fix(responses): preserve citation ordering during tool stream sync

This commit is contained in:
haiyangbg 2026-10-29 15:09:25 +08:00
commit 42c95eb679
8 changed files with 570 additions and 82 deletions

View file

@ -11,3 +11,12 @@ ignore:
package:
name: zlib
type: apk
- vulnerability: GHSA-v4x8-3wr5-3wq8
reason: >-
filelock ReadWriteLock advisory, and no package in the image takes ReadWriteLock;
the fix (4.2.1) is blocked by litellm's own filelock<4.0 cap until #48635 or #48802 lifts it,
and the image then moves off 3.32.7, which retires this version-scoped rule
package:
name: filelock
version: 3.32.7
type: python

View file

@ -32,6 +32,7 @@ from litellm.responses.utils import ResponsesAPIRequestUtils
from litellm.types.llms.openai import (
PART_UNION_TYPES,
BaseLiteLLMOpenAIResponseObject,
ChatCompletionAnnotation,
ContentPartAddedEvent,
ContentPartDoneEvent,
ContentPartDonePartOutputText,
@ -59,6 +60,7 @@ from litellm.types.llms.openai import (
WebSearchCallInProgressEvent,
WebSearchCallSearchingEvent,
)
from litellm.types.responses.main import GenericResponseOutputItemContentAnnotation
from litellm.types.utils import Delta as ChatCompletionDelta
from litellm.types.utils import (
ModelResponse,
@ -108,7 +110,6 @@ def _delta_carries_output(delta: ChatCompletionDelta) -> bool:
getattr(delta, "reasoning_content", None),
delta.tool_calls,
delta.function_call,
getattr(delta, "annotations", None),
getattr(delta, "images", None),
getattr(delta, "audio", None),
)
@ -153,7 +154,8 @@ class LiteLLMCompletionStreamingIterator(ResponsesAPIStreamingIterator):
self.sent_output_text_done_event: bool = False
self.sent_output_content_part_done_event: bool = False
self.sent_output_item_done_event: bool = False
self.sent_annotation_events: bool = False
self._annotation_index: int = 0
self._annotations_awaiting_message: tuple[ChatCompletionAnnotation, ...] = ()
self.litellm_model_response: ModelResponse | TextCompletionResponse | None = None
self.completed_response = None
self.final_text: str = ""
@ -344,7 +346,7 @@ class LiteLLMCompletionStreamingIterator(ResponsesAPIStreamingIterator):
namespace_kwargs: Final = {"namespace": tool_namespace} if tool_namespace else {}
return {**item_kwargs, "name": tool_name, **namespace_kwargs}
def _is_reasoning_end(self, chunk):
def _is_reasoning_end(self, chunk: ModelResponseStream) -> bool:
if not chunk.choices:
return False
delta: Final = chunk.choices[0].delta
@ -355,7 +357,9 @@ class LiteLLMCompletionStreamingIterator(ResponsesAPIStreamingIterator):
if hasattr(delta, "thinking_blocks") and delta.thinking_blocks:
return False
return delta.content or delta.function_call or delta.tool_calls or chunk.choices[0].finish_reason is not None
return bool(
delta.content or delta.function_call or delta.tool_calls or chunk.choices[0].finish_reason is not None
)
def _reserve_web_search_indexes(self, provider_fields: object) -> None:
if not isinstance(provider_fields, dict):
@ -436,6 +440,8 @@ class LiteLLMCompletionStreamingIterator(ResponsesAPIStreamingIterator):
self._queue_named_tool_call_delta(call_id, fn_name, fn_args_delta)
def _queue_named_tool_call_delta(self, call_id: str, fn_name: str, fn_args_delta: str) -> None:
if call_id not in self._tool_output_index_by_call_id:
self._queue_held_annotations()
output_index: Final = self._get_or_assign_tool_output_index(call_id)
if not fn_name and call_id not in self._tool_item_id_by_call_id:
self._unnamed_tool_args_by_call_id = MappingProxyType(
@ -820,8 +826,7 @@ class LiteLLMCompletionStreamingIterator(ResponsesAPIStreamingIterator):
return event
def _queue_message_item_added_events(self) -> None:
if self._cached_item_id is None:
self._cached_item_id = f"msg_{uuid.uuid4()}"
item_id: Final = self._message_item_id()
self.sent_message_item_added_event = True
self.sent_content_part_added_event = True
self._sequence_number += 1
@ -830,7 +835,7 @@ class LiteLLMCompletionStreamingIterator(ResponsesAPIStreamingIterator):
output_index=self._message_index(),
item=BaseLiteLLMOpenAIResponseObject(
**{
"id": self._cached_item_id,
"id": item_id,
"type": "message",
"role": "assistant",
"status": "in_progress",
@ -1119,6 +1124,7 @@ class LiteLLMCompletionStreamingIterator(ResponsesAPIStreamingIterator):
if self.litellm_model_response:
self._queue_compaction_item_added()
self._queue_compaction_item_done()
self._queue_held_annotations()
if self._pending_response_events:
return self._pending_response_events.pop(0)
# If tool calls exist, emit tool events before finishing/response.completed.
@ -1201,7 +1207,11 @@ class LiteLLMCompletionStreamingIterator(ResponsesAPIStreamingIterator):
delta: Final = chunk.choices[0].delta if chunk.choices else None
if delta and getattr(delta, "reasoning_content", None):
self._accumulated_reasoning_content_parts.append(delta.reasoning_content)
if not self._is_reasoning_end(chunk):
if self._is_reasoning_end(chunk):
self._queue_reasoning_done_events()
def _queue_reasoning_done_events(self) -> None:
if not self._reasoning_active or self._reasoning_done_emitted:
return
reasoning_content: Final = "".join(self._accumulated_reasoning_content_parts)
self._cached_reasoning_item_id = (
@ -1275,9 +1285,7 @@ class LiteLLMCompletionStreamingIterator(ResponsesAPIStreamingIterator):
self._snapshot_chunk_for_stream_chunk_builder(chunk)
)
self._queue_reasoning_done_events_on_reasoning_end(chunk)
response_api_chunk = self._transform_chat_completion_chunk_to_response_api_chunk(chunk)
if response_api_chunk:
self._pending_response_events.append(response_api_chunk)
self._queue_chunk_events(chunk)
if self._pending_response_events:
return self._pending_response_events.pop(0)
@ -1338,9 +1346,7 @@ class LiteLLMCompletionStreamingIterator(ResponsesAPIStreamingIterator):
self._snapshot_chunk_for_stream_chunk_builder(cast(ModelResponseStream, chunk))
)
self._queue_reasoning_done_events_on_reasoning_end(chunk)
response_api_chunk = self._transform_chat_completion_chunk_to_response_api_chunk(chunk)
if response_api_chunk:
self._pending_response_events.append(response_api_chunk)
self._queue_chunk_events(chunk)
if self._pending_response_events:
return self._pending_response_events.pop(0)
# Otherwise, loop to next chunk
@ -1351,6 +1357,58 @@ class LiteLLMCompletionStreamingIterator(ResponsesAPIStreamingIterator):
self.finished = True
raise e
def _queue_chunk_events(self, chunk: ModelResponseStream) -> None:
self._hold_chunk_annotations(chunk)
response_api_chunk: Final = self._transform_chat_completion_chunk_to_response_api_chunk(chunk)
if response_api_chunk:
self._pending_response_events.append(response_api_chunk)
if self.sent_message_item_added_event:
self._queue_held_annotations()
def _hold_chunk_annotations(self, chunk: ModelResponseStream) -> None:
chunk_annotations: Final = (getattr(chunk.choices[0].delta, "annotations", None) or ()) if chunk.choices else ()
self._annotations_awaiting_message = (*self._annotations_awaiting_message, *chunk_annotations)
def _queue_held_annotations(self) -> None:
annotations: Final = self._annotations_awaiting_message
if not annotations:
return
self._annotations_awaiting_message = ()
self._queue_reasoning_done_events()
self._queue_annotation_added_events(annotations)
def _queue_annotation_added_events(self, annotations: Sequence[ChatCompletionAnnotation]) -> None:
if not self.sent_message_item_added_event:
self._queue_message_item_added_events()
item_id: Final = self._message_item_id()
response_annotations: Final = (
LiteLLMCompletionResponsesConfig.transform_chat_completion_annotations_to_response_output_annotations(
annotations=annotations
)
)
self._pending_response_events.extend(
self._annotation_added_event(item_id, annotation) for annotation in response_annotations
)
def _message_item_id(self) -> str:
if self._cached_item_id is None:
self._cached_item_id = f"msg_{uuid.uuid4()}"
return self._cached_item_id
def _annotation_added_event(
self, item_id: str, annotation: GenericResponseOutputItemContentAnnotation
) -> OutputTextAnnotationAddedEvent:
event: Final = OutputTextAnnotationAddedEvent(
type=ResponsesAPIStreamEvents.OUTPUT_TEXT_ANNOTATION_ADDED,
item_id=item_id,
output_index=self._message_index(),
content_index=0,
annotation_index=self._annotation_index,
annotation=annotation.model_dump(),
)
self._annotation_index += 1
return event
def _transform_chat_completion_chunk_to_response_api_chunk(
self, chunk: ModelResponseStream
) -> ResponsesAPIStreamingResponse | None:
@ -1359,37 +1417,9 @@ class LiteLLMCompletionStreamingIterator(ResponsesAPIStreamingIterator):
This currently handles emitting the OutputTextDeltaEvent, which is used by other tools using the responses API
and the ReasoningSummaryTextDeltaEvent, which is used by the responses API to emit reasoning content.
It also handles emitting annotation.added events when annotations are detected in the chunk.
"""
if self._cached_item_id is None:
self._cached_item_id = f"msg_{uuid.uuid4()}"
item_id: Final = self._cached_item_id
item_id: Final = self._message_item_id()
# Check if this chunk has annotations first (before processing text/reasoning)
# This ensures we detect and queue annotation events from the annotation chunk
if chunk.choices and hasattr(chunk.choices[0].delta, "annotations"):
annotations: Final = chunk.choices[0].delta.annotations
if annotations and self.sent_annotation_events is False:
self.sent_annotation_events = True
# Store annotation events to emit them one by one
if not hasattr(self, "_pending_annotation_events"):
response_annotations = LiteLLMCompletionResponsesConfig.transform_chat_completion_annotations_to_response_output_annotations(
annotations=annotations
)
self._pending_annotation_events = []
for idx, annotation in enumerate(response_annotations):
annotation_dict = (
annotation.model_dump() if hasattr(annotation, "model_dump") else dict(annotation)
)
event = OutputTextAnnotationAddedEvent(
type=ResponsesAPIStreamEvents.OUTPUT_TEXT_ANNOTATION_ADDED,
item_id=item_id,
output_index=self._message_index(),
content_index=0,
annotation_index=idx,
annotation=annotation_dict,
)
self._pending_annotation_events.append(event)
# Priority 1: Handle reasoning content (highest priority)
if (
chunk.choices
@ -1433,13 +1463,6 @@ class LiteLLMCompletionStreamingIterator(ResponsesAPIStreamingIterator):
if self._pending_tool_events:
return self._pending_tool_events.pop(0)
# Priority 4: If we have pending annotation events, emit the next one
# This happens when the current chunk has no text/reasoning content
if hasattr(self, "_pending_annotation_events") and self._pending_annotation_events:
event = self._pending_annotation_events.pop(0)
return event
# Priority 5: If we have pending tool events (from earlier chunk), emit the next one
if self._pending_tool_events:
return self._pending_tool_events.pop(0)
@ -1482,7 +1505,10 @@ class LiteLLMCompletionStreamingIterator(ResponsesAPIStreamingIterator):
def _is_unstreamed_empty_message(self, item: object) -> bool:
if getattr(item, "type", None) != "message" or self._message_output_index is not None:
return False
return not any(getattr(part, "text", None) for part in attribute_items(item, "content"))
return not any(
getattr(part, "text", None) or getattr(part, "annotations", None)
for part in attribute_items(item, "content")
)
def finished_round_event(self) -> ResponseCompletedEvent | ResponseIncompleteEvent | None:
if not isinstance(self.litellm_model_response, ModelResponse):

View file

@ -3555,7 +3555,7 @@ class LiteLLMCompletionResponsesConfig:
choice=choice,
)
message_output_items.extend(image_generation_items)
elif choice.message.content is not None:
elif choice.message.content is not None or getattr(choice.message, "annotations", None):
message_output_items.append(
GenericResponseOutputItem(
type="message",
@ -3629,13 +3629,13 @@ class LiteLLMCompletionResponsesConfig:
return OutputText(
type="output_text",
text=message.content,
text=message.content or "",
annotations=transformed_annotations,
)
@staticmethod
def transform_chat_completion_annotations_to_response_output_annotations(
annotations: list[ChatCompletionAnnotation] | None,
annotations: Sequence[ChatCompletionAnnotation] | None,
) -> list[GenericResponseOutputItemContentAnnotation]:
response_output_annotations: Final[list[GenericResponseOutputItemContentAnnotation]] = []

View file

@ -1,27 +1,102 @@
[[IgnoredVulns]]
id = "GHSA-w8v5-vhqr-4h9v"
ignoreUntil = 2026-11-01
reason = "diskcache has no fixed release published; remove this entry once one exists"
[[IgnoredVulns]]
id = "GHSA-v4x8-3wr5-3wq8"
ignoreUntil = 2026-11-20
reason = "the only fixed filelock is 4.2.1, a new major that requires Python >= 3.11 while litellm still supports 3.10, and the advisory is in ReadWriteLock, which litellm never uses (only FileLock); bump filelock and remove this entry once 3.10 is dropped or a 3.x fix ships"
reason = "filelock ReadWriteLock advisory, which no package in the lock takes (litellm and huggingface_hub, filelock's only dependents, use just FileLock and SoftFileLock); fixed only in 4.2.1, and filelock 4.1.1+ needs Python >= 3.11; litellm's own filelock<4.0 cap keeps every Python on 3.32.7 until it is lifted for 3.11+ (#48635, #48802); remove once that lands and 3.10 is dropped"
[[IgnoredVulns]]
id = "GHSA-vfj7-8cjw-p6xm"
ignoreUntil = 2026-11-03
reason = "braces has no fixed release published (3.0.3, 2024-05-21, is latest and last_affected); dev-only via knip > fast-glob > micromatch; remove this entry once one exists"
id = "GHSA-567q-5qvw-pf92"
ignoreUntil = 2026-11-20
reason = "Windows-only local DoS (CVSS 2.5) in filelock soft locks; litellm never takes SoftFileLock, but huggingface_hub falls back to it in the Hugging Face cache when the filesystem has no flock, which litellm reaches only to download the Kimi K3 tokenizer for Bedrock token counting, and the attack needs another local user who can write that cache; fixed only in 4.2.1, and filelock 4.1.1+ needs Python >= 3.11; litellm's own filelock<4.0 cap keeps every Python on 3.32.7 until it is lifted for 3.11+ (#48635, #48802); remove once that lands and 3.10 is dropped"
[[IgnoredVulns]]
id = "GHSA-6q48-6c7w-9gw2"
ignoreUntil = 2026-10-23
reason = "certifi 2026.10.19 (published 2026-10-20) is inside the 3-day uv exclude-newer window until 2026-10-23 02:24 UTC; bump certifi and remove this entry once it clears"
id = "GHSA-9vhv-jxww-xfm5"
ignoreUntil = 2026-11-20
reason = "Windows-only local DoS (CVSS 3.3) in filelock ReadWriteLock, which no package in the lock takes (litellm and huggingface_hub, filelock's only dependents, use just FileLock and SoftFileLock); fixed only in 4.2.1, and filelock 4.1.1+ needs Python >= 3.11; litellm's own filelock<4.0 cap keeps every Python on 3.32.7 until it is lifted for 3.11+ (#48635, #48802); remove once that lands and 3.10 is dropped"
[[IgnoredVulns]]
id = "GHSA-6627-m3xw-93mp"
ignoreUntil = 2026-11-20
reason = "filelock StrictSoftFileLock claim directories ignore the lock mode, which no package in the lock takes (litellm and huggingface_hub, filelock's only dependents, use just FileLock and SoftFileLock); fixed only in 4.2.1, and filelock 4.1.1+ needs Python >= 3.11; litellm's own filelock<4.0 cap keeps every Python on 3.32.7 until it is lifted for 3.11+ (#48635, #48802); remove once that lands and 3.10 is dropped"
[[IgnoredVulns]]
id = "GHSA-69hj-f68j-2879"
ignoreUntil = 2026-11-20
reason = "filelock UnixFileLock truncates and chmods a file another user renames or hard-links in at the lock path; litellm's CLI takes FileLock only inside ~/.litellm, which it creates owner-only, and huggingface_hub takes it in the Hugging Face cache, which litellm reaches only to download the Kimi K3 tokenizer for Bedrock token counting; the attack needs another local user who can rename or link files in that cache; fixed only in 4.2.1, and filelock 4.1.1+ needs Python >= 3.11; litellm's own filelock<4.0 cap keeps every Python on 3.32.7 until it is lifted for 3.11+ (#48635, #48802); remove once that lands and 3.10 is dropped"
[[IgnoredVulns]]
id = "GHSA-8mx5-7m5m-r77g"
ignoreUntil = 2026-11-20
reason = "filelock WindowsFileLock ignores mode; the proxy image is Linux, so only Windows installs run it; there litellm's CLI takes FileLock only inside the owner-only ~/.litellm, whose lock file holds no data, and huggingface_hub takes it in the Hugging Face cache, which litellm reaches only to download the Kimi K3 tokenizer for Bedrock token counting; fixed only in 4.2.1, and filelock 4.1.1+ needs Python >= 3.11; litellm's own filelock<4.0 cap keeps every Python on 3.32.7 until it is lifted for 3.11+ (#48635, #48802); remove once that lands and 3.10 is dropped"
[[IgnoredVulns]]
id = "GHSA-9c4x-gm8v-wgc4"
ignoreUntil = 2026-11-20
reason = "filelock StrictSoftFileLock and SoftReadWriteLock adopt another user's protocol directory, which no package in the lock takes (litellm and huggingface_hub, filelock's only dependents, use just FileLock and SoftFileLock); fixed only in 4.2.1, and filelock 4.1.1+ needs Python >= 3.11; litellm's own filelock<4.0 cap keeps every Python on 3.32.7 until it is lifted for 3.11+ (#48635, #48802); remove once that lands and 3.10 is dropped"
[[IgnoredVulns]]
id = "GHSA-9j6q-wvj5-jgjg"
ignoreUntil = 2026-11-20
reason = "filelock SoftFileLock symlink swap leaks whether a path exists; litellm never takes SoftFileLock, but huggingface_hub falls back to it in the Hugging Face cache when the filesystem has no flock, which litellm reaches only to download the Kimi K3 tokenizer for Bedrock token counting, and the attack needs another local user who can write that cache; fixed only in 4.2.1, and filelock 4.1.1+ needs Python >= 3.11; litellm's own filelock<4.0 cap keeps every Python on 3.32.7 until it is lifted for 3.11+ (#48635, #48802); remove once that lands and 3.10 is dropped"
[[IgnoredVulns]]
id = "GHSA-cwqr-hh33-gh44"
ignoreUntil = 2026-11-20
reason = "filelock ReadWriteLock stalls on a crafted lock database, which no package in the lock takes (litellm and huggingface_hub, filelock's only dependents, use just FileLock and SoftFileLock); fixed only in 4.2.1, and filelock 4.1.1+ needs Python >= 3.11; litellm's own filelock<4.0 cap keeps every Python on 3.32.7 until it is lifted for 3.11+ (#48635, #48802); remove once that lands and 3.10 is dropped"
[[IgnoredVulns]]
id = "GHSA-fq56-3cx3-fjxp"
ignoreUntil = 2026-11-20
reason = "filelock SoftFileLock stale-lock break removes a live lock on a reused inode; litellm never takes SoftFileLock, but huggingface_hub falls back to it in the Hugging Face cache when the filesystem has no flock, which litellm reaches only to download the Kimi K3 tokenizer for Bedrock token counting, and the attack needs another local user who can write that cache; fixed only in 4.2.1, and filelock 4.1.1+ needs Python >= 3.11; litellm's own filelock<4.0 cap keeps every Python on 3.32.7 until it is lifted for 3.11+ (#48635, #48802); remove once that lands and 3.10 is dropped"
[[IgnoredVulns]]
id = "GHSA-j8f7-rjxc-mr56"
ignoreUntil = 2026-11-20
reason = "filelock ReadWriteLock follows a symlink into SQLite, which no package in the lock takes (litellm and huggingface_hub, filelock's only dependents, use just FileLock and SoftFileLock); fixed in 4.0.9, which litellm's own filelock<4.0 cap blocks on every Python until it is lifted (#48635, #48802); remove once that lands"
[[IgnoredVulns]]
id = "GHSA-m462-7rxw-wvpg"
ignoreUntil = 2026-11-20
reason = "filelock StrictSoftFileLock follows a symlink swapped in at <path>.filelock, which no package in the lock takes (litellm and huggingface_hub, filelock's only dependents, use just FileLock and SoftFileLock); fixed only in 4.2.1, and filelock 4.1.1+ needs Python >= 3.11; litellm's own filelock<4.0 cap keeps every Python on 3.32.7 until it is lifted for 3.11+ (#48635, #48802); remove once that lands and 3.10 is dropped"
[[IgnoredVulns]]
id = "GHSA-rfpc-qcw5-3pc4"
ignoreUntil = 2026-11-20
reason = "filelock SoftFileLock removes a live holder's marker when another user forges its PID; litellm never takes SoftFileLock, but huggingface_hub falls back to it in the Hugging Face cache when the filesystem has no flock, which litellm reaches only to download the Kimi K3 tokenizer for Bedrock token counting, and the attack needs another local user who can write that cache; fixed only in 4.2.1, and filelock 4.1.1+ needs Python >= 3.11; litellm's own filelock<4.0 cap keeps every Python on 3.32.7 until it is lifted for 3.11+ (#48635, #48802); remove once that lands and 3.10 is dropped"
[[IgnoredVulns]]
id = "GHSA-pcg2-g6cj-q92c"
ignoreUntil = 2026-11-27
reason = "tornado 6.6 (published 2026-10-21) is the only fix and requires Python >= 3.11, so the Python 3.10 resolution stays on 6.5.10 while 3.11+ locks 6.6; litellm starts no tornado server or client, since semantic-router and aurelio-sdk only declare it and tenacity and sentry-sdk import it only to wrap tornado code when it is present, so tornado's HTTP header parsing is never reached; remove this entry once 3.10 is dropped"
[[IgnoredVulns]]
id = "GHSA-pw9p-v7r6-gx4p"
ignoreUntil = 2026-11-27
reason = "tornado 6.6 (published 2026-10-21) is the only fix and requires Python >= 3.11, so the Python 3.10 resolution stays on 6.5.10 while 3.11+ locks 6.6; litellm starts no tornado server or client, since semantic-router and aurelio-sdk only declare it and tenacity and sentry-sdk import it only to wrap tornado code when it is present, so tornado's RedirectHandler is never reached; remove this entry once 3.10 is dropped"
[[IgnoredVulns]]
id = "GHSA-r4j5-9xpp-3985"
ignoreUntil = 2026-11-27
reason = "tornado 6.6 (published 2026-10-21) is the only fix and requires Python >= 3.11, so the Python 3.10 resolution stays on 6.5.10 while 3.11+ locks 6.6; litellm starts no tornado server or client, since semantic-router and aurelio-sdk only declare it and tenacity and sentry-sdk import it only to wrap tornado code when it is present, so tornado's chunked body reading is never reached; remove this entry once 3.10 is dropped"
[[IgnoredVulns]]
id = "GHSA-w8v5-vhqr-4h9v"
ignoreUntil = 2026-11-27
reason = "diskcache still has no fixed release published (checked 2026-10-28), and it is reached only through the optional caching extra; remove this entry once one exists"
[[IgnoredVulns]]
id = "GHSA-45c3-73f7-pv4m"
ignoreUntil = 2026-11-27
reason = "dnspython 2.9.0 (published 2026-10-09) is the only fix and requires Python >= 3.11, so the Python 3.10 resolution stays on 2.8.0 while 3.11+ locks 2.9.0; the advisory is CPU use in dns.name.from_wire_parser and litellm only reaches dnspython through pydantic EmailStr, which never does DNS lookups; remove this entry once 3.10 is dropped or a 3.10-compatible fix ships"
[[IgnoredVulns]]
id = "GHSA-xcx6-jx8p-x3h5"
ignoreUntil = 2026-11-24
reason = "@azure/msal-node is fixed only in 7.1.0 (published 2026-10-13), but it is dev-only in vscode-extension via @vscode/vsce > @azure/identity, and the latest @azure/identity (4.14.0) still requires msal-node ^6.0.0, so the lock cannot move without forcing a major override; the advisory is in the auth_code flow, which extension packaging does not run; bump and remove this entry once @azure/identity accepts msal-node >= 7.1.0"
[[PackageOverrides]]
name = "pillow"
version = "12.3.0"
ecosystem = "PyPI"
vulnerability.ignore = true
effectiveUntil = 2026-10-27
reason = "pillow 13.0.0 requires Python >= 3.11, so the Python 3.10 resolution of the ci and dev groups stays on 12.3.0; 3.11+ is on 13.0.0"
effectiveUntil = 2026-11-27
reason = "pillow 13.0.0 requires Python >= 3.11, so the Python 3.10 resolution of the ci and dev groups and of the mlflow extra stays on 12.3.0; 3.11+ is on 13.0.0; pillow has never shipped an older-major release after the next major (no 10.x after 11.0.0, no 11.x after 12.0.0), so 13.x is the only fix line; extended to 2026-11-27 with maintainer approval, same date as the dnspython 3.10 entry; remove once 3.10 is dropped"

View file

@ -360,11 +360,13 @@ cache-keys = [
]
constraint-dependencies = [
"tornado>=6.5.8",
"tornado>=6.6; python_version >= '3.11'",
"aiohttp>=3.14.2,<4.0",
"packaging>=24.0",
"soupsieve>=2.8.4",
"httplib2>=0.32.0",
"setuptools>=84.1.0",
"dnspython>=2.9.0; python_version >= '3.11'",
]
override-dependencies = [
# a2a-sdk 1.x requires packaging>=24.0; lunary 1.4.x still caps at <24.0.

View file

@ -663,6 +663,46 @@ class TestLiteLLMCompletionResponsesConfig:
assert "message" not in output_types
assert "function_call" in output_types
def test_cited_tool_call_only_response_keeps_its_citations_in_a_message_item(self) -> None:
citation: Final = {"start_index": 0, "end_index": 0, "url": "https://example.com", "title": "Example"}
chat_completion_response: Final = ModelResponse(
id="test-response-id",
created=1234567890,
model="gemini-3.8-flash",
object="chat.completion",
choices=[
Choices(
finish_reason="tool_calls",
index=0,
message=Message(
content=None,
role="assistant",
annotations=[{"type": "url_citation", "url_citation": citation}],
tool_calls=[
ChatCompletionMessageToolCall(
id="call_cited",
type="function",
function=Function(name="save_museum", arguments='{"name": "Louvre"}'),
)
],
),
)
],
)
responses_api_response: Final = (
LiteLLMCompletionResponsesConfig.transform_chat_completion_response_to_responses_api_response(
request_input="which museum?",
responses_api_request={},
chat_completion_response=chat_completion_response,
)
)
output: Final = responses_api_response.output
assert [item.type for item in output] == ["message", "function_call"]
assert output[0].content[0].text == ""
assert [annotation.url for annotation in output[0].content[0].annotations] == [citation["url"]]
def test_content_bearing_response_still_emits_message_item(self):
"""Turns with real text content must keep their message output item."""
chat_completion_response = ModelResponse(

View file

@ -13,6 +13,7 @@ spend tracking stores, so a follow-up previous_response_id still finds the conve
import itertools
import json
from collections.abc import Callable
from itertools import accumulate, chain, islice, product
from typing import Final
from unittest.mock import AsyncMock, MagicMock
@ -239,14 +240,16 @@ async def test_custom_tool_input_events_round_trip_with_interleaved_function(
@pytest.mark.parametrize("sync_mode", [True, False])
@pytest.mark.parametrize("tool_type", ["custom", "function"])
@pytest.mark.parametrize("late_name", [True, False])
@pytest.mark.parametrize("citation_first", [True, False])
@pytest.mark.asyncio
async def test_emitted_tool_call_prevents_text_only_continuation(
sync_mode: bool, tool_type: str, late_name: bool
sync_mode: bool, tool_type: str, late_name: bool, citation_first: bool
) -> None:
iterator: Final = LiteLLMCompletionStreamingIterator(
model="test-model",
litellm_custom_stream_wrapper=_FakeStreamWrapper(
[
*([_annotation_only_chunk()] if citation_first else []),
_custom_tool_stream_chunk(
Delta(
content=None,
@ -279,10 +282,12 @@ async def test_emitted_tool_call_prevents_text_only_continuation(
)
events: Final = await _collect_events(iterator, sync_mode)
expected_type: Final = "custom_tool_call" if tool_type == "custom" else "function_call"
assert any(
event.type == "response.output_item.added" and event.item.type == expected_type
added_types: Final = tuple(
event.item.type
for event in events
if event.type == "response.output_item.added"
)
assert added_types == (("message", expected_type) if citation_first else (expected_type,))
assert iterator.started_tool_call
@ -1176,6 +1181,15 @@ def _with_reasoning(chunk: ModelResponseStream, reasoning: str) -> ModelResponse
)
def _with_citation(chunk: ModelResponseStream) -> ModelResponseStream:
choice: Final = chunk.choices[0]
annotations: Final = _annotation_only_chunk().choices[0].delta.annotations
delta: Final = Delta(**{**choice.delta.model_dump(), "annotations": annotations})
return chunk.model_copy(
update={"choices": [StreamingChoices(index=0, delta=delta, finish_reason=choice.finish_reason)]}
)
def _reasoning_then_text_chunks() -> tuple[ModelResponseStream, ...]:
return (_reasoning_chunk("let me think"), _chunk("Hello"), _chunk("!", finish_reason="stop"))
@ -1849,6 +1863,266 @@ async def test_a_compaction_block_without_its_summary_streams_no_compaction_item
assert [item["type"] for item in events[-1]["response"]["output"]] == ["message"]
@pytest.mark.parametrize("sync_mode", [True, False])
@pytest.mark.parametrize("batches", [(5,), (2, 3)])
@pytest.mark.parametrize("with_text", [True, False])
@pytest.mark.parametrize("reasoning_first", [True, False])
@pytest.mark.asyncio
async def test_all_citation_batches_stream_before_message_done(
sync_mode: bool, batches: tuple[int, ...], with_text: bool, reasoning_first: bool
) -> None:
annotations: Final = [
{
"type": "url_citation",
"url_citation": {
"start_index": 0,
"end_index": 5,
"url": f"https://example.com/{index}",
"title": f"Source {index}",
},
}
for index in range(sum(batches))
]
citation_chunks: Final = tuple(
ModelResponseStream(
id=CHAT_COMPLETION_ID,
created=1748575031,
model="claude-haiku-4-5",
object="chat.completion.chunk",
choices=[
StreamingChoices(
index=0,
delta=Delta(
role="assistant",
content="!" if with_text else "",
annotations=annotations[end - size : end],
),
finish_reason="stop" if end == len(annotations) else None,
)
],
)
for size, end in zip(batches, accumulate(batches), strict=True)
)
reasoning_chunks: Final = _reasoning_then_text_chunks()[:1] if reasoning_first else ()
chunks: Final = (*reasoning_chunks, _chunk("Claim"), *citation_chunks)
events: Final = await _collect_events(_build_iterator(chunks), sync_mode)
added: Final = [event for event in events if event.type == "response.output_text.annotation.added"]
terminal: Final = events[-1]
message: Final = next(item for item in terminal.response.output if item.type == "message")
final_annotations: Final = [annotation.model_dump() for annotation in message.content[0].annotations]
assert [event.annotation for event in added] == final_annotations
assert len(added) == len(annotations)
assert [event.annotation_index for event in added] == list(range(len(annotations)))
message_index: Final = next(
event.output_index
for event in events
if event.type == "response.output_item.added" and event.item.type == "message"
)
assert all(event.item_id == message.id and event.output_index == message_index for event in added)
done_index: Final = next(index for index, event in enumerate(events) if event.type == "response.output_text.done")
assert all(events.index(event) < done_index for event in added)
assert message.content[0].text == "Claim" + ("!" * len(batches) if with_text else "")
def _citation_after_reasoning_chunks() -> tuple[ModelResponseStream, ...]:
return (_reasoning_chunk("let me think"), _annotation_only_chunk())
def _citation_with_reasoning_chunks() -> tuple[ModelResponseStream, ...]:
return (_with_reasoning(_annotation_only_chunk(), "let me think"),)
def _citation_between_reasoning_chunks() -> tuple[ModelResponseStream, ...]:
return (_reasoning_chunk("let me "), _annotation_only_chunk(), _reasoning_chunk("think"))
@pytest.mark.parametrize("sync_mode", [True, False])
@pytest.mark.parametrize(
"stream_chunks",
[_citation_after_reasoning_chunks, _citation_with_reasoning_chunks, _citation_between_reasoning_chunks],
ids=["citation_after_reasoning", "citation_with_reasoning", "citation_between_reasoning"],
)
@pytest.mark.asyncio
async def test_citation_during_reasoning_streams_once_reasoning_closes(
stream_chunks: StreamChunks, sync_mode: bool
) -> None:
chunks: Final = (*stream_chunks(), _chunk("Claim", finish_reason="stop"))
events: Final = await _collect_events(_build_iterator(chunks), sync_mode)
output: Final = events[-1].response.output
message_index: Final = next(index for index, item in enumerate(output) if item.type == "message")
message: Final = output[message_index]
reasoning: Final = next(item for item in output if item.type == "reasoning")
reasoning_done: Final = next(
index
for index, event in enumerate(events)
if event.type == "response.output_item.done" and event.item.type == "reasoning"
)
message_added: Final = next(index for index, event in enumerate(events) if _is_message_item(event))
part_added: Final = next(index for index, event in enumerate(events) if event.type == "response.content_part.added")
text_done: Final = next(index for index, event in enumerate(events) if event.type == "response.output_text.done")
citations: Final = [event for event in events if event.type == "response.output_text.annotation.added"]
reasoning_deltas: Final = [
index for index, event in enumerate(events) if event.type == "response.reasoning_summary_text.delta"
]
summary_done: Final = next(event for event in events if event.type == "response.reasoning_summary_text.done")
text_delta: Final = next(event for event in events if event.type == "response.output_text.delta")
assert len(citations) == 1
assert reasoning_done < message_added < part_added < events.index(citations[0]) < text_done
assert all(index < reasoning_done for index in reasoning_deltas)
assert summary_done.text == reasoning.content[0].text == "let me think"
assert citations[0].item_id == message.id
assert citations[0].output_index == message_index
assert citations[0].annotation_index == 0
assert [event.annotation for event in citations] == [entry.model_dump() for entry in message.content[0].annotations]
assert ("sequence_number" in citations[0].model_dump()) == ("sequence_number" in text_delta.model_dump())
assert message.content[0].text == "Claim"
def _citation_then_tool_call_finishing_chunks() -> tuple[ModelResponseStream, ...]:
return (_annotation_only_chunk(), _tool_call_chunk(finish_reason="tool_calls"))
def _reasoning_citation_then_tool_call_finishing_chunks() -> tuple[ModelResponseStream, ...]:
return (_reasoning_chunk("let me think"), *_citation_then_tool_call_finishing_chunks())
def _reasoning_citation_tool_call_then_empty_finish_chunks() -> tuple[ModelResponseStream, ...]:
return (
_reasoning_chunk("let me think"),
_annotation_only_chunk(),
_tool_call_chunk(),
_chunk("", finish_reason="tool_calls"),
)
def _citation_between_reasoning_then_empty_finish_chunks() -> tuple[ModelResponseStream, ...]:
return (*_citation_between_reasoning_chunks(), _chunk("", finish_reason="stop"))
def _tool_call_then_citation_finishing_chunks() -> tuple[ModelResponseStream, ...]:
return (_tool_call_chunk(), _annotation_only_chunk())
@pytest.mark.parametrize("sync_mode", [True, False])
@pytest.mark.parametrize(
"stream_chunks",
[
_citation_then_tool_call_finishing_chunks,
_reasoning_citation_then_tool_call_finishing_chunks,
_reasoning_citation_tool_call_then_empty_finish_chunks,
_citation_between_reasoning_then_empty_finish_chunks,
_citation_between_reasoning_chunks,
_tool_call_then_citation_finishing_chunks,
],
ids=[
"citation_then_tool_call",
"reasoning_citation_then_tool_call",
"reasoning_citation_tool_call_then_empty_finish",
"citation_between_reasoning_then_empty_finish",
"citation_between_reasoning_without_finish",
"tool_call_then_citation_without_text",
],
)
@pytest.mark.asyncio
async def test_streamed_citations_match_the_completed_message(stream_chunks: StreamChunks, sync_mode: bool) -> None:
events: Final = await _collect_events(_build_iterator(stream_chunks()), sync_mode)
output: Final = events[-1].response.output
completed_message_indexes: Final = [index for index, item in enumerate(output) if item.type == "message"]
completed_citations: Final = chain.from_iterable(
output[index].content[0].annotations for index in completed_message_indexes
)
completed_annotations: Final = [annotation.model_dump() for annotation in completed_citations]
streamed_message_indexes: Final = [event.output_index for event in events if _is_message_item(event)]
citations: Final = [event for event in events if event.type == "response.output_text.annotation.added"]
reasoning_done: Final = [
index
for index, event in enumerate(events)
if event.type == "response.output_item.done" and event.item.type == "reasoning"
]
message_added: Final = [index for index, event in enumerate(events) if _is_message_item(event)]
assert len(citations) == 1
assert [event.annotation for event in citations] == completed_annotations
assert sorted(set(streamed_message_indexes)) == completed_message_indexes
assert all(event.output_index in completed_message_indexes for event in citations)
assert len(reasoning_done) == sum(item.type == "reasoning" for item in output)
assert all(index < added for index, added in product(reasoning_done, message_added))
def _empty_text_citation_then_tool_call_chunks() -> tuple[ModelResponseStream, ...]:
return (_chunk(""), *_citation_then_tool_call_finishing_chunks())
def _reasoning_empty_text_citation_then_tool_call_chunks() -> tuple[ModelResponseStream, ...]:
return (_reasoning_chunk("let me think"), *_empty_text_citation_then_tool_call_chunks())
def _empty_text_then_cited_tool_call_chunks() -> tuple[ModelResponseStream, ...]:
return (_chunk(""), _with_citation(_tool_call_chunk(finish_reason="tool_calls")))
def _citation_tool_call_then_text_chunks() -> tuple[ModelResponseStream, ...]:
return (_annotation_only_chunk(), _tool_call_chunk(), _chunk("Claim", finish_reason="tool_calls"))
@pytest.mark.parametrize("sync_mode", [True, False])
@pytest.mark.parametrize(
("stream_chunks", "expected_order"),
[
(_empty_text_citation_then_tool_call_chunks, ("message", "function_call")),
(_reasoning_empty_text_citation_then_tool_call_chunks, ("reasoning", "message", "function_call")),
(_empty_text_then_cited_tool_call_chunks, ("message", "function_call")),
(_citation_tool_call_then_text_chunks, ("message", "function_call")),
(_citation_then_tool_call_finishing_chunks, ("message", "function_call")),
(_reasoning_citation_tool_call_then_empty_finish_chunks, ("reasoning", "message", "function_call")),
],
ids=[
"empty_text_citation_then_tool_call",
"reasoning_empty_text_citation_then_tool_call",
"empty_text_then_cited_tool_call",
"citation_tool_call_then_text",
"citation_then_tool_call_without_text",
"reasoning_citation_tool_call_then_empty_finish",
],
)
@pytest.mark.asyncio
async def test_citation_before_a_tool_call_keeps_the_message_ahead_of_it(
stream_chunks: StreamChunks, expected_order: tuple[str, ...], sync_mode: bool
) -> None:
events: Final = await _collect_events(_build_iterator(stream_chunks()), sync_mode)
added: Final = [event for event in events if event.type == "response.output_item.added"]
output: Final = events[-1].response.output
message_index: Final = expected_order.index("message")
citations: Final = [event for event in events if event.type == "response.output_text.annotation.added"]
assert tuple(event.item.type for event in added) == expected_order
assert [event.output_index for event in added] == list(range(len(expected_order)))
assert tuple(item.type for item in output) == expected_order
assert [event.output_index for event in citations] == [message_index]
assert [event.annotation for event in citations] == [
annotation.model_dump() for annotation in output[message_index].content[0].annotations
]
def test_disconnect_after_a_held_citation_still_bills_it() -> None:
iterator: Final = _build_iterator((*_citation_between_reasoning_chunks(), _chunk("Claim", finish_reason="stop")))
reasoning_deltas: Final = (event for event in iterator if event.type == "response.reasoning_summary_text.delta")
assert len(list(islice(reasoning_deltas, 2))) == 2
incomplete: Final = iterator.partial_response_on_disconnect()
assert incomplete is not None
messages: Final = [item for item in incomplete.response.output if item.type == "message"]
annotations: Final = chain.from_iterable(message.content[0].annotations for message in messages)
assert [annotation.url for annotation in annotations] == ["https://example.com"]
def test_the_partial_response_billed_on_disconnect_carries_the_logging_objects_bill_as_hidden_response_cost() -> None:
from litellm.litellm_core_utils.hidden_params import get_hidden_params

84
uv.lock generated
View file

@ -25,11 +25,13 @@ members = [
]
constraints = [
{ name = "aiohttp", specifier = ">=3.14.2,<4.0" },
{ name = "dnspython", marker = "python_full_version >= '3.11'", specifier = ">=2.9.0" },
{ name = "httplib2", specifier = ">=0.32.0" },
{ name = "packaging", specifier = ">=24.0" },
{ name = "setuptools", specifier = ">=84.1.0" },
{ name = "soupsieve", specifier = ">=2.8.4" },
{ name = "tornado", specifier = ">=6.5.8" },
{ name = "tornado", marker = "python_full_version >= '3.11'", specifier = ">=6.6" },
]
overrides = [
{ name = "cryptography", specifier = ">=50.0.0,<51.0" },
@ -579,7 +581,8 @@ dependencies = [
{ name = "python-dotenv" },
{ name = "requests" },
{ name = "requests-toolbelt" },
{ name = "tornado" },
{ name = "tornado", version = "6.5.10", source = { registry = "https://pypi.org/simple" }, marker = "python_full_version < '3.11' or python_full_version >= '3.14'" },
{ name = "tornado", version = "6.6", source = { registry = "https://pypi.org/simple" }, marker = "python_full_version >= '3.11' and python_full_version < '3.14'" },
]
sdist = { url = "https://files.pythonhosted.org/packages/27/0e/c2e369ad173fb3d76448e46d10beb3dcc53388318933ddf8169a3f21a810/aurelio_sdk-0.0.19.tar.gz", hash = "sha256:14107e7440ff2efd0b4a08c52fb595e7680bd4bc973a0ddfb3b64157c6666b91", size = 15258, upload-time = "2025-03-24T14:37:32.203Z" }
wheels = [
@ -4702,7 +4705,7 @@ wheels = [
[[package]]
name = "langgraph"
version = "1.2.6"
version = "1.2.17"
source = { registry = "https://pypi.org/simple" }
dependencies = [
{ name = "langchain-core" },
@ -4713,22 +4716,22 @@ dependencies = [
{ name = "pydantic", version = "2.14.1", source = { registry = "https://pypi.org/simple" }, marker = "python_full_version >= '3.15'" },
{ name = "xxhash" },
]
sdist = { url = "https://files.pythonhosted.org/packages/02/7a/ea09b05bb0cbddfa43bd34fc581357e87fc3f21a751cc0d419688c3106da/langgraph-1.2.6.tar.gz", hash = "sha256:f9b45a34f13930c94d96cdb76277447ad2cc70ec2d18cd2764d7fdadb36cdc1b", size = 714400, upload-time = "2026-06-18T20:58:21.514Z" }
sdist = { url = "https://files.pythonhosted.org/packages/75/39/38b60bfc6fd064bc172693e19b3a6dccdaf9141bc9b984a9f83cdb06d014/langgraph-1.2.17.tar.gz", hash = "sha256:311516e313abd5424596d6c5250507c89be2e1973cbe813061696728803b0e10", size = 774002, upload-time = "2026-10-24T01:07:50.567Z" }
wheels = [
{ url = "https://files.pythonhosted.org/packages/89/32/772db1b00a9fe42f50320d1aa20caefb76e621eff1f7218b9918093d631d/langgraph-1.2.6-py3-none-any.whl", hash = "sha256:1cf94d3ca124f84f77ce408fa1b06c3dee680a8aafffe364a8fd5d7d03eb8695", size = 246132, upload-time = "2026-06-18T20:58:20.335Z" },
{ url = "https://files.pythonhosted.org/packages/00/d9/2e5b99319e2e8d39bc44d84836497c2e8f616554bb47843c8c819c8fed28/langgraph-1.2.17-py3-none-any.whl", hash = "sha256:e022a9fea64631fed946557a23f3fa5ac2ff38c72dcb999f426fc77d55ccde60", size = 260489, upload-time = "2026-10-24T01:07:49.155Z" },
]
[[package]]
name = "langgraph-checkpoint"
version = "4.1.1"
version = "4.3.0"
source = { registry = "https://pypi.org/simple" }
dependencies = [
{ name = "langchain-core" },
{ name = "ormsgpack" },
]
sdist = { url = "https://files.pythonhosted.org/packages/83/47/886af6f886f0bff2273164a45f008694e48a96ff3cd25ff0228f2aa9480e/langgraph_checkpoint-4.1.1.tar.gz", hash = "sha256:6c2bdb530c91f91d7d9c1bd100925d0fc4f498d418c17f3587d1526279482a25", size = 184020, upload-time = "2026-05-22T16:57:38.503Z" }
sdist = { url = "https://files.pythonhosted.org/packages/0f/69/31fdbdc65a85bbd6178afa193c772bb926620f47b4869638bc2bc80afaaa/langgraph_checkpoint-4.3.0.tar.gz", hash = "sha256:c75965d84cc2c1d549163e910a15bcb577758001b141619d05297c463280b018", size = 182652, upload-time = "2026-10-12T22:26:31.478Z" }
wheels = [
{ url = "https://files.pythonhosted.org/packages/bd/b4/71425e3e38be92611300b9cc5e46a5bf98ab23f5ea8a75b73d02a2f1413c/langgraph_checkpoint-4.1.1-py3-none-any.whl", hash = "sha256:25d29144b082827218e7bc3f1e9b0566a4bb007895cd6cc26f66a8428739f56e", size = 56212, upload-time = "2026-05-22T16:57:37.203Z" },
{ url = "https://files.pythonhosted.org/packages/1f/0c/84747e340bf4f29291c84cdd5733fc8d0a822f3d33bb24e664a18afa4a7c/langgraph_checkpoint-4.3.0-py3-none-any.whl", hash = "sha256:bedfafe2f997ded60e4fa593e79f56f436a6e45586392dc382aa810d0c751c64", size = 58063, upload-time = "2026-10-12T22:26:30.429Z" },
]
[[package]]
@ -4746,7 +4749,7 @@ wheels = [
[[package]]
name = "langgraph-sdk"
version = "0.4.4"
version = "0.4.9"
source = { registry = "https://pypi.org/simple" }
dependencies = [
{ name = "httpx" },
@ -4755,9 +4758,9 @@ dependencies = [
{ name = "orjson" },
{ name = "websockets" },
]
sdist = { url = "https://files.pythonhosted.org/packages/3f/ae/91446c1fffa04a2dc1f81afbfb5bfff3590452891a72bd9098a656ab2657/langgraph_sdk-0.4.4.tar.gz", hash = "sha256:4e651ffa09de695681579396375377bdde23bedbc8e35b070e615cbd5af7da8b", size = 345789, upload-time = "2026-08-27T21:25:22.593Z" }
sdist = { url = "https://files.pythonhosted.org/packages/4e/ef/e8f63b62dea6ecb75bef9d336f62e3e8f853b17a9943e4727c1bbeb09b11/langgraph_sdk-0.4.9.tar.gz", hash = "sha256:24d775ac1382b94c2595b8eb4eb4e0aa8dde95c905f18e017dc6574e1d79070d", size = 358868, upload-time = "2026-10-23T14:19:05.963Z" }
wheels = [
{ url = "https://files.pythonhosted.org/packages/43/d5/2ad7f3a835c6fcebf58b6669552536ecd52d2dfe4cf49c6c36648fd83572/langgraph_sdk-0.4.4-py3-none-any.whl", hash = "sha256:39afe416c91742925e6f8a93715f566d499b36e1b636b804a4ffe3190e4f4e64", size = 162270, upload-time = "2026-08-27T21:25:21.072Z" },
{ url = "https://files.pythonhosted.org/packages/57/7f/3102130eab099da9653ebe4a49123087a5889f4070eeee99b402255b1786/langgraph_sdk-0.4.9-py3-none-any.whl", hash = "sha256:c009243f4784758a07287fb2fd21fe47381c3d85f30479df2ae7f345c8703bb1", size = 167613, upload-time = "2026-10-23T14:19:04.82Z" },
]
[[package]]
@ -10228,7 +10231,8 @@ dependencies = [
{ name = "pyyaml" },
{ name = "regex" },
{ name = "tiktoken" },
{ name = "tornado" },
{ name = "tornado", version = "6.5.10", source = { registry = "https://pypi.org/simple" }, marker = "python_full_version < '3.11' or python_full_version >= '3.14'" },
{ name = "tornado", version = "6.6", source = { registry = "https://pypi.org/simple" }, marker = "python_full_version >= '3.11' and python_full_version < '3.14'" },
{ name = "urllib3" },
]
sdist = { url = "https://files.pythonhosted.org/packages/dc/a9/1a689e916e8b280f1fd8fb335cc059be626a22fe4533baa045d32fcd6de5/semantic_router-0.1.15.tar.gz", hash = "sha256:328256ddc3c2b713101ec69561d6585aecbf1198ea3461e1486289d8c3a35288", size = 95605, upload-time = "2026-05-23T12:58:15.444Z" }
@ -10920,6 +10924,9 @@ wheels = [
name = "tornado"
version = "6.5.10"
source = { registry = "https://pypi.org/simple" }
resolution-markers = [
"python_full_version < '3.11'",
]
sdist = { url = "https://files.pythonhosted.org/packages/06/61/53d562a57b28c08eda40b258c0f975e360541943ad7c7bef897a40caafda/tornado-6.5.10.tar.gz", hash = "sha256:a6b1ccd08c04b4a06fb5aeb381be99de5ad1e5375c1785e31d78c880feb57687", size = 537910, upload-time = "2026-09-15T13:47:48.73Z" }
wheels = [
{ url = "https://files.pythonhosted.org/packages/cd/5b/ff5fc58fa2427c30dea74c90053f4fc5eda1e7f3833ed3ecc7147fe2b311/tornado-6.5.10-cp39-abi3-macosx_10_9_universal2.whl", hash = "sha256:9261783640e23258694a9ff0795df430a5a7b0a651d3dd53dd0969ad6be16da7", size = 465883, upload-time = "2026-09-15T13:47:35.463Z" },
@ -10933,6 +10940,61 @@ wheels = [
{ url = "https://files.pythonhosted.org/packages/f3/1a/fd497f3a7f7b74bb04f4b94536b5c9f80742b5d50501fd27977652ddec16/tornado-6.5.10-cp39-abi3-win_arm64.whl", hash = "sha256:ce045d3c298fddd30e89a2777f97039d1b641eb9518ac7b26a4721903539c694", size = 467847, upload-time = "2026-09-15T13:47:47.283Z" },
]
[[package]]
name = "tornado"
version = "6.6"
source = { registry = "https://pypi.org/simple" }
resolution-markers = [
"python_full_version == '3.13.*'",
"python_full_version == '3.12.*'",
"python_full_version == '3.11.*'",
]
sdist = { url = "https://files.pythonhosted.org/packages/d0/02/5f549b1ad5da6267ed1e1fac2eef69257b1979cc812497002499b7895ff9/tornado-6.6.tar.gz", hash = "sha256:34c4aa916d7f2b2d5cadee2c6850f5a4a7f3d688cbf58399ea9bd3057af4abd5", size = 617998, upload-time = "2026-10-21T15:31:04.733Z" }
wheels = [
{ url = "https://files.pythonhosted.org/packages/df/f5/d3e6f76b945f9812842db742d0708a3b6287e850e81be3ca10e290c103c6/tornado-6.6-cp311-abi3-macosx_10_9_universal2.whl", hash = "sha256:01abf15429d64f04ccfafc238b191cc2c177108b7a331f4461707ad77198c8c2", size = 528326, upload-time = "2026-10-21T15:30:05.528Z" },
{ url = "https://files.pythonhosted.org/packages/30/69/b20cf3a8bf641ec792a5a6050aa153e8f120604f9c23471142de535113e5/tornado-6.6-cp311-abi3-manylinux1_x86_64.manylinux_2_28_x86_64.manylinux_2_5_x86_64.whl", hash = "sha256:11bb97c00981eb8254e55bd6a5ad960c983ae343fef8ae3ae5584cd1fb70236a", size = 529544, upload-time = "2026-10-21T15:30:07.126Z" },
{ url = "https://files.pythonhosted.org/packages/0f/f4/d9d8d3ac345aa9df3a132ad03bd443160c2ab24ad4003b4e8d6d5af06371/tornado-6.6-cp311-abi3-manylinux2014_aarch64.manylinux_2_17_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:4e152a3b2d3eae113b836d979823f7bb2320f7e334a84449729964921f374a4c", size = 530523, upload-time = "2026-10-21T15:30:08.478Z" },
{ url = "https://files.pythonhosted.org/packages/ad/82/1296ade1c14dee5a4906e5ca092997622ece7c4dc768c68ed9caaf895d1b/tornado-6.6-cp311-abi3-manylinux2014_ppc64le.manylinux_2_17_ppc64le.manylinux_2_28_ppc64le.whl", hash = "sha256:e912d4c4babd04db6f34b25b748174a4a176debe86c8610beedf1815f9af7fdd", size = 530739, upload-time = "2026-10-21T15:30:09.98Z" },
{ url = "https://files.pythonhosted.org/packages/66/30/6ac1eab66c81af46dafd93c9b24bd311f82e199d9d7b8ce2a64a91b6a200/tornado-6.6-cp311-abi3-manylinux2014_s390x.manylinux_2_17_s390x.manylinux_2_28_s390x.whl", hash = "sha256:33aabe8a03d426cef4e0fe5669f0643cadd983b54957f1a7ac2789ec326e9644", size = 529738, upload-time = "2026-10-21T15:30:11.498Z" },
{ url = "https://files.pythonhosted.org/packages/24/b4/6c1f6aa1fd148e4d0f500a10cd0c9949e7bf385e161ff497c81de5b3ae4f/tornado-6.6-cp311-abi3-manylinux_2_31_riscv64.manylinux_2_39_riscv64.whl", hash = "sha256:9aaf0e1c2fb9f046f30e2036b8d2060666399e8464d0a72769e43c754089567b", size = 527908, upload-time = "2026-10-21T15:30:13.298Z" },
{ url = "https://files.pythonhosted.org/packages/54/5b/67d3d452b1179bc9e265cebd381a56ab4cf2d7b2a54607aedd8d6e1ebc2e/tornado-6.6-cp311-abi3-musllinux_1_2_aarch64.whl", hash = "sha256:61fc5513cea97720ab9bf54b04827ac0f4500c63f0b072b963d6145c798b4661", size = 530367, upload-time = "2026-10-21T15:30:15.183Z" },
{ url = "https://files.pythonhosted.org/packages/74/14/4e8e9a14f77bcafb2408634b2d4c8c676ad7a43bf8dac4ee2d48370698f9/tornado-6.6-cp311-abi3-musllinux_1_2_ppc64le.whl", hash = "sha256:31e1a0d5cd704d2fe8a008c1cc27a0e2b67269ca263630a27985dddc160b5537", size = 530456, upload-time = "2026-10-21T15:30:16.741Z" },
{ url = "https://files.pythonhosted.org/packages/56/12/e3ddef99a2089ccfc86f32bd3f39caaad87bee4868be1bacbf213a0b07a2/tornado-6.6-cp311-abi3-musllinux_1_2_riscv64.whl", hash = "sha256:74416e41e9bdf70fbc2fb242f6fbbc7a58ec11e59d39f15fc5a421ecb6845118", size = 529610, upload-time = "2026-10-21T15:30:18.031Z" },
{ url = "https://files.pythonhosted.org/packages/35/3d/62a692343d46b37e852e5078e0f6513534d925f70a04580072e3c127b129/tornado-6.6-cp311-abi3-musllinux_1_2_s390x.whl", hash = "sha256:8cdd69104ae9d705b3f04d2563aa96e357043e267a455802daae9a1e633a37c8", size = 529803, upload-time = "2026-10-21T15:30:19.326Z" },
{ url = "https://files.pythonhosted.org/packages/49/04/6c59a9413adfe4810f648a4dd9fdf476f13b0495f9cc25ef856b71f9ce1e/tornado-6.6-cp311-abi3-musllinux_1_2_x86_64.whl", hash = "sha256:fc7c8a3263501fbe97518cd699101bc3c5f499a7dd4f75fc6cbb3bd9ee1db1f1", size = 529751, upload-time = "2026-10-21T15:30:20.826Z" },
{ url = "https://files.pythonhosted.org/packages/31/36/a1b3b84cd347ae43f982d3bf95628de6c6365668dba2543af47c6a1c888e/tornado-6.6-cp311-abi3-win32.whl", hash = "sha256:b54ca69ff93d0374efb083a12a1e9ec2dd2036e8496d197e0e8ef77b8d9135c3", size = 530643, upload-time = "2026-10-21T15:30:22.428Z" },
{ url = "https://files.pythonhosted.org/packages/b7/50/dc813fe35a672e14ee8bed8b5c634eeb343a05c98eb455318a169de24567/tornado-6.6-cp311-abi3-win_amd64.whl", hash = "sha256:48f26ed9d7beb64e5f06adc864524e581ab6d62ba8ff46eaaa1ac2d0e6d753fd", size = 530886, upload-time = "2026-10-21T15:30:23.975Z" },
{ url = "https://files.pythonhosted.org/packages/0d/71/62a9b9e980026d38874740d0a7c162f53eb0563f6a441e3aff84c48db7e8/tornado-6.6-cp311-abi3-win_arm64.whl", hash = "sha256:29114ffd4e4de869d390393df9e9a0f191065aae1a96f38a58d35e464c04a867", size = 530899, upload-time = "2026-10-21T15:30:25.264Z" },
{ url = "https://files.pythonhosted.org/packages/ac/4c/633e3568eb574df555d6ddd2dac44b7793a5c26cd63b444e653fcfb1d6ec/tornado-6.6-cp314-cp314t-macosx_10_15_universal2.whl", hash = "sha256:63fc8d243b52fd8dd9ed4ad02d0df08f2024f9aea632a433a8a6c91075076d44", size = 528432, upload-time = "2026-10-21T15:30:26.578Z" },
{ url = "https://files.pythonhosted.org/packages/37/eb/b525d1870ad28a56f47b664dd8fd3fab4f7ddd237c87c14e97248bf4762f/tornado-6.6-cp314-cp314t-manylinux1_x86_64.manylinux_2_28_x86_64.manylinux_2_5_x86_64.whl", hash = "sha256:8d4b85cac795bfc6a7f20ba81871abb8c2ae8fba019061b8a9ae1c2f97e29637", size = 532586, upload-time = "2026-10-21T15:30:28.027Z" },
{ url = "https://files.pythonhosted.org/packages/b7/35/909a33c253352346ca081a27ca66b45c1398468adf784b767597ca773c28/tornado-6.6-cp314-cp314t-manylinux2014_aarch64.manylinux_2_17_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:e035c065edadfa696529d05eebcbeb25154c129b5aee6dbfb697a6abf5d4f697", size = 533641, upload-time = "2026-10-21T15:30:29.318Z" },
{ url = "https://files.pythonhosted.org/packages/6f/64/100202666d6e3d96288ab184827762415456c139798a15c962547f9fabf9/tornado-6.6-cp314-cp314t-manylinux2014_ppc64le.manylinux_2_17_ppc64le.manylinux_2_28_ppc64le.whl", hash = "sha256:5b94e589ab855090c448b5c0937b3a75f1c1a604b7b36207e8236d209f720b9c", size = 533869, upload-time = "2026-10-21T15:30:30.882Z" },
{ url = "https://files.pythonhosted.org/packages/b1/e0/af3ed62bc15830d60328d5b4a1a9076b35ea0ba142cb17b1c1772cb15c59/tornado-6.6-cp314-cp314t-manylinux2014_s390x.manylinux_2_17_s390x.manylinux_2_28_s390x.whl", hash = "sha256:1a394c8c9db9137f2a2b42869d9f7b91398fcf19da120d0dde4c4546172710f7", size = 532767, upload-time = "2026-10-21T15:30:32.273Z" },
{ url = "https://files.pythonhosted.org/packages/24/d7/1e2e1a6932eb62d63324b64dba2f62e6eaf79ad883dd209352c297b7848c/tornado-6.6-cp314-cp314t-manylinux_2_31_riscv64.manylinux_2_39_riscv64.whl", hash = "sha256:0215d1df6e4b5e736ddad5adf0af4138cf7437fdd39c77c5e69768e37850a632", size = 530956, upload-time = "2026-10-21T15:30:33.746Z" },
{ url = "https://files.pythonhosted.org/packages/70/65/3b9a0e53d7a4c39f2ebe95486540fa8eeb67bfb6e8c437d603c00d35b128/tornado-6.6-cp314-cp314t-musllinux_1_2_aarch64.whl", hash = "sha256:4d711790e0a9576ad924d942c4607e4e401d697588ea6f30f6662824008a7add", size = 533372, upload-time = "2026-10-21T15:30:35.085Z" },
{ url = "https://files.pythonhosted.org/packages/44/48/90979cbfaa3ff941324a2605eb794af0f087c670cd99f18a69ab40498073/tornado-6.6-cp314-cp314t-musllinux_1_2_ppc64le.whl", hash = "sha256:034a66bae0f55421d007803e2385d2a59707d8829b2f026373d42f08256e9922", size = 533425, upload-time = "2026-10-21T15:30:36.425Z" },
{ url = "https://files.pythonhosted.org/packages/bb/c4/ce743a86d91f85c255b77a4282eb5b2716b61e74430eb9d1df8578a539d2/tornado-6.6-cp314-cp314t-musllinux_1_2_riscv64.whl", hash = "sha256:5cb38483b077129056a94dd415a71d29ab4e42b46da894fee8b4657003f10b21", size = 532545, upload-time = "2026-10-21T15:30:37.789Z" },
{ url = "https://files.pythonhosted.org/packages/02/5e/a82131e35778540fc19139229d02aa8cf2453a8022e202be50c88f0f8ac3/tornado-6.6-cp314-cp314t-musllinux_1_2_s390x.whl", hash = "sha256:46bbbd107efa37777ac4b83601866d19b4dbd5b770a7a9123582633d95959ad3", size = 532700, upload-time = "2026-10-21T15:30:39.336Z" },
{ url = "https://files.pythonhosted.org/packages/3a/09/6dac8c2c677c65d60b3d7dae94176dbfe392f314e7c8e2d709623edd6a26/tornado-6.6-cp314-cp314t-musllinux_1_2_x86_64.whl", hash = "sha256:b85e9d0263cdbc027b95947c8b873a7bc7fd6a08d16bfb4f5047350b0808e464", size = 532712, upload-time = "2026-10-21T15:30:40.698Z" },
{ url = "https://files.pythonhosted.org/packages/b2/ac/71e5776cbcdc3264e81aa6db28c69717ba50607a9eb8d5b766f11a4b020e/tornado-6.6-cp314-cp314t-win32.whl", hash = "sha256:4276eef0880e5fe4488e6799c2842f59af3b06637e9313ab4fb506f9c9c2d6cb", size = 530491, upload-time = "2026-10-21T15:30:42.115Z" },
{ url = "https://files.pythonhosted.org/packages/d8/0f/0b6a6908a5b5aa8320009dfdb84434b1cf160b577a00456e59ab9e6ea598/tornado-6.6-cp314-cp314t-win_amd64.whl", hash = "sha256:10d4af98c6df09f8429b1340ccaea3a72887b86f695dfb8056bda517c8115a00", size = 530780, upload-time = "2026-10-21T15:30:43.414Z" },
{ url = "https://files.pythonhosted.org/packages/3c/39/c152e1cc5371f75f5f9f0a4f44bf2dc9396be9407d384de7428446fb4651/tornado-6.6-cp314-cp314t-win_arm64.whl", hash = "sha256:b1f0ef20a5278854f5fdac31637bafadb055405df28b7fc945e54add12b172e3", size = 530781, upload-time = "2026-10-21T15:30:44.741Z" },
{ url = "https://files.pythonhosted.org/packages/0c/c5/f96c7d7a8895745bb37dfb6e6107e1ee40ad02a4b162d7d6c143644455ab/tornado-6.6-cp315-cp315t-macosx_10_15_universal2.whl", hash = "sha256:8f8e382da30fe042f27f27a871e21aefc2613870e5a41648de18446c72d6ea59", size = 528446, upload-time = "2026-10-21T15:30:46.112Z" },
{ url = "https://files.pythonhosted.org/packages/c0/dd/ab8f6500d274b9c55b025798eb54e92591bcff27515799bec443bd54e35a/tornado-6.6-cp315-cp315t-manylinux1_x86_64.manylinux_2_28_x86_64.manylinux_2_5_x86_64.whl", hash = "sha256:8c7a8a328396c2195454a2ecd9499512441bbce99505544a717e6a87d16ff6f9", size = 532774, upload-time = "2026-10-21T15:30:47.393Z" },
{ url = "https://files.pythonhosted.org/packages/87/23/b78e7d1bf329160b8157284005395421ea941864c16c7f04c3e48c98bcf9/tornado-6.6-cp315-cp315t-manylinux2014_aarch64.manylinux_2_17_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:ca025bb0da81c294aef26f27427316df1525220820315f4bf32dc1e84a4d2232", size = 533872, upload-time = "2026-10-21T15:30:48.884Z" },
{ url = "https://files.pythonhosted.org/packages/e0/4f/ab66b614ad0bc9a219484be6bc2c002c8de9916a7b6c7b27fe6273838e5c/tornado-6.6-cp315-cp315t-manylinux2014_ppc64le.manylinux_2_17_ppc64le.manylinux_2_28_ppc64le.whl", hash = "sha256:59c22ebc905ac2a267f8d9d8a878f86afbcbbde329d75c485f1422fc45a81b83", size = 534027, upload-time = "2026-10-21T15:30:50.157Z" },
{ url = "https://files.pythonhosted.org/packages/da/63/501fa15a36bace2f4b9158228cf922d606403261bf71c960c433ec3ca0a4/tornado-6.6-cp315-cp315t-manylinux2014_s390x.manylinux_2_17_s390x.manylinux_2_28_s390x.whl", hash = "sha256:f89802bc0bcd15f988d9325b4e6d2171bb04ea68d188732556335e10d3fd00bd", size = 532936, upload-time = "2026-10-21T15:30:51.45Z" },
{ url = "https://files.pythonhosted.org/packages/7f/b7/82d643f635f0300bda98613ca2660b4f3134b6fb594c6fddfa4de706806c/tornado-6.6-cp315-cp315t-manylinux_2_31_riscv64.manylinux_2_39_riscv64.whl", hash = "sha256:70043256d5958f9c655066153d23d60ea3e29bf9854f7fb75850e223c477758a", size = 531254, upload-time = "2026-10-21T15:30:52.865Z" },
{ url = "https://files.pythonhosted.org/packages/d4/53/92c40891a94750bc09257310af964bf25c59e3531d6199736c318baa864f/tornado-6.6-cp315-cp315t-musllinux_1_2_aarch64.whl", hash = "sha256:82b49495e3849d644b841ef33b35491160055997deaa93720ff647ef019421db", size = 533566, upload-time = "2026-10-21T15:30:54.15Z" },
{ url = "https://files.pythonhosted.org/packages/7e/88/b82f5c5efde731b88d839ad99cb6ace2a521f03bb6d7346d2de089c76da8/tornado-6.6-cp315-cp315t-musllinux_1_2_ppc64le.whl", hash = "sha256:07b2bcaa4bf447bdbfb664e4c28344226540db3f0ebccd74886528338a240765", size = 533567, upload-time = "2026-10-21T15:30:55.451Z" },
{ url = "https://files.pythonhosted.org/packages/02/2c/e9a7f5d0fca0bdbcd48b2d9c7884dd0b32669c9945d98bc6d303a80ecfea/tornado-6.6-cp315-cp315t-musllinux_1_2_riscv64.whl", hash = "sha256:c36b921374c302e54909bae17d9f354061d80e7ae7d8c94f73abbcc217866460", size = 532849, upload-time = "2026-10-21T15:30:56.729Z" },
{ url = "https://files.pythonhosted.org/packages/d0/f5/4ade737340ca6ce9c677ea0c2843591c8ad065d07f0913c28d922c7e078e/tornado-6.6-cp315-cp315t-musllinux_1_2_s390x.whl", hash = "sha256:7412d9e37aa9f7048e4b591f520a95f7a097278ffab15f1ddf3f5c2320cf6648", size = 532873, upload-time = "2026-10-21T15:30:58.067Z" },
{ url = "https://files.pythonhosted.org/packages/8e/04/6acdc937d0bb6189ddb34ceded517d7bd1c7b3bded5d4e1b3442160e6b6f/tornado-6.6-cp315-cp315t-musllinux_1_2_x86_64.whl", hash = "sha256:15bc95112b2d24b0bfaee9a004e3f974da929e33b4244a7f695a22f45b08a5e5", size = 532883, upload-time = "2026-10-21T15:30:59.395Z" },
{ url = "https://files.pythonhosted.org/packages/ab/27/0a766fac6b1cc629d96346329a0485287fd2857c51a887a3a154f2a69938/tornado-6.6-cp315-cp315t-win32.whl", hash = "sha256:f661c21117b875d612371f3eb243d0c69f56d9bdd6aee38b7c24453419dfe554", size = 530488, upload-time = "2026-10-21T15:31:00.77Z" },
{ url = "https://files.pythonhosted.org/packages/5c/80/3e4142570883da27b3f60ae82c638ac96e60997eac684740613557519c96/tornado-6.6-cp315-cp315t-win_amd64.whl", hash = "sha256:ee8853ab35da510dc952879ebd9995c82bc10ffb7925fd87930dc0d6916f4b39", size = 530777, upload-time = "2026-10-21T15:31:02.087Z" },
{ url = "https://files.pythonhosted.org/packages/17/57/4a85c617b6349962320b4afbe1386d5b05fd57218ee8ad6e82e0110ec9ee/tornado-6.6-cp315-cp315t-win_arm64.whl", hash = "sha256:7de62e2dfdf559fd7154d2b9aa58ff3d9cff8304ae0519341feb082618bc2a33", size = 530780, upload-time = "2026-10-21T15:31:03.425Z" },
]
[[package]]
name = "tqdm"
version = "4.67.3"