chore: ran linting

surfsense_backend/app/tasks/chat/llm_history_normalizer.py

@@ -86,4 +86,3 @@ def user_content_to_llm_content(
     if allow_images and any(part.get("type") == "image_url" for part in parts):
         return parts
     return "\n".join(text_chunks)
-

surfsense_backend/app/tasks/chat/message_parts_normalizer.py

@@ -31,7 +31,9 @@ def _text_from_content(content: Any) -> str:
     return "".join(text_parts)
 
 
-def normalize_ai_message_to_parts(message: AIMessage | Any | None) -> list[dict[str, Any]]:
+def normalize_ai_message_to_parts(
+    message: AIMessage | Any | None,
+) -> list[dict[str, Any]]:
     """Return user-visible assistant-ui parts for a final AI message.
 
     We intentionally do not backfill provider ``thinking`` /
@@ -60,7 +62,9 @@ def last_ai_message(messages: Iterable[Any] | None) -> AIMessage | Any | None:
     return None
 
 
-def final_assistant_parts_from_messages(messages: Iterable[Any] | None) -> list[dict[str, Any]]:
+def final_assistant_parts_from_messages(
+    messages: Iterable[Any] | None,
+) -> list[dict[str, Any]]:
     return normalize_ai_message_to_parts(last_ai_message(messages))
 
 
@@ -83,4 +87,3 @@ def merge_streamed_and_final_parts(
     if not has_non_empty_text_part(final_parts):
         return streamed_parts
     return [*streamed_parts, *final_parts]
-

surfsense_backend/app/tasks/chat/streaming/agent/event_loop.py

@@ -16,6 +16,9 @@ from app.agents.chat.multi_agent_chat.main_agent.middleware.kb_persistence impor
 )
 from app.agents.chat.multi_agent_chat.shared.filesystem_selection import FilesystemMode
 from app.services.new_streaming_service import VercelStreamingService
+from app.tasks.chat.message_parts_normalizer import (
+    final_assistant_parts_from_messages,
+)
 from app.tasks.chat.streaming.contract.file_contract import (
     contract_enforcement_active,
     evaluate_file_contract_outcome,
@@ -25,9 +28,6 @@ from app.tasks.chat.streaming.graph_stream.event_stream import stream_output
 from app.tasks.chat.streaming.helpers.interrupt_inspector import (
     all_interrupt_values,
 )
-from app.tasks.chat.message_parts_normalizer import (
-    final_assistant_parts_from_messages,
-)
 from app.tasks.chat.streaming.shared.stream_result import StreamResult
 from app.tasks.chat.streaming.shared.utils import safe_float
 from app.utils.perf import get_perf_logger

surfsense_backend/app/tasks/chat/streaming/errors/classifier.py

@@ -146,9 +146,10 @@ def _provider_error_extra(adapted: Any) -> dict[str, Any] | None:
 
 def _classify_provider_exception(
     exc: Exception,
-) -> tuple[
-    str, str, Literal["info", "warn", "error"], bool, str, dict[str, Any] | None
-] | None:
+) -> (
+    tuple[str, str, Literal["info", "warn", "error"], bool, str, dict[str, Any] | None]
+    | None
+):
     adapted = adapt_llm_exception(exc)
 
     if adapted.category is LLMErrorCategory.RATE_LIMITED:

surfsense_backend/app/tasks/chat/streaming/flows/shared/llm_bundle.py

@@ -55,8 +55,12 @@ def _agent_config_from_resolved(
     )
 
 
-async def _load_search_space(session: AsyncSession, search_space_id: int) -> SearchSpace | None:
-    result = await session.execute(select(SearchSpace).where(SearchSpace.id == search_space_id))
+async def _load_search_space(
+    session: AsyncSession, search_space_id: int
+) -> SearchSpace | None:
+    result = await session.execute(
+        select(SearchSpace).where(SearchSpace.id == search_space_id)
+    )
     return result.scalars().first()
 
 
@@ -131,7 +135,9 @@ async def load_llm_bundle(
             None,
         )
 
-    global_model = next((m for m in config.GLOBAL_MODELS if m.get("id") == config_id), None)
+    global_model = next(
+        (m for m in config.GLOBAL_MODELS if m.get("id") == config_id), None
+    )
     if not global_model or not has_capability(global_model, "chat"):
         return None, None, f"Failed to load global chat model with id {config_id}"
     global_connection = next(
@@ -144,7 +150,9 @@ async def load_llm_bundle(
     )
     if not global_connection:
         return None, None, f"Failed to load global connection for model {config_id}"
-    model_string, litellm_kwargs = to_litellm(global_connection, global_model["model_id"])
+    model_string, litellm_kwargs = to_litellm(
+        global_connection, global_model["model_id"]
+    )
     display_name = global_model.get("display_name") or global_model.get("model_id")
     provider = global_connection.get("provider") or ""
     register_model_usage_metadata(