Python: Move SessionStore to core and persist Foundry Responses sessions

python/.github/skills/agent-framework-py-release/SKILL.md

@@ -39,6 +39,8 @@ If the user states target versions or a date explicitly, use exactly what they s
 
 ## Non-negotiable rules
 
+- **Release workflow owns the CHANGELOG**: individual feature/fix PRs do not edit
+  `python/CHANGELOG.md`; this workflow creates the entries centrally from merged changes.
 - **CHANGELOG-driven bumps**: only packages mentioned in the new CHANGELOG section get version bumps. Exceptions: root follows core (==pin); user-opted cohort bump on betas.
 - **Follow `python-package-management` for package lifecycle and versioning rules** — do not duplicate those
   rules in this release workflow.
@@ -136,6 +138,9 @@ Root `agent-framework` is touched when `python/pyproject.toml`, `python/agent_fr
 
 ### 4. Draft CHANGELOG entries (THIS DRIVES THE BUMP LIST)
 
+Individual feature/fix PRs intentionally do not add CHANGELOG entries. During release preparation, derive this section
+centrally from the merged PRs since the last released tag.
+
 Locate `## [Unreleased]` and the top existing release header. INSERT a new section between them.
 
 **New section structure:**

python/AGENTS.md

@@ -37,6 +37,13 @@ team norms from a single conversation without explicit confirmation.
   match the feature-lifecycle stages documented in the
   `python-feature-lifecycle` skill.
 
+## Changelog Ownership
+
+- Individual feature, fix, documentation, and dependency PRs must not update
+  `python/CHANGELOG.md`.
+- CHANGELOG entries and release sections are assembled centrally during the
+  Python release-preparation workflow.
+
 ## Pull Request Description Guidance
 
 When preparing a PR description:

python/PACKAGE_STATUS.md

@@ -80,12 +80,61 @@ listed below.
   helper checks defined in `agent_framework/_evaluation.py`
 - `agent-framework-foundry`: `FoundryEvals`, `evaluate_traces`, and `evaluate_foundry_target`
 
-#### `SKILLS`
+#### `FILE_HISTORY`
 
-- `agent-framework-core`: exported skills APIs from `agent_framework`, including `Skill`,
-  `SkillResource`, `SkillScript`, `SkillScriptRunner`, and `SkillsProvider` from
+- `agent-framework-core`: `FileHistoryProvider` from `agent_framework/_sessions.py`
+
+#### `FIDES`
+
+- `agent-framework-core`: security labeling, content indirection, policy enforcement, and secure MCP
+  APIs from `agent_framework/security.py`, including `IntegrityLabel`, `ConfidentialityLabel`,
+  `ContentLabel`, `ContentVariableStore`, `SecureAgentConfig`, and `SecureMCPToolProxy`
+
+#### `FOUNDRY_TOOLS`
+
+- `agent-framework-foundry`: released-service tool helpers on `FoundryChatClient`, currently
+  `get_bing_grounding_tool` and `get_azure_ai_search_tool`
+
+#### `FOUNDRY_PREVIEW_TOOLS`
+
+- `agent-framework-foundry`: preview-service tool helpers on `FoundryChatClient`, including Bing
+  Custom Search, SharePoint, Fabric, Memory Search, Computer Use, Browser Automation, and A2A
+
+#### `FUNCTIONAL_WORKFLOWS`
+
+- `agent-framework-core`: functional workflow APIs from
+  `agent_framework/_workflows/_functional.py`, including `RunContext`, `step`,
+  `FunctionalWorkflow`, `workflow`, and `FunctionalWorkflowAgent`
+
+#### `HARNESS`
+
+- `agent-framework-core`: experimental harness APIs for background agents, file access, looping,
+  memory, and file-backed todo storage under `agent_framework/_harness/`
+
+#### `MCP_LONG_RUNNING_TASKS`
+
+- `agent-framework-core`: `MCPTaskOptions` from `agent_framework/_mcp.py`
+
+#### `MCP_SKILLS`
+
+- `agent-framework-core`: `MCPSkillResource`, `MCPSkill`, and `MCPSkillsSource` from
   `agent_framework/_skills.py`
 
+#### `PROGRESSIVE_TOOLS`
+
+- `agent-framework-core`: `FunctionInvocationContext.add_tools` and
+  `FunctionInvocationContext.remove_tools` from `agent_framework/_middleware.py`
+
+#### `SESSION_STORE`
+
+- `agent-framework-core`: `SessionStore` and `FileSessionStore` from
+  `agent_framework/_sessions.py`
+
+#### `TO_PROMPT_AGENT`
+
+- `agent-framework-foundry`: `to_prompt_agent` from
+  `agent_framework_foundry/_to_prompt_agent.py`
+
 ### Release-candidate features
 
 There are currently no feature-level `rc` APIs.

python/packages/core/AGENTS.md

@@ -74,13 +74,16 @@ agent_framework/
 ### Sessions (`_sessions.py`)
 
 - **`AgentSession`** - Manages conversation state and session metadata
+- **`SessionStore`** - Experimental in-memory opaque `session_id -> AgentSession` snapshot store; reads return independent copies
+- **`FileSessionStore`** - Experimental msgspec file-backed session snapshot store with atomic last-writer-wins updates; JSON is the default, `serialization_format="msgpack"` enables binary MessagePack, file keys use a restricted portable shape, and corrupt snapshots are quarantined before an error is raised
+- **`register_state_type`** - Registers custom `AgentSession.state` classes with stable type IDs and optional mapping codecs. Implicit Pydantic registration remains temporarily with `DeprecationWarning`, but module-level registration is needed to guarantee cold-start restoration.
 - **`ServiceSessionId`** - Mapping alias for structured service-owned continuation handles used in `AgentSession.service_session_id`
 - **`SessionContext`** - Context object for session-scoped data during agent runs. `extend_messages(...)` can attach
   ordered, deduplicated `origin_session_ids` attribution when a provider injects content from other sessions.
 - **`ContextProvider`** - Base class for context providers (RAG, memory systems)
 - **`HistoryProvider`** - Base class for conversation history storage
 - **`InMemoryHistoryProvider`** - Built-in session-state history provider for local runs
-- **`FileHistoryProvider`** - JSON Lines file-backed history provider storing one file per session with one message record per line
+- **`FileHistoryProvider`** - Experimental append-only file-backed history provider; msgspec JSON Lines is the default and `serialization_format="msgpack"` uses length-prefixed binary MessagePack records. Custom `dumps`/`loads` remain as deprecated JSON-only compatibility hooks and emit `DeprecationWarning` when supplied.
 
 ### Skills (`_skills.py`)
 

python/packages/core/agent_framework/__init__.py

@@ -194,11 +194,13 @@
         "AgentSession",
         "ContextProvider",
         "FileHistoryProvider",
+        "FileSessionStore",
         "HistoryProvider",
         "InMemoryHistoryProvider",
         "MESSAGE_INJECTION_PENDING_MESSAGES_STATE_KEY",
         "MessageInjectionMiddleware",
         "ServiceSessionId",
+        "SessionStore",
         "SessionContext",
         "enqueue_messages",
         "register_state_type",
@@ -451,6 +453,7 @@
     "FileMemoryProvider",
     "FileSearchMatch",
     "FileSearchResult",
+    "FileSessionStore",
     "FileSkill",
     "FileSkillScript",
     "FileSkillsSource",
@@ -516,6 +519,7 @@
     "SelectiveToolCallCompactionStrategy",
     "ServiceSessionId",
     "SessionContext",
+    "SessionStore",
     "SingleEdgeGroup",
     "Skill",
     "SkillFrontmatter",

python/packages/core/agent_framework/__init__.pyi

@@ -154,11 +154,13 @@ from ._sessions import (
     AgentSession,
     ContextProvider,
     FileHistoryProvider,
+    FileSessionStore,
     HistoryProvider,
     InMemoryHistoryProvider,
     MessageInjectionMiddleware,
     ServiceSessionId,
     SessionContext,
+    SessionStore,
     enqueue_messages,
     register_state_type,
 )
@@ -418,6 +420,7 @@ __all__ = [
     "FileMemoryProvider",
     "FileSearchMatch",
     "FileSearchResult",
+    "FileSessionStore",
     "FileSkill",
     "FileSkillScript",
     "FileSkillsSource",
@@ -483,6 +486,7 @@ __all__ = [
     "SelectiveToolCallCompactionStrategy",
     "ServiceSessionId",
     "SessionContext",
+    "SessionStore",
     "SingleEdgeGroup",
     "Skill",
     "SkillFrontmatter",

python/packages/core/agent_framework/_feature_stage.py

@@ -61,6 +61,7 @@ class ExperimentalFeature(str, Enum):
     MCP_LONG_RUNNING_TASKS = "MCP_LONG_RUNNING_TASKS"
     MCP_SKILLS = "MCP_SKILLS"
     PROGRESSIVE_TOOLS = "PROGRESSIVE_TOOLS"
+    SESSION_STORE = "SESSION_STORE"
     TO_PROMPT_AGENT = "TO_PROMPT_AGENT"
 
 
@@ -140,6 +141,15 @@ def _get_descriptor_callable(obj: Any) -> Callable[..., Any]:
     return cast(Callable[..., Any], obj.__func__)
 
 
+def _is_internal_framework_subclass(args: tuple[Any, ...]) -> bool:
+    """Return whether an experimental subclass is defined inside an Agent Framework package."""
+    subclass = next((arg for arg in args if isinstance(arg, type)), None)
+    if subclass is None:
+        return False
+    module = subclass.__module__
+    return module == "agent_framework" or module.startswith(("agent_framework.", "agent_framework_"))
+
+
 def _is_protocol_class(obj: Any) -> bool:
     return isinstance(obj, type) and bool(getattr(obj, "_is_protocol", False))
 
@@ -326,28 +336,30 @@ def __new__(cls: type[Any], /, *args: Any, **kwargs: Any) -> Any:
 
             @functools.wraps(original_init_subclass_func)
             def bound_init_subclass_wrapper(*args: Any, **kwargs: Any) -> Any:
-                _warn_on_feature_use(
-                    stage=stage,
-                    feature_id=feature_id,
-                    object_name=object_name,
-                    category=category,
-                )
+                if not _is_internal_framework_subclass(args):
+                    _warn_on_feature_use(
+                        stage=stage,
+                        feature_id=feature_id,
+                        object_name=object_name,
+                        category=category,
+                    )
                 return original_init_subclass_func(*args, **kwargs)
 
             experimental_class.__init_subclass__ = classmethod(bound_init_subclass_wrapper)  # type: ignore[assignment]
         else:
 
             @functools.wraps(original_init_subclass)
-            def init_subclass_wrapper(*args: Any, **kwargs: Any) -> Any:
-                _warn_on_feature_use(
-                    stage=stage,
-                    feature_id=feature_id,
-                    object_name=object_name,
-                    category=category,
-                )
+            def init_subclass_wrapper(subclass: type[Any], /, *args: Any, **kwargs: Any) -> Any:
+                if not _is_internal_framework_subclass((subclass, *args)):
+                    _warn_on_feature_use(
+                        stage=stage,
+                        feature_id=feature_id,
+                        object_name=object_name,
+                        category=category,
+                    )
                 return original_init_subclass(*args, **kwargs)
 
-            experimental_class.__init_subclass__ = init_subclass_wrapper
+            experimental_class.__init_subclass__ = classmethod(init_subclass_wrapper)  # type: ignore[assignment]
 
         return cast(FeatureStageT, experimental_class)
 

python/packages/core/agent_framework/_harness/_memory.py

@@ -976,8 +976,10 @@ def __init__(
             consolidation_client: Optional chat client override used only for consolidation so the
                 cleanup pass can use a cheaper or faster model than the main agent client.
             history_message_filter: Optional callback that can rewrite or drop messages before transcript save.
-            history_dumps: Callable used to serialize transcript JSONL.
-            history_loads: Callable used to deserialize transcript JSONL and state JSON.
+            history_dumps: Deprecated callback forwarded to ``FileHistoryProvider.dumps``.
+                Omit it to use msgspec JSON.
+            history_loads: Deprecated callback forwarded to ``FileHistoryProvider.loads``.
+                Omit it to use msgspec JSON.
         """
         super().__init__(
             source_id=source_id,