unique-toolkit 2026.30.0.dev11__tar.gz → 2026.30.0.dev13__tar.gz

{unique_toolkit-2026.30.0.dev11 → unique_toolkit-2026.30.0.dev13}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: unique-toolkit
-Version: 2026.30.0.dev11
+Version: 2026.30.0.dev13
 Summary: 
 Author: Cedric Klinkert, Martin Fadler, Sadique Sheik, Fabian Schläpfer, Pascal Hauri, Ejiroghene Oki-Peter, Rami Azouz, Ricardo Visini, Sanzio Monti
 Author-email: Cedric Klinkert <cedric.klinkert@unique.ch>, Martin Fadler <martin.fadler@unique.ch>, Sadique Sheik <sadique@unique.ch>, Fabian Schläpfer <fabian@unique.ch>, Pascal Hauri <pascal@unique.ch>, Ejiroghene Oki-Peter <oki-peter.ejiroghene@unique.ai>, Rami Azouz <rami.ext@unique.ch>, Ricardo Visini <ricardo.visini@unique.ch>, Sanzio Monti <sanzio.monti@unique.ai>

{unique_toolkit-2026.30.0.dev11 → unique_toolkit-2026.30.0.dev13}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "unique_toolkit"
-version = "2026.30.0.dev11"
+version = "2026.30.0.dev13"
 description = ""
 readme = "README.md"
 requires-python = ">=3.12"

{unique_toolkit-2026.30.0.dev11 → unique_toolkit-2026.30.0.dev13}/unique_toolkit/agentic/history_manager/history_construction_with_contents.py

@@ -6,7 +6,7 @@ import mimetypes
 from datetime import datetime
 from enum import StrEnum
 from itertools import groupby
-from typing import TYPE_CHECKING, Iterator
+from typing import TYPE_CHECKING, Callable, Iterator
 
 if TYPE_CHECKING:
     from unique_toolkit.content.schemas import ContentChunk
@@ -200,9 +200,7 @@ async def download_encoded_images_async(
     return base64_encoded_images
 
 
-class FileContentSerialization(StrEnum):
-    NONE = "none"
-    FILE_NAME = "file_name"
+FileContentSerializer = Callable[[Content], str | None]
 
 
 class ImageContentInclusion(StrEnum):
@@ -210,23 +208,17 @@ class ImageContentInclusion(StrEnum):
     ALL = "all"
 
 
-def file_content_serialization(
+def _serialize_file_contents(
+    text: str,
     file_contents: list[Content],
-    file_content_serialization: FileContentSerialization,
+    file_content_serializer: FileContentSerializer | None,
 ) -> str:
-    match file_content_serialization:
-        case FileContentSerialization.NONE:
-            return ""
-        case FileContentSerialization.FILE_NAME:
-            file_names = [
-                f"- Uploaded file: {f.key} at {f.created_at}" for f in file_contents
-            ]
-            return "\n".join(
-                [
-                    "Files Uploaded to Chat can be accessed by internal search tool if available:\n",
-                ]
-                + file_names,
-            )
+    serialized_files: list[str] = []
+    if file_content_serializer is not None:
+        for file_content in file_contents:
+            if serialized_file := file_content_serializer(file_content):
+                serialized_files.append(serialized_file)
+    return (text + "\n\n" + "\n".join(serialized_files)).strip()
 
 
 def _append_element_to_builder(
@@ -234,9 +226,9 @@ def _append_element_to_builder(
     c: ChatMessageWithContents,
     text: str,
     include_images: ImageContentInclusion,
-    file_content_serialization_type: FileContentSerialization,
     content_service: ContentService,
     chat_id: str,
+    file_content_serializer: FileContentSerializer | None = None,
     selected_content_ids: set[str] | None = None,
 ) -> None:
     if len(c.contents) > 0:
@@ -249,15 +241,12 @@ def _append_element_to_builder(
             image_contents = [
                 co for co in image_contents if co.id in selected_content_ids
             ]
-        content = (
-            text
-            + "\n\n"
-            + file_content_serialization(
-                file_contents,
-                file_content_serialization_type,
-            )
-        ).strip()
-        if include_images and len(image_contents) > 0:
+        content = _serialize_file_contents(
+            text,
+            file_contents,
+            file_content_serializer,
+        )
+        if include_images and image_contents:
             builder.image_message_append(
                 content=content,
                 images=download_encoded_images(
@@ -285,9 +274,9 @@ async def _append_element_to_builder_async(
     c: ChatMessageWithContents,
     text: str,
     include_images: ImageContentInclusion,
-    file_content_serialization_type: FileContentSerialization,
     content_service: ContentService,
     chat_id: str,
+    file_content_serializer: FileContentSerializer | None = None,
     selected_content_ids: set[str] | None = None,
 ) -> None:
     if len(c.contents) > 0:
@@ -300,15 +289,12 @@ async def _append_element_to_builder_async(
             image_contents = [
                 co for co in image_contents if co.id in selected_content_ids
             ]
-        content = (
-            text
-            + "\n\n"
-            + file_content_serialization(
-                file_contents,
-                file_content_serialization_type,
-            )
-        ).strip()
-        if include_images and len(image_contents) > 0:
+        content = _serialize_file_contents(
+            text,
+            file_contents,
+            file_content_serializer,
+        )
+        if include_images and image_contents:
             builder.image_message_append(
                 content=content,
                 images=await download_encoded_images_async(
@@ -336,7 +322,7 @@ def get_full_history_with_contents(
     chat_service: ChatService,
     content_service: ContentService,
     include_images: ImageContentInclusion = ImageContentInclusion.ALL,
-    file_content_serialization_type: FileContentSerialization = FileContentSerialization.FILE_NAME,
+    file_content_serializer: FileContentSerializer | None = None,
     selected_content_ids: set[str] | None = None,
 ) -> LanguageModelMessages:
     grouped_elements = get_chat_history_with_contents(
@@ -361,7 +347,7 @@ def get_full_history_with_contents(
             c=c,
             text=text,
             include_images=include_images,
-            file_content_serialization_type=file_content_serialization_type,
+            file_content_serializer=file_content_serializer,
             content_service=content_service,
             chat_id=chat_id,
             selected_content_ids=selected_content_ids,
@@ -376,7 +362,7 @@ async def get_full_history_with_contents_async(
     chat_service: ChatService,
     content_service: ContentService,
     include_images: ImageContentInclusion = ImageContentInclusion.ALL,
-    file_content_serialization_type: FileContentSerialization = FileContentSerialization.FILE_NAME,
+    file_content_serializer: FileContentSerializer | None = None,
     selected_content_ids: set[str] | None = None,
 ) -> LanguageModelMessages:
     grouped_elements = await get_chat_history_with_contents_async(
@@ -400,7 +386,7 @@ async def get_full_history_with_contents_async(
             c=c,
             text=text,
             include_images=include_images,
-            file_content_serialization_type=file_content_serialization_type,
+            file_content_serializer=file_content_serializer,
             content_service=content_service,
             chat_id=chat_id,
             selected_content_ids=selected_content_ids,
@@ -415,7 +401,7 @@ def get_full_history_with_contents_and_tool_calls(
     chat_service: ChatService,
     content_service: ContentService,
     include_images: ImageContentInclusion = ImageContentInclusion.ALL,
-    file_content_serialization_type: FileContentSerialization = FileContentSerialization.FILE_NAME,
+    file_content_serializer: FileContentSerializer | None = None,
     selected_content_ids: set[str] | None = None,
 ) -> tuple[LanguageModelMessages, int, dict[int, "ContentChunk"]]:
     """Build the full LLM message history, including persisted tool call rounds.
@@ -544,7 +530,7 @@ def get_full_history_with_contents_and_tool_calls(
             c=c,
             text=text,
             include_images=include_images,
-            file_content_serialization_type=file_content_serialization_type,
+            file_content_serializer=file_content_serializer,
             content_service=content_service,
             chat_id=chat_id,
             selected_content_ids=selected_content_ids,
@@ -559,7 +545,7 @@ async def get_full_history_with_contents_and_tool_calls_async(
     chat_service: ChatService,
     content_service: ContentService,
     include_images: ImageContentInclusion = ImageContentInclusion.ALL,
-    file_content_serialization_type: FileContentSerialization = FileContentSerialization.FILE_NAME,
+    file_content_serializer: FileContentSerializer | None = None,
     selected_content_ids: set[str] | None = None,
 ) -> tuple[LanguageModelMessages, int, dict[int, "ContentChunk"]]:
     """Async version of get_full_history_with_contents_and_tool_calls."""
@@ -662,7 +648,7 @@ async def get_full_history_with_contents_and_tool_calls_async(
             c=c,
             text=text,
             include_images=include_images,
-            file_content_serialization_type=file_content_serialization_type,
+            file_content_serializer=file_content_serializer,
             content_service=content_service,
             chat_id=chat_id,
             selected_content_ids=selected_content_ids,

{unique_toolkit-2026.30.0.dev11 → unique_toolkit-2026.30.0.dev13}/unique_toolkit/agentic/history_manager/history_manager.py

@@ -10,6 +10,9 @@ from unique_toolkit._common.feature_flags.schema import (
 )
 from unique_toolkit._common.pydantic_helpers import DeactivatedNone
 from unique_toolkit._common.validators import LMI
+from unique_toolkit.agentic.history_manager.history_construction_with_contents import (
+    FileContentSerializer,
+)
 from unique_toolkit.agentic.history_manager.loop_token_reducer import LoopTokenReducer
 from unique_toolkit.agentic.history_manager.utils import (
     serialize_tool_content_json,
@@ -115,6 +118,7 @@ class HistoryManager:
         config: HistoryManagerConfig,
         language_model: LMI,
         reference_manager: ReferenceManager,
+        file_content_serializer: FileContentSerializer | None = None,
     ):
         self._config = config
         self._logger = logger
@@ -123,10 +127,10 @@ class HistoryManager:
             logger=self._logger,
             event=event,
             max_history_tokens=self._config.max_history_tokens,
-            has_uploaded_content_config=bool(self._config.uploaded_content_config),
             language_model=self._language_model,
             reference_manager=reference_manager,
             enable_tool_call_persistence=self._config.enable_tool_call_persistence,
+            file_content_serializer=file_content_serializer,
         )
         self._reference_manager = reference_manager
         self._tool_call_result_history: list[ToolCallResponse] = []

{unique_toolkit-2026.30.0.dev11 → unique_toolkit-2026.30.0.dev13}/unique_toolkit/agentic/history_manager/loop_token_reducer.py

@@ -9,7 +9,7 @@ from unique_toolkit._common.token.token_counting import (
 )
 from unique_toolkit._common.validators import LMI
 from unique_toolkit.agentic.history_manager.history_construction_with_contents import (
-    FileContentSerialization,
+    FileContentSerializer,
     get_full_history_with_contents_and_tool_calls_async,
     get_full_history_with_contents_async,
 )
@@ -54,13 +54,13 @@ class LoopTokenReducer:
         logger: Logger,
         event: ChatEvent,
         max_history_tokens: int,
-        has_uploaded_content_config: bool,
         reference_manager: ReferenceManager,
         language_model: LMI,
         enable_tool_call_persistence: bool = False,
+        file_content_serializer: FileContentSerializer | None = None,
     ):
         self._max_history_tokens = max_history_tokens
-        self._has_uploaded_content_config = has_uploaded_content_config
+        self._file_content_serializer = file_content_serializer
         self._logger = logger
         self._reference_manager = reference_manager
         self._language_model = language_model
@@ -309,11 +309,6 @@ class LoopTokenReducer:
         Returns:
             list[LanguageModelMessage]: The history
         """
-        file_content_serialization_type = (
-            FileContentSerialization.NONE
-            if self._has_uploaded_content_config
-            else FileContentSerialization.FILE_NAME
-        )
         if self._enable_tool_call_persistence:
             (
                 full_history,
@@ -324,7 +319,7 @@ class LoopTokenReducer:
                 chat_id=self._chat_id,
                 chat_service=self._chat_service,
                 content_service=self._content_service,
-                file_content_serialization_type=file_content_serialization_type,
+                file_content_serializer=self._file_content_serializer,
                 selected_content_ids=self._selected_content_ids,
             )
             self._max_db_source_number = max_src
@@ -335,7 +330,7 @@ class LoopTokenReducer:
                 chat_id=self._chat_id,
                 chat_service=self._chat_service,
                 content_service=self._content_service,
-                file_content_serialization_type=file_content_serialization_type,
+                file_content_serializer=self._file_content_serializer,
                 selected_content_ids=self._selected_content_ids,
             )
 

{unique_toolkit-2026.30.0.dev11 → unique_toolkit-2026.30.0.dev13}/unique_toolkit/language_model/infos.py

@@ -61,6 +61,9 @@ class LanguageModelName(StrEnum):
     AZURE_GPT_54_PRO_2026_0305 = "AZURE_GPT_54_PRO_2026_0305"
     AZURE_GPT_55_2026_0424 = "AZURE_GPT_55_2026_0424"
     AZURE_GPT_55_PRO_2026_0424 = "AZURE_GPT_55_PRO_2026_0424"
+    AZURE_GPT_56_SOL_2026_0709 = "AZURE_GPT_56_SOL_2026_0709"
+    AZURE_GPT_56_TERRA_2026_0709 = "AZURE_GPT_56_TERRA_2026_0709"
+    AZURE_GPT_56_LUNA_2026_0709 = "AZURE_GPT_56_LUNA_2026_0709"
     AZURE_GPT_4o_2024_0513 = "AZURE_GPT_4o_2024_0513"
     AZURE_GPT_4o_2024_0806 = "AZURE_GPT_4o_2024_0806"
     AZURE_GPT_4o_2024_1120 = "AZURE_GPT_4o_2024_1120"
@@ -117,6 +120,9 @@ class LanguageModelName(StrEnum):
     LITELLM_OPENAI_GPT_54_THINKING = "litellm:openai-gpt-5-4-thinking"
     LITELLM_OPENAI_GPT_55 = "litellm:openai-gpt-5-5"
     LITELLM_OPENAI_GPT_55_PRO = "litellm:openai-gpt-5-5-pro"
+    LITELLM_OPENAI_GPT_56_SOL = "litellm:openai-gpt-5-6-sol"
+    LITELLM_OPENAI_GPT_56_TERRA = "litellm:openai-gpt-5-6-terra"
+    LITELLM_OPENAI_GPT_56_LUNA = "litellm:openai-gpt-5-6-luna"
     LITELLM_OPENAI_O1 = "litellm:openai-o1"
     LITELLM_OPENAI_O3 = "litellm:openai-o3"
     LITELLM_OPENAI_O3_DEEP_RESEARCH = "litellm:openai-o3-deep-research"
@@ -201,6 +207,9 @@ def get_encoder_name(model_name: LanguageModelName) -> EncoderName:
             | LMN.AZURE_GPT_54_PRO_2026_0305
             | LMN.AZURE_GPT_55_2026_0424
             | LMN.AZURE_GPT_55_PRO_2026_0424
+            | LMN.AZURE_GPT_56_SOL_2026_0709
+            | LMN.AZURE_GPT_56_TERRA_2026_0709
+            | LMN.AZURE_GPT_56_LUNA_2026_0709
             | LMN.AZURE_MODEL_ROUTER_2025_1118
             | LMN.LITELLM_OPENAI_GPT_5
             | LMN.LITELLM_OPENAI_GPT_5_MINI
@@ -215,6 +224,9 @@ def get_encoder_name(model_name: LanguageModelName) -> EncoderName:
             | LMN.LITELLM_OPENAI_GPT_54_THINKING
             | LMN.LITELLM_OPENAI_GPT_55
             | LMN.LITELLM_OPENAI_GPT_55_PRO
+            | LMN.LITELLM_OPENAI_GPT_56_SOL
+            | LMN.LITELLM_OPENAI_GPT_56_TERRA
+            | LMN.LITELLM_OPENAI_GPT_56_LUNA
             | LMN.LITELLM_OPENAI_O1
             | LMN.LITELLM_OPENAI_O3
             | LMN.LITELLM_OPENAI_O3_DEEP_RESEARCH
@@ -1122,6 +1134,46 @@ class LanguageModelInfo(BaseModel):
                     },
                     supported_reasoning_efforts=["medium", "high", "xhigh"],
                 )
+            case (
+                LanguageModelName.AZURE_GPT_56_SOL_2026_0709
+                | LanguageModelName.AZURE_GPT_56_TERRA_2026_0709
+                | LanguageModelName.AZURE_GPT_56_LUNA_2026_0709
+            ):
+                return cls(
+                    name=model_name,
+                    provider=LanguageModelProvider.AZURE,
+                    family=ModelFamily.OPENAI,
+                    version="2026-07-09",
+                    encoder_name=EncoderName.O200K_BASE,
+                    capabilities=[
+                        ModelCapabilities.CHAT_COMPLETIONS_API,
+                        ModelCapabilities.FUNCTION_CALLING,
+                        ModelCapabilities.PARALLEL_FUNCTION_CALLING,
+                        ModelCapabilities.REASONING,
+                        ModelCapabilities.RESPONSES_API,
+                        ModelCapabilities.STREAMING,
+                        ModelCapabilities.STRUCTURED_OUTPUT,
+                        ModelCapabilities.VISION,
+                    ],
+                    token_limits=LanguageModelTokenLimits(
+                        token_limit_input=922_000, token_limit_output=128_000
+                    ),
+                    info_cutoff_at=date(2026, 6, 1),
+                    published_at=date(2026, 7, 9),
+                    temperature_bounds=TemperatureBounds(
+                        min_temperature=0.0, max_temperature=1.0
+                    ),
+                    default_options={
+                        "reasoning_effort": "none",
+                    },
+                    supported_reasoning_efforts=[
+                        "none",
+                        "low",
+                        "medium",
+                        "high",
+                        "xhigh",
+                    ],
+                )
             case LanguageModelName.AZURE_GPT_4_TURBO_2024_0409:
                 return cls(
                     name=model_name,
@@ -2462,6 +2514,46 @@ class LanguageModelInfo(BaseModel):
                     },
                     supported_reasoning_efforts=["medium", "high", "xhigh"],
                 )
+            case (
+                LanguageModelName.LITELLM_OPENAI_GPT_56_SOL
+                | LanguageModelName.LITELLM_OPENAI_GPT_56_TERRA
+                | LanguageModelName.LITELLM_OPENAI_GPT_56_LUNA
+            ):
+                return cls(
+                    name=model_name,
+                    provider=LanguageModelProvider.LITELLM,
+                    family=ModelFamily.OPENAI,
+                    version="2026-07-09",
+                    encoder_name=EncoderName.O200K_BASE,
+                    capabilities=[
+                        ModelCapabilities.CHAT_COMPLETIONS_API,
+                        ModelCapabilities.FUNCTION_CALLING,
+                        ModelCapabilities.PARALLEL_FUNCTION_CALLING,
+                        ModelCapabilities.REASONING,
+                        ModelCapabilities.RESPONSES_API,
+                        ModelCapabilities.STREAMING,
+                        ModelCapabilities.STRUCTURED_OUTPUT,
+                        ModelCapabilities.VISION,
+                    ],
+                    token_limits=LanguageModelTokenLimits(
+                        token_limit_input=922_000, token_limit_output=128_000
+                    ),
+                    info_cutoff_at=date(2026, 6, 1),
+                    published_at=date(2026, 7, 9),
+                    temperature_bounds=TemperatureBounds(
+                        min_temperature=0.0, max_temperature=1.0
+                    ),
+                    default_options={
+                        "reasoning_effort": "none",
+                    },
+                    supported_reasoning_efforts=[
+                        "none",
+                        "low",
+                        "medium",
+                        "high",
+                        "xhigh",
+                    ],
+                )
             case LanguageModelName.LITELLM_OPENAI_O1:
                 return cls(
                     name=model_name,