generated_content == "partial "
@@ -316,14 +313,13 @@ async def test_async_iterator_error_after_first_chunk_carries_generated_content(
         ]
     )
 
-    chunks = []
-    async def _drain():
-        async for chunk in iterator:
-            chunks.append(chunk)
+    stream = aiter(iterator)
+    first_chunk = await anext(stream)
+    second_chunk = await anext(stream)
+    assert first_chunk is not None and second_chunk is not None
 
     with pytest.raises(MidStreamFallbackError) as exc_info:
-        await _drain()
-    assert len(chunks) == 2
+        await anext(stream)
     assert exc_info.value.status_code == 500
     assert exc_info.value.is_pre_first_chunk is False
     assert exc_info.value.generated_content == "hello world"