Fix lint

changelog.d/19041.feature

@@ -0,0 +1 @@
+Add companion endpoint for MSC4360: Sliding Sync Threads Extension.

synapse/handlers/relations.py

@@ -20,6 +20,7 @@
 #
 import enum
 import logging
+from collections import defaultdict
 from typing import (
     TYPE_CHECKING,
     Collection,
@@ -30,24 +31,59 @@ from typing import (
 
 import attr
 
-from synapse.api.constants import Direction, EventTypes, RelationTypes
+from synapse.api.constants import Direction, EventTypes, Membership, RelationTypes
 from synapse.api.errors import SynapseError
 from synapse.events import EventBase, relation_from_event
 from synapse.events.utils import SerializeEventConfig
 from synapse.logging.context import make_deferred_yieldable, run_in_background
 from synapse.logging.opentracing import trace
-from synapse.storage.databases.main.relations import ThreadsNextBatch, _RelatedEvent
+from synapse.storage.databases.main.relations import (
+    ThreadsNextBatch,
+    ThreadUpdateInfo,
+    _RelatedEvent,
+)
 from synapse.streams.config import PaginationConfig
-from synapse.types import JsonDict, Requester, UserID
+from synapse.types import (
+    JsonDict,
+    Requester,
+    RoomStreamToken,
+    StreamKeyType,
+    StreamToken,
+    UserID,
+)
 from synapse.util.async_helpers import gather_results
 from synapse.visibility import filter_events_for_client
 
 if TYPE_CHECKING:
+    from synapse.events.utils import EventClientSerializer
+    from synapse.handlers.sliding_sync.room_lists import RoomsForUserType
     from synapse.server import HomeServer
+    from synapse.storage.databases.main import DataStore
 
 
 logger = logging.getLogger(__name__)
 
+# Type aliases for thread update processing
+ThreadUpdatesMap = dict[str, list[ThreadUpdateInfo]]
+ThreadRootsMap = dict[str, EventBase]
+AggregationsMap = dict[str, "BundledAggregations"]
+
+
+@attr.s(slots=True, frozen=True, auto_attribs=True)
+class ThreadUpdate:
+    """
+    Data for a single thread update.
+
+    Attributes:
+        thread_root: The thread root event, or None if not requested/not visible
+        prev_batch: Per-thread pagination token for fetching older events in this thread
+        bundled_aggregations: Bundled aggregations for the thread root event
+    """
+
+    thread_root: EventBase | None
+    prev_batch: StreamToken | None
+    bundled_aggregations: "BundledAggregations | None" = None
+
 
 class ThreadsListInclude(str, enum.Enum):
     """Valid values for the 'include' flag of /threads."""
@@ -544,6 +580,367 @@ class RelationsHandler:
 
         return results
 
+    async def _filter_thread_updates_for_user(
+        self,
+        all_thread_updates: ThreadUpdatesMap,
+        user_id: str,
+    ) -> ThreadUpdatesMap:
+        """Process thread updates by filtering for visibility.
+
+        Takes raw thread updates from storage and filters them based on whether the
+        user can see the events. Preserves the ordering of updates within each thread.
+
+        Args:
+            all_thread_updates: Map of thread_id to list of ThreadUpdateInfo objects
+            user_id: The user ID to filter events for
+
+        Returns:
+            Filtered map of thread_id to list of ThreadUpdateInfo objects, containing
+            only updates for events the user can see.
+        """
+        # Build a mapping of event_id -> (thread_id, update) for efficient lookup
+        # during visibility filtering.
+        event_to_thread_map: dict[str, tuple[str, ThreadUpdateInfo]] = {}
+        for thread_id, updates in all_thread_updates.items():
+            for update in updates:
+                event_to_thread_map[update.event_id] = (thread_id, update)
+
+        # Fetch and filter events for visibility
+        all_events = await self._main_store.get_events_as_list(
+            event_to_thread_map.keys()
+        )
+        filtered_events = await filter_events_for_client(
+            self._storage_controllers, user_id, all_events
+        )
+
+        # Rebuild thread updates from filtered events
+        filtered_updates: ThreadUpdatesMap = defaultdict(list)
+        for event in filtered_events:
+            if event.event_id in event_to_thread_map:
+                thread_id, update = event_to_thread_map[event.event_id]
+                filtered_updates[thread_id].append(update)
+
+        return filtered_updates
+
+    def _build_thread_updates_response(
+        self,
+        filtered_updates: ThreadUpdatesMap,
+        thread_root_event_map: ThreadRootsMap,
+        aggregations_map: AggregationsMap,
+        global_prev_batch_token: StreamToken | None,
+    ) -> dict[str, dict[str, ThreadUpdate]]:
+        """Build thread update response structure with per-thread prev_batch tokens.
+
+        Args:
+            filtered_updates: Map of thread_root_id to list of ThreadUpdateInfo
+            thread_root_event_map: Map of thread_root_id to EventBase
+            aggregations_map: Map of thread_root_id to BundledAggregations
+            global_prev_batch_token: Global pagination token, or None if no more results
+
+        Returns:
+            Map of room_id to thread_root_id to ThreadUpdate
+        """
+        thread_updates: dict[str, dict[str, ThreadUpdate]] = {}
+
+        for thread_root_id, updates in filtered_updates.items():
+            # We only care about the latest update for the thread
+            # Updates are already sorted by stream_ordering DESC from the database query,
+            # and filter_events_for_client preserves order, so updates[0] is guaranteed to be
+            # the latest event for each thread.
+            latest_update = updates[0]
+            room_id = latest_update.room_id
+
+            # Generate per-thread prev_batch token if this thread has multiple visible updates
+            # or if we hit the global limit.
+            # When we hit the global limit, we generate prev_batch tokens for all threads, even if
+            # we only saw 1 update for them. This is to cover the case where we only saw
+            # a single update for a given thread, but the global limit prevents us from
+            # obtaining other updates which would have otherwise been included in the range.
+            per_thread_prev_batch = None
+            if len(updates) > 1 or global_prev_batch_token is not None:
+                # Create a token pointing to one position before the latest event's stream position.
+                # This makes it exclusive - /relations with dir=b won't return the latest event again.
+                # Use StreamToken.START as base (all other streams at 0) since only room position matters.
+                per_thread_prev_batch = StreamToken.START.copy_and_replace(
+                    StreamKeyType.ROOM,
+                    RoomStreamToken(stream=latest_update.stream_ordering - 1),
+                )
+
+            if room_id not in thread_updates:
+                thread_updates[room_id] = {}
+
+            thread_updates[room_id][thread_root_id] = ThreadUpdate(
+                thread_root=thread_root_event_map.get(thread_root_id),
+                prev_batch=per_thread_prev_batch,
+                bundled_aggregations=aggregations_map.get(thread_root_id),
+            )
+
+        return thread_updates
+
+    async def _fetch_thread_updates(
+        self,
+        room_ids: frozenset[str],
+        room_membership_map: Mapping[str, "RoomsForUserType"],
+        from_token: StreamToken | None,
+        to_token: StreamToken,
+        limit: int,
+        exclude_thread_ids: set[str] | None = None,
+    ) -> tuple[ThreadUpdatesMap, StreamToken | None]:
+        """Fetch thread updates across multiple rooms, handling membership states properly.
+
+        This method separates rooms based on membership status (LEAVE/BAN vs others)
+        and queries them appropriately to prevent data leaks. For rooms where the user
+        has left or been banned, we bound the query to their leave/ban event position.
+
+        Args:
+            room_ids: The set of room IDs to fetch thread updates for
+            room_membership_map: Map of room_id to RoomsForUserType containing membership info
+            from_token: Lower bound (exclusive) for the query, or None for no lower bound
+            to_token: Upper bound for the query (for joined/invited/knocking rooms)
+            limit: Maximum number of thread updates to return across all rooms
+            exclude_thread_ids: Optional set of thread IDs to exclude from results
+
+        Returns:
+            A tuple of:
+            - Map of thread_id to list of ThreadUpdateInfo objects
+            - Global prev_batch token if there are more results, None otherwise
+        """
+        # Separate rooms based on membership to handle LEAVE/BAN rooms specially
+        leave_ban_rooms: set[str] = set()
+        other_rooms: set[str] = set()
+
+        for room_id in room_ids:
+            membership_info = room_membership_map.get(room_id)
+            if membership_info and membership_info.membership in (
+                Membership.LEAVE,
+                Membership.BAN,
+            ):
+                leave_ban_rooms.add(room_id)
+            else:
+                other_rooms.add(room_id)
+
+        # Fetch thread updates from storage, handling LEAVE/BAN rooms separately
+        all_thread_updates: ThreadUpdatesMap = {}
+        prev_batch_token: StreamToken | None = None
+        remaining_limit = limit
+
+        # Query LEAVE/BAN rooms with bounded to_token to prevent data leaks
+        if leave_ban_rooms:
+            for room_id in leave_ban_rooms:
+                if remaining_limit <= 0:
+                    # We've hit the limit, set prev_batch to indicate more results
+                    prev_batch_token = to_token
+                    break
+
+                membership_info = room_membership_map[room_id]
+                bounded_to_token = membership_info.event_pos.to_room_stream_token()
+
+                (
+                    room_thread_updates,
+                    room_prev_batch,
+                ) = await self._main_store.get_thread_updates_for_rooms(
+                    room_ids={room_id},
+                    from_token=from_token.room_key if from_token else None,
+                    to_token=bounded_to_token,
+                    limit=remaining_limit,
+                    exclude_thread_ids=exclude_thread_ids,
+                )
+
+                # Count updates and reduce remaining limit
+                num_updates = sum(
+                    len(updates) for updates in room_thread_updates.values()
+                )
+                remaining_limit -= num_updates
+
+                # Merge updates
+                for thread_id, updates in room_thread_updates.items():
+                    all_thread_updates.setdefault(thread_id, []).extend(updates)
+
+                # Merge prev_batch tokens (take the maximum for backward pagination)
+                if room_prev_batch is not None:
+                    if prev_batch_token is None:
+                        prev_batch_token = room_prev_batch
+                    elif (
+                        room_prev_batch.room_key.stream
+                        > prev_batch_token.room_key.stream
+                    ):
+                        prev_batch_token = room_prev_batch
+
+        # Query other rooms (joined/invited/knocking) with normal to_token
+        if other_rooms and remaining_limit > 0:
+            (
+                other_thread_updates,
+                other_prev_batch,
+            ) = await self._main_store.get_thread_updates_for_rooms(
+                room_ids=other_rooms,
+                from_token=from_token.room_key if from_token else None,
+                to_token=to_token.room_key,
+                limit=remaining_limit,
+                exclude_thread_ids=exclude_thread_ids,
+            )
+
+            # Merge updates
+            for thread_id, updates in other_thread_updates.items():
+                all_thread_updates.setdefault(thread_id, []).extend(updates)
+
+            # Merge prev_batch tokens
+            if other_prev_batch is not None:
+                if prev_batch_token is None:
+                    prev_batch_token = other_prev_batch
+                elif (
+                    other_prev_batch.room_key.stream > prev_batch_token.room_key.stream
+                ):
+                    prev_batch_token = other_prev_batch
+
+        return all_thread_updates, prev_batch_token
+
+    async def get_thread_updates_for_rooms(
+        self,
+        room_ids: frozenset[str],
+        room_membership_map: Mapping[str, "RoomsForUserType"],
+        user_id: str,
+        from_token: StreamToken | None,
+        to_token: StreamToken,
+        limit: int,
+        include_roots: bool = False,
+        exclude_thread_ids: set[str] | None = None,
+    ) -> tuple[dict[str, dict[str, ThreadUpdate]], StreamToken | None]:
+        """Get thread updates across multiple rooms with full processing pipeline.
+
+        This is the main entry point for fetching thread updates. It handles:
+        - Fetching updates with membership-based security
+        - Filtering for visibility
+        - Optionally fetching thread roots and aggregations
+        - Building the response structure
+
+        Args:
+            room_ids: The set of room IDs to fetch updates for
+            room_membership_map: Map of room_id to RoomsForUserType for membership info
+            user_id: The user requesting the updates
+            from_token: Lower bound (exclusive) for the query
+            to_token: Upper bound for the query
+            limit: Maximum number of updates to return
+            include_roots: Whether to fetch and include thread root events (default: False)
+            exclude_thread_ids: Optional set of thread IDs to exclude
+
+        Returns:
+            A tuple of:
+            - Map of room_id to thread_root_id to ThreadUpdate
+            - Global prev_batch token if there are more results, None otherwise
+        """
+        # Fetch thread updates with membership handling
+        all_thread_updates, prev_batch_token = await self._fetch_thread_updates(
+            room_ids=room_ids,
+            room_membership_map=room_membership_map,
+            from_token=from_token,
+            to_token=to_token,
+            limit=limit,
+            exclude_thread_ids=exclude_thread_ids,
+        )
+
+        if not all_thread_updates:
+            return {}, prev_batch_token
+
+        # Filter thread updates for visibility
+        filtered_updates = await self._filter_thread_updates_for_user(
+            all_thread_updates, user_id
+        )
+
+        if not filtered_updates:
+            return {}, prev_batch_token
+
+        # Optionally fetch thread root events and their bundled aggregations
+        thread_root_event_map: ThreadRootsMap = {}
+        aggregations_map: AggregationsMap = {}
+        if include_roots:
+            # Fetch thread root events
+            thread_root_events = await self._main_store.get_events_as_list(
+                filtered_updates.keys()
+            )
+            thread_root_event_map = {e.event_id: e for e in thread_root_events}
+
+            # Fetch bundled aggregations for the thread roots
+            if thread_root_event_map:
+                aggregations_map = await self.get_bundled_aggregations(
+                    thread_root_event_map.values(),
+                    user_id,
+                )
+
+        # Build response structure with per-thread prev_batch tokens
+        thread_updates = self._build_thread_updates_response(
+            filtered_updates=filtered_updates,
+            thread_root_event_map=thread_root_event_map,
+            aggregations_map=aggregations_map,
+            global_prev_batch_token=prev_batch_token,
+        )
+
+        return thread_updates, prev_batch_token
+
+    @staticmethod
+    async def serialize_thread_updates(
+        thread_updates: Mapping[str, Mapping[str, ThreadUpdate]],
+        prev_batch_token: StreamToken | None,
+        event_serializer: "EventClientSerializer",
+        time_now: int,
+        store: "DataStore",
+        serialize_options: SerializeEventConfig,
+    ) -> JsonDict:
+        """
+        Serialize thread updates to JSON format.
+
+        This helper handles serialization of ThreadUpdate objects for both the
+        companion endpoint and the sliding sync extension.
+
+        Args:
+            thread_updates: Map of room_id to thread_root_id to ThreadUpdate
+            prev_batch_token: Global pagination token for fetching more updates
+            event_serializer: The event serializer to use
+            time_now: Current time in milliseconds for event serialization
+            store: Datastore for serializing stream tokens
+            serialize_options: Serialization config
+
+        Returns:
+            JSON-serializable dict with "updates" and optionally "prev_batch"
+        """
+        updates_dict: JsonDict = {}
+
+        for room_id, room_threads in thread_updates.items():
+            room_updates: JsonDict = {}
+            for thread_root_id, update in room_threads.items():
+                update_dict: JsonDict = {}
+
+                # Serialize thread_root event if present
+                if update.thread_root is not None:
+                    bundle_aggs_map = (
+                        {thread_root_id: update.bundled_aggregations}
+                        if update.bundled_aggregations is not None
+                        else None
+                    )
+                    serialized_events = await event_serializer.serialize_events(
+                        [update.thread_root],
+                        time_now,
+                        config=serialize_options,
+                        bundle_aggregations=bundle_aggs_map,
+                    )
+                    if serialized_events:
+                        update_dict["thread_root"] = serialized_events[0]
+
+                # Add per-thread prev_batch if present
+                if update.prev_batch is not None:
+                    update_dict["prev_batch"] = await update.prev_batch.to_string(store)
+
+                room_updates[thread_root_id] = update_dict
+
+            updates_dict[room_id] = room_updates
+
+        result: JsonDict = {"updates": updates_dict}
+
+        # Add global prev_batch token if present
+        if prev_batch_token is not None:
+            result["prev_batch"] = await prev_batch_token.to_string(store)
+
+        return result
+
     async def get_threads(
         self,
         requester: Requester,

synapse/handlers/sliding_sync/extensions.py

@@ -14,7 +14,6 @@
 
 import itertools
 import logging
-from collections import defaultdict
 from typing import (
     TYPE_CHECKING,
     AbstractSet,
@@ -31,7 +30,6 @@ from synapse.api.constants import (
     AccountDataTypes,
     EduTypes,
     EventContentFields,
-    Membership,
     MRelatesToFields,
     RelationTypes,
 )
@@ -40,15 +38,12 @@ from synapse.handlers.receipts import ReceiptEventSource
 from synapse.handlers.sliding_sync.room_lists import RoomsForUserType
 from synapse.logging.opentracing import trace
 from synapse.storage.databases.main.receipts import ReceiptInRoom
-from synapse.storage.databases.main.relations import ThreadUpdateInfo
 from synapse.types import (
     DeviceListUpdates,
     JsonMapping,
     MultiWriterStreamToken,
-    RoomStreamToken,
     SlidingSyncStreamToken,
     StrCollection,
-    StreamKeyType,
     StreamToken,
     ThreadSubscriptionsToken,
 )
@@ -64,7 +59,6 @@ from synapse.util.async_helpers import (
     concurrently_execute,
     gather_optional_coroutines,
 )
-from synapse.visibility import filter_events_for_client
 
 _ThreadSubscription: TypeAlias = (
     SlidingSyncResult.Extensions.ThreadSubscriptionsExtension.ThreadSubscription
@@ -72,7 +66,6 @@ _ThreadSubscription: TypeAlias = (
 _ThreadUnsubscription: TypeAlias = (
     SlidingSyncResult.Extensions.ThreadSubscriptionsExtension.ThreadUnsubscription
 )
-_ThreadUpdate: TypeAlias = SlidingSyncResult.Extensions.ThreadsExtension.ThreadUpdate
 
 if TYPE_CHECKING:
     from synapse.server import HomeServer
@@ -1040,42 +1033,6 @@ class SlidingSyncExtensionHandler:
                         threads_in_timeline.add(thread_id)
         return threads_in_timeline
 
-    def _merge_prev_batch_token(
-        self,
-        current_token: StreamToken | None,
-        new_token: StreamToken | None,
-    ) -> StreamToken | None:
-        """Merge two prev_batch tokens, taking the maximum (latest) for backwards pagination.
-
-        Args:
-            current_token: The current prev_batch token (may be None)
-            new_token: The new prev_batch token to merge (may be None)
-
-        Returns:
-            The merged token (maximum of the two, or None if both are None)
-        """
-        if new_token is None:
-            return current_token
-        if current_token is None:
-            return new_token
-        if new_token.room_key.stream > current_token.room_key.stream:
-            return new_token
-        return current_token
-
-    def _merge_thread_updates(
-        self,
-        target: dict[str, list[ThreadUpdateInfo]],
-        source: dict[str, list[ThreadUpdateInfo]],
-    ) -> None:
-        """Merge thread updates from source into target.
-
-        Args:
-            target: The target dict to merge into (modified in place)
-            source: The source dict to merge from
-        """
-        for thread_id, updates in source.items():
-            target.setdefault(thread_id, []).extend(updates)
-
     async def get_threads_extension_response(
         self,
         sync_config: SlidingSyncConfig,
@@ -1117,159 +1074,26 @@ class SlidingSyncExtensionHandler:
                 actual_room_response_map
             )
 
-        # Separate rooms into groups based on membership status.
-        # For LEAVE/BAN rooms, we need to bound the to_token to prevent leaking events
-        # that occurred after the user left/was banned.
-        leave_ban_rooms: set[str] = set()
-        other_rooms: set[str] = set()
-
-        for room_id in actual_room_ids:
-            membership_info = room_membership_for_user_at_to_token_map.get(room_id)
-            if membership_info and membership_info.membership in (
-                Membership.LEAVE,
-                Membership.BAN,
-            ):
-                leave_ban_rooms.add(room_id)
-            else:
-                other_rooms.add(room_id)
-
-        # Fetch thread updates, handling LEAVE/BAN rooms separately to avoid data leaks.
-        all_thread_updates: dict[str, list[ThreadUpdateInfo]] = {}
-        prev_batch_token: StreamToken | None = None
-        remaining_limit = threads_request.limit
-
-        # Query for rooms where the user has left or been banned, using their leave/ban
-        # event position as the upper bound to prevent seeing events after they left.
-        if leave_ban_rooms:
-            for room_id in leave_ban_rooms:
-                if remaining_limit <= 0:
-                    # We've already fetched enough updates, but we still need to set
-                    # prev_batch to indicate there are more results.
-                    prev_batch_token = to_token
-                    break
-
-                membership_info = room_membership_for_user_at_to_token_map[room_id]
-                bounded_to_token = membership_info.event_pos.to_room_stream_token()
-
-                (
-                    room_thread_updates,
-                    room_prev_batch,
-                ) = await self.store.get_thread_updates_for_rooms(
-                    room_ids={room_id},
-                    from_token=from_token.stream_token.room_key if from_token else None,
-                    to_token=bounded_to_token,
-                    limit=remaining_limit,
-                    exclude_thread_ids=threads_to_exclude,
-                )
-
-                # Count how many updates we fetched and reduce the remaining limit
-                num_updates = sum(
-                    len(updates) for updates in room_thread_updates.values()
-                )
-                remaining_limit -= num_updates
-
-                self._merge_thread_updates(all_thread_updates, room_thread_updates)
-                prev_batch_token = self._merge_prev_batch_token(
-                    prev_batch_token, room_prev_batch
-                )
-
-        # Query for rooms where the user is joined, invited, or knocking, using the
-        # normal to_token as the upper bound.
-        if other_rooms and remaining_limit > 0:
-            (
-                other_thread_updates,
-                other_prev_batch,
-            ) = await self.store.get_thread_updates_for_rooms(
-                room_ids=other_rooms,
-                from_token=from_token.stream_token.room_key if from_token else None,
-                to_token=to_token.room_key,
-                limit=remaining_limit,
-                exclude_thread_ids=threads_to_exclude,
-            )
-
-            self._merge_thread_updates(all_thread_updates, other_thread_updates)
-            prev_batch_token = self._merge_prev_batch_token(
-                prev_batch_token, other_prev_batch
-            )
-
-        if len(all_thread_updates) == 0:
-            return None
-
-        # Build a mapping of event_id -> (thread_id, update) for efficient lookup
-        # during visibility filtering.
-        event_to_thread_map: dict[str, tuple[str, ThreadUpdateInfo]] = {}
-        for thread_id, updates in all_thread_updates.items():
-            for update in updates:
-                event_to_thread_map[update.event_id] = (thread_id, update)
-
-        # Fetch and filter events for visibility
-        all_events = await self.store.get_events_as_list(event_to_thread_map.keys())
-        filtered_events = await filter_events_for_client(
-            self._storage_controllers, sync_config.user.to_string(), all_events
+        # Get thread updates using unified helper
+        user_id = sync_config.user.to_string()
+        (
+            thread_updates_response,
+            prev_batch_token,
+        ) = await self.relations_handler.get_thread_updates_for_rooms(
+            room_ids=frozenset(actual_room_ids),
+            room_membership_map=room_membership_for_user_at_to_token_map,
+            user_id=user_id,
+            from_token=from_token.stream_token if from_token else None,
+            to_token=to_token,
+            limit=threads_request.limit,
+            include_roots=threads_request.include_roots,
+            exclude_thread_ids=threads_to_exclude,
         )
 
-        # Rebuild thread updates from filtered events
-        filtered_updates: dict[str, list[ThreadUpdateInfo]] = defaultdict(list)
-        for event in filtered_events:
-            if event.event_id in event_to_thread_map:
-                thread_id, update = event_to_thread_map[event.event_id]
-                filtered_updates[thread_id].append(update)
-
-        if not filtered_updates:
+        if not thread_updates_response:
             return None
 
-        # Note: Updates are already sorted by stream_ordering DESC from the database query,
-        # and filter_events_for_client preserves order, so updates[0] is guaranteed to be
-        # the latest event for each thread.
-
-        # Optionally fetch thread root events and their bundled aggregations
-        thread_root_event_map = {}
-        aggregations_map = {}
-        if threads_request.include_roots:
-            thread_root_events = await self.store.get_events_as_list(
-                filtered_updates.keys()
-            )
-            thread_root_event_map = {e.event_id: e for e in thread_root_events}
-
-            if thread_root_event_map:
-                aggregations_map = (
-                    await self.relations_handler.get_bundled_aggregations(
-                        thread_root_event_map.values(),
-                        sync_config.user.to_string(),
-                    )
-                )
-
-        thread_updates: dict[str, dict[str, _ThreadUpdate]] = {}
-        for thread_root, updates in filtered_updates.items():
-            # We only care about the latest update for the thread.
-            # After sorting above, updates[0] is guaranteed to be the latest (highest stream_ordering).
-            latest_update = updates[0]
-
-            # Generate per-thread prev_batch token if this thread has multiple visible updates.
-            # When we hit the global limit, we generate prev_batch tokens for all threads, even if
-            # we only saw 1 update for them. This is to cover the case where we only saw
-            # a single update for a given thread, but the global limit prevents us from
-            # obtaining other updates which would have otherwise been included in the
-            # range.
-            per_thread_prev_batch = None
-            if len(updates) > 1 or prev_batch_token is not None:
-                # Create a token pointing to one position before the latest event's stream position.
-                # This makes it exclusive - /relations with dir=b won't return the latest event again.
-                # Use StreamToken.START as base (all other streams at 0) since only room position matters.
-                per_thread_prev_batch = StreamToken.START.copy_and_replace(
-                    StreamKeyType.ROOM,
-                    RoomStreamToken(stream=latest_update.stream_ordering - 1),
-                )
-
-            thread_updates.setdefault(latest_update.room_id, {})[thread_root] = (
-                _ThreadUpdate(
-                    thread_root=thread_root_event_map.get(thread_root),
-                    prev_batch=per_thread_prev_batch,
-                    bundled_aggregations=aggregations_map.get(thread_root),
-                )
-            )
-
         return SlidingSyncResult.Extensions.ThreadsExtension(
-            updates=thread_updates,
+            updates=thread_updates_response,
             prev_batch=prev_batch_token,
         )

synapse/rest/client/relations.py

@@ -20,17 +20,33 @@
 
 import logging
 import re
-from typing import TYPE_CHECKING
+from typing import TYPE_CHECKING, Annotated
 
-from synapse.api.constants import Direction
+from pydantic import StrictBool, StrictStr
+from pydantic.types import StringConstraints
+
+from synapse.api.constants import Direction, Membership
+from synapse.api.errors import SynapseError
+from synapse.events.utils import SerializeEventConfig
 from synapse.handlers.relations import ThreadsListInclude
 from synapse.http.server import HttpServer
-from synapse.http.servlet import RestServlet, parse_boolean, parse_integer, parse_string
+from synapse.http.servlet import (
+    RestServlet,
+    parse_and_validate_json_object_from_request,
+    parse_boolean,
+    parse_integer,
+    parse_string,
+)
 from synapse.http.site import SynapseRequest
 from synapse.rest.client._base import client_patterns
 from synapse.storage.databases.main.relations import ThreadsNextBatch
-from synapse.streams.config import PaginationConfig
-from synapse.types import JsonDict
+from synapse.streams.config import (
+    PaginationConfig,
+    extract_stream_token_from_pagination_token,
+)
+from synapse.types import JsonDict, RoomStreamToken, StreamKeyType, StreamToken, UserID
+from synapse.types.handlers.sliding_sync import PerConnectionState, SlidingSyncConfig
+from synapse.types.rest.client import RequestBodyModel, SlidingSyncBody
 
 if TYPE_CHECKING:
     from synapse.server import HomeServer
@@ -38,6 +54,39 @@ if TYPE_CHECKING:
 logger = logging.getLogger(__name__)
 
 
+class ThreadUpdatesBody(RequestBodyModel):
+    """
+    Thread updates companion endpoint request body (MSC4360).
+
+    Allows paginating thread updates using the same room selection as a sliding sync
+    request. This enables clients to fetch thread updates for the same set of rooms
+    that were included in their sliding sync response.
+
+    Attributes:
+        lists: Sliding window API lists, using the same structure as SlidingSyncBody.lists.
+            If provided along with room_subscriptions, the union of rooms from both will
+            be used.
+        room_subscriptions: Room subscription API rooms, using the same structure as
+            SlidingSyncBody.room_subscriptions. If provided along with lists, the union
+            of rooms from both will be used.
+        include_roots: Whether to include the thread root events in the response.
+            Defaults to False.
+
+    If neither lists nor room_subscriptions are provided, thread updates from all
+    joined rooms are returned.
+    """
+
+    lists: (
+        dict[
+            Annotated[str, StringConstraints(max_length=64, strict=True)],
+            SlidingSyncBody.SlidingSyncList,
+        ]
+        | None
+    ) = None
+    room_subscriptions: dict[StrictStr, SlidingSyncBody.RoomSubscription] | None = None
+    include_roots: StrictBool = False
+
+
 class RelationPaginationServlet(RestServlet):
     """API to paginate relations on an event by topological ordering, optionally
     filtered by relation type and event type.
@@ -133,6 +182,167 @@ class ThreadsServlet(RestServlet):
         return 200, result
 
 
+class ThreadUpdatesServlet(RestServlet):
+    """
+    Companion endpoint to the Sliding Sync threads extension (MSC4360).
+    Allows clients to bulk fetch thread updates across all joined rooms.
+    """
+
+    PATTERNS = client_patterns(
+        "/io.element.msc4360/thread_updates$",
+        unstable=True,
+        releases=(),
+    )
+    CATEGORY = "Client API requests"
+
+    def __init__(self, hs: "HomeServer"):
+        super().__init__()
+        self.clock = hs.get_clock()
+        self.auth = hs.get_auth()
+        self.store = hs.get_datastores().main
+        self.relations_handler = hs.get_relations_handler()
+        self.event_serializer = hs.get_event_client_serializer()
+        self._storage_controllers = hs.get_storage_controllers()
+        self.sliding_sync_handler = hs.get_sliding_sync_handler()
+
+    async def on_POST(self, request: SynapseRequest) -> tuple[int, JsonDict]:
+        requester = await self.auth.get_user_by_req(request)
+
+        # Parse request body
+        body = parse_and_validate_json_object_from_request(request, ThreadUpdatesBody)
+
+        # Parse query parameters
+        dir_str = parse_string(request, "dir", default="b")
+        if dir_str != "b":
+            raise SynapseError(
+                400,
+                "The 'dir' parameter must be 'b' (backward). Forward pagination is not supported.",
+            )
+
+        limit = parse_integer(request, "limit", default=100)
+        if limit <= 0:
+            raise SynapseError(400, "The 'limit' parameter must be positive.")
+
+        from_token_str = parse_string(request, "from")
+        to_token_str = parse_string(request, "to")
+
+        # Parse pagination tokens
+        from_token: StreamToken | None = None
+        to_token: StreamToken | None = None
+
+        if from_token_str:
+            try:
+                stream_token_str = extract_stream_token_from_pagination_token(
+                    from_token_str
+                )
+                from_token = await StreamToken.from_string(self.store, stream_token_str)
+            except Exception as e:
+                logger.exception("Error parsing 'from' token: %s", from_token_str)
+                raise SynapseError(400, "'from' parameter is invalid") from e
+
+        if to_token_str:
+            try:
+                stream_token_str = extract_stream_token_from_pagination_token(
+                    to_token_str
+                )
+                to_token = await StreamToken.from_string(self.store, stream_token_str)
+            except Exception:
+                raise SynapseError(400, "'to' parameter is invalid")
+
+        # Get the list of rooms to fetch thread updates for
+        user_id = requester.user.to_string()
+        user = UserID.from_string(user_id)
+
+        # Get the current stream token for membership lookup
+        if from_token is None:
+            max_stream_ordering = self.store.get_room_max_stream_ordering()
+            current_token = StreamToken.START.copy_and_replace(
+                StreamKeyType.ROOM, RoomStreamToken(stream=max_stream_ordering)
+            )
+        else:
+            current_token = from_token
+
+        # Get room membership information to properly handle LEAVE/BAN rooms
+        (
+            room_membership_for_user_at_to_token_map,
+            _,
+            _,
+        ) = await self.sliding_sync_handler.room_lists.get_room_membership_for_user_at_to_token(
+            user=user,
+            to_token=current_token,
+            from_token=None,
+        )
+
+        # Determine which rooms to fetch updates for based on lists/room_subscriptions
+        if body.lists is not None or body.room_subscriptions is not None:
+            # Use sliding sync room selection logic
+            sync_config = SlidingSyncConfig(
+                user=user,
+                requester=requester,
+                lists=body.lists,
+                room_subscriptions=body.room_subscriptions,
+            )
+
+            # Use the sliding sync room list handler to get the same set of rooms
+            interested_rooms = (
+                await self.sliding_sync_handler.room_lists.compute_interested_rooms(
+                    sync_config=sync_config,
+                    previous_connection_state=PerConnectionState(),
+                    to_token=current_token,
+                    from_token=None,
+                )
+            )
+
+            room_ids = frozenset(interested_rooms.relevant_room_map.keys())
+        else:
+            # No lists or room_subscriptions, use only joined rooms
+            room_ids = frozenset(
+                room_id
+                for room_id, membership_info in room_membership_for_user_at_to_token_map.items()
+                if membership_info.membership == Membership.JOIN
+            )
+
+        # Get thread updates using unified helper
+        (
+            thread_updates,
+            prev_batch_token,
+        ) = await self.relations_handler.get_thread_updates_for_rooms(
+            room_ids=room_ids,
+            room_membership_map=room_membership_for_user_at_to_token_map,
+            user_id=user_id,
+            from_token=to_token,
+            to_token=from_token if from_token else current_token,
+            limit=limit,
+            include_roots=body.include_roots,
+        )
+
+        if not thread_updates:
+            return 200, {"chunk": {}}
+
+        # Serialize thread updates using shared helper
+        time_now = self.clock.time_msec()
+        serialize_options = SerializeEventConfig(requester=requester)
+
+        serialized = await self.relations_handler.serialize_thread_updates(
+            thread_updates=thread_updates,
+            prev_batch_token=prev_batch_token,
+            event_serializer=self.event_serializer,
+            time_now=time_now,
+            store=self.store,
+            serialize_options=serialize_options,
+        )
+
+        # Build response with "chunk" wrapper and "next_batch" key
+        # (companion endpoint uses different key names than sliding sync)
+        response: JsonDict = {"chunk": serialized["updates"]}
+        if "prev_batch" in serialized:
+            response["next_batch"] = serialized["prev_batch"]
+
+        return 200, response
+
+
 def register_servlets(hs: "HomeServer", http_server: HttpServer) -> None:
     RelationPaginationServlet(hs).register(http_server)
     ThreadsServlet(hs).register(http_server)
+    if hs.config.experimental.msc4360_enabled:
+        ThreadUpdatesServlet(hs).register(http_server)

synapse/rest/client/sync.py

@@ -37,6 +37,7 @@ from synapse.events.utils import (
     format_event_raw,
 )
 from synapse.handlers.presence import format_user_presence_state
+from synapse.handlers.relations import RelationsHandler
 from synapse.handlers.sliding_sync import SlidingSyncConfig, SlidingSyncResult
 from synapse.handlers.sync import (
     ArchivedSyncResult,
@@ -1107,6 +1108,7 @@ class SlidingSyncRestServlet(RestServlet):
                 time_now,
                 extensions.threads,
                 self.store,
+                requester,
             )
 
         return serialized_extensions
@@ -1150,6 +1152,7 @@ async def _serialise_threads(
     time_now: int,
     threads: SlidingSyncResult.Extensions.ThreadsExtension,
     store: "DataStore",
+    requester: Requester,
 ) -> JsonDict:
     """
     Serialize the threads extension response for sliding sync.
@@ -1159,6 +1162,7 @@ async def _serialise_threads(
         time_now: The current time in milliseconds, used for event serialization.
         threads: The threads extension data containing thread updates and pagination tokens.
         store: The datastore, needed for serializing stream tokens.
+        requester: The user making the request, used for transaction_id inclusion.
 
     Returns:
         A JSON-serializable dict containing:
@@ -1169,46 +1173,24 @@ async def _serialise_threads(
           - "prev_batch": A pagination token for fetching older events in the thread.
         - "prev_batch": A pagination token for fetching older thread updates (if available).
     """
-    out: JsonDict = {}
+    if not threads.updates:
+        out: JsonDict = {}
+        if threads.prev_batch:
+            out["prev_batch"] = await threads.prev_batch.to_string(store)
+        return out
 
-    if threads.updates:
-        updates_dict: JsonDict = {}
-        for room_id, thread_updates in threads.updates.items():
-            room_updates: JsonDict = {}
-            for thread_root_id, update in thread_updates.items():
-                # Serialize the update
-                update_dict: JsonDict = {}
+    # Create serialization config to include transaction_id for requester's events
+    serialize_options = SerializeEventConfig(requester=requester)
 
-                # Serialize the thread_root event if present
-                if update.thread_root is not None:
-                    # Create a mapping of event_id to bundled_aggregations
-                    bundle_aggs_map = (
-                        {thread_root_id: update.bundled_aggregations}
-                        if update.bundled_aggregations
-                        else None
-                    )
-                    serialized_events = await event_serializer.serialize_events(
-                        [update.thread_root],
-                        time_now,
-                        bundle_aggregations=bundle_aggs_map,
-                    )
-                    if serialized_events:
-                        update_dict["thread_root"] = serialized_events[0]
-
-                # Add prev_batch if present
-                if update.prev_batch is not None:
-                    update_dict["prev_batch"] = await update.prev_batch.to_string(store)
-
-                room_updates[thread_root_id] = update_dict
-
-            updates_dict[room_id] = room_updates
-
-        out["updates"] = updates_dict
-
-    if threads.prev_batch:
-        out["prev_batch"] = await threads.prev_batch.to_string(store)
-
-    return out
+    # Use shared serialization helper (static method)
+    return await RelationsHandler.serialize_thread_updates(
+        thread_updates=threads.updates,
+        prev_batch_token=threads.prev_batch,
+        event_serializer=event_serializer,
+        time_now=time_now,
+        store=store,
+        serialize_options=serialize_options,
+    )
 
 
 def register_servlets(hs: "HomeServer", http_server: HttpServer) -> None:

synapse/streams/config.py

@@ -35,6 +35,32 @@ logger = logging.getLogger(__name__)
 MAX_LIMIT = 1000
 
 
+def extract_stream_token_from_pagination_token(token_str: str) -> str:
+    """
+    Extract the StreamToken portion from a pagination token string.
+
+    Handles both:
+    - StreamToken format: "s123_456_..."
+    - SlidingSyncStreamToken format: "5/s123_456_..." (extracts part after /)
+
+    This allows clients using sliding sync to use their pos tokens
+    with endpoints like /relations and /messages.
+
+    Args:
+        token_str: The token string to parse
+
+    Returns:
+        The StreamToken portion of the token
+    """
+    if "/" in token_str:
+        # SlidingSyncStreamToken format: "connection_position/stream_token"
+        # Split and return just the stream_token part
+        parts = token_str.split("/", 1)
+        if len(parts) == 2:
+            return parts[1]
+    return token_str
+
+
 @attr.s(slots=True, auto_attribs=True)
 class PaginationConfig:
     """A configuration object which stores pagination parameters."""
@@ -57,32 +83,14 @@ class PaginationConfig:
         from_tok_str = parse_string(request, "from")
         to_tok_str = parse_string(request, "to")
 
-        # Helper function to extract StreamToken from either StreamToken or SlidingSyncStreamToken format
-        def extract_stream_token(token_str: str) -> str:
-            """
-            Extract the StreamToken portion from a token string.
-
-            Handles both:
-            - StreamToken format: "s123_456_..."
-            - SlidingSyncStreamToken format: "5/s123_456_..." (extracts part after /)
-
-            This allows clients using sliding sync to use their pos tokens
-            with endpoints like /relations and /messages.
-            """
-            if "/" in token_str:
-                # SlidingSyncStreamToken format: "connection_position/stream_token"
-                # Split and return just the stream_token part
-                parts = token_str.split("/", 1)
-                if len(parts) == 2:
-                    return parts[1]
-            return token_str
-
         try:
             from_tok = None
             if from_tok_str == "END":
                 from_tok = None  # For backwards compat.
             elif from_tok_str:
-                stream_token_str = extract_stream_token(from_tok_str)
+                stream_token_str = extract_stream_token_from_pagination_token(
+                    from_tok_str
+                )
                 from_tok = await StreamToken.from_string(store, stream_token_str)
         except Exception:
             raise SynapseError(400, "'from' parameter is invalid")
@@ -90,7 +98,9 @@ class PaginationConfig:
         try:
             to_tok = None
             if to_tok_str:
-                stream_token_str = extract_stream_token(to_tok_str)
+                stream_token_str = extract_stream_token_from_pagination_token(
+                    to_tok_str
+                )
                 to_tok = await StreamToken.from_string(store, stream_token_str)
         except Exception:
             raise SynapseError(400, "'to' parameter is invalid")

synapse/types/handlers/sliding_sync.py

@@ -37,7 +37,8 @@ from synapse.api.constants import EventTypes
 from synapse.events import EventBase
 
 if TYPE_CHECKING:
-    from synapse.handlers.relations import BundledAggregations
+    from synapse.handlers.relations import BundledAggregations, ThreadUpdate
+
 from synapse.types import (
     DeviceListUpdates,
     JsonDict,
@@ -409,30 +410,7 @@ class SlidingSyncResult:
                     to paginate through older thread updates.
             """
 
-            @attr.s(slots=True, frozen=True, auto_attribs=True)
-            class ThreadUpdate:
-                """Information about a single thread that has new activity.
-
-                Attributes:
-                    thread_root: The thread root event, if requested via include_roots in the
-                        request. This is the event that started the thread.
-                    prev_batch: A pagination token (exclusive) for fetching older events in this
-                        specific thread. Only present if the thread has multiple updates in the
-                        sync window. This token can be used with the /relations endpoint with
-                        dir=b to paginate backwards through the thread's history.
-                    bundled_aggregations: Bundled aggregations for the thread root event,
-                        including the latest_event in the thread (found in
-                        unsigned.m.relations.m.thread). Only present if thread_root is included.
-                """
-
-                thread_root: EventBase | None
-                prev_batch: StreamToken | None
-                bundled_aggregations: "BundledAggregations | None" = None
-
-                def __bool__(self) -> bool:
-                    return bool(self.thread_root) or bool(self.prev_batch)
-
-            updates: Mapping[str, Mapping[str, ThreadUpdate]] | None
+            updates: Mapping[str, Mapping[str, "ThreadUpdate"]] | None
             prev_batch: StreamToken | None
 
             def __bool__(self) -> bool:

tests/rest/client/sliding_sync/test_extension_threads.py

@@ -924,6 +924,250 @@ class SlidingSyncThreadsExtensionTestCase(SlidingSyncBase):
         # Verify the thread root event is present
         self.assertIn("thread_root", thread_updates[thread_root_id])
 
+    def test_thread_updates_initial_sync(self) -> None:
+        """
+        Test that prev_batch from the threads extension response can be used
+        with the /thread_updates endpoint to get additional thread updates during
+        initial sync. This verifies:
+        1. The from parameter boundary is exclusive (no duplicates)
+        2. Using prev_batch as 'from' provides complete coverage (no gaps)
+        3. Works correctly with different numbers of threads
+        """
+        user1_id = self.register_user("user1", "pass")
+        user1_tok = self.login(user1_id, "pass")
+        room_id = self.helper.create_room_as(user1_id, tok=user1_tok)
+
+        # Create 5 thread roots
+        thread_ids = []
+        for i in range(5):
+            thread_root_id = self.helper.send(
+                room_id, body=f"Thread {i}", tok=user1_tok
+            )["event_id"]
+            thread_ids.append(thread_root_id)
+
+            # Add reply to each thread
+            self.helper.send_event(
+                room_id,
+                type="m.room.message",
+                content={
+                    "msgtype": "m.text",
+                    "body": f"Reply to thread {i}",
+                    "m.relates_to": {
+                        "rel_type": RelationTypes.THREAD,
+                        "event_id": thread_root_id,
+                    },
+                },
+                tok=user1_tok,
+            )
+
+        # Do initial sync with threads extension enabled and limit=2
+        sync_body = {
+            "lists": {
+                "all-rooms": {
+                    "ranges": [[0, 10]],
+                    "required_state": [],
+                    "timeline_limit": 0,
+                }
+            },
+            "extensions": {
+                EXT_NAME: {
+                    "enabled": True,
+                    "limit": 2,
+                }
+            },
+        }
+        response_body, _ = self.do_sync(sync_body, tok=user1_tok)
+
+        # Should get 2 thread updates
+        thread_updates = response_body["extensions"][EXT_NAME]["updates"][room_id]
+        self.assertEqual(len(thread_updates), 2)
+        first_sync_threads = set(thread_updates.keys())
+
+        # Get the top-level prev_batch token from the extension
+        self.assertIn("prev_batch", response_body["extensions"][EXT_NAME])
+        prev_batch = response_body["extensions"][EXT_NAME]["prev_batch"]
+
+        # Use prev_batch with /thread_updates endpoint to get remaining updates
+        # Note: prev_batch should be used as 'from' parameter (upper bound for backward pagination)
+        channel = self.make_request(
+            "POST",
+            f"/_matrix/client/unstable/io.element.msc4360/thread_updates?dir=b&from={prev_batch}",
+            access_token=user1_tok,
+            content={},
+        )
+        self.assertEqual(channel.code, 200)
+
+        # Should get the remaining 3 thread updates
+        chunk = channel.json_body["chunk"]
+        self.assertIn(room_id, chunk)
+        self.assertEqual(len(chunk[room_id]), 3)
+
+        thread_updates_response_threads = set(chunk[room_id].keys())
+
+        # Verify no overlap - the from parameter boundary should be exclusive
+        self.assertEqual(
+            len(first_sync_threads & thread_updates_response_threads),
+            0,
+            "from parameter boundary should be exclusive - no thread should appear in both responses",
+        )
+
+        # Verify no gaps - all threads should be accounted for
+        all_threads = set(thread_ids)
+        combined_threads = first_sync_threads | thread_updates_response_threads
+        self.assertEqual(
+            combined_threads,
+            all_threads,
+            "Combined responses should include all thread updates with no gaps",
+        )
+
+    def test_thread_updates_incremental_sync(self) -> None:
+        """
+        Test the intended usage pattern from MSC4360: using prev_batch as 'from'
+        and a previous sync pos as 'to' with /thread_updates to fill gaps between
+        syncs. This verifies that using both bounds together provides complete
+        coverage with no gaps or duplicates.
+        """
+        user1_id = self.register_user("user1", "pass")
+        user1_tok = self.login(user1_id, "pass")
+        room_id = self.helper.create_room_as(user1_id, tok=user1_tok)
+
+        # Create 3 threads initially
+        initial_thread_ids = []
+        for i in range(3):
+            thread_root_id = self.helper.send(
+                room_id, body=f"Thread {i}", tok=user1_tok
+            )["event_id"]
+            initial_thread_ids.append(thread_root_id)
+
+            self.helper.send_event(
+                room_id,
+                type="m.room.message",
+                content={
+                    "msgtype": "m.text",
+                    "body": f"Reply to thread {i}",
+                    "m.relates_to": {
+                        "rel_type": RelationTypes.THREAD,
+                        "event_id": thread_root_id,
+                    },
+                },
+                tok=user1_tok,
+            )
+
+        # First sync
+        sync_body = {
+            "lists": {
+                "all-rooms": {
+                    "ranges": [[0, 10]],
+                    "required_state": [],
+                    "timeline_limit": 0,
+                }
+            },
+            "extensions": {
+                EXT_NAME: {
+                    "enabled": True,
+                }
+            },
+        }
+        response_body, pos1 = self.do_sync(sync_body, tok=user1_tok)
+
+        # Should get 3 thread updates
+        first_sync_threads = set(
+            response_body["extensions"][EXT_NAME]["updates"][room_id].keys()
+        )
+        self.assertEqual(len(first_sync_threads), 3)
+
+        # Create 3 more threads after the first sync
+        new_thread_ids = []
+        for i in range(3, 6):
+            thread_root_id = self.helper.send(
+                room_id, body=f"Thread {i}", tok=user1_tok
+            )["event_id"]
+            new_thread_ids.append(thread_root_id)
+
+            self.helper.send_event(
+                room_id,
+                type="m.room.message",
+                content={
+                    "msgtype": "m.text",
+                    "body": f"Reply to thread {i}",
+                    "m.relates_to": {
+                        "rel_type": RelationTypes.THREAD,
+                        "event_id": thread_root_id,
+                    },
+                },
+                tok=user1_tok,
+            )
+
+        # Second sync with limit=1 to get only some of the new threads
+        sync_body_with_limit = {
+            "lists": {
+                "all-rooms": {
+                    "ranges": [[0, 10]],
+                    "required_state": [],
+                    "timeline_limit": 0,
+                }
+            },
+            "extensions": {
+                EXT_NAME: {
+                    "enabled": True,
+                    "limit": 1,
+                }
+            },
+        }
+        response_body, pos2 = self.do_sync(
+            sync_body_with_limit, tok=user1_tok, since=pos1
+        )
+
+        # Should get 1 thread update
+        second_sync_threads = set(
+            response_body["extensions"][EXT_NAME]["updates"][room_id].keys()
+        )
+        self.assertEqual(len(second_sync_threads), 1)
+
+        # Get prev_batch from the extension
+        self.assertIn("prev_batch", response_body["extensions"][EXT_NAME])
+        prev_batch = response_body["extensions"][EXT_NAME]["prev_batch"]
+
+        # Now use /thread_updates with from=prev_batch and to=pos1
+        # This should get the 2 remaining new threads (created after pos1, not returned in second sync)
+        channel = self.make_request(
+            "POST",
+            f"/_matrix/client/unstable/io.element.msc4360/thread_updates?dir=b&from={prev_batch}&to={pos1}",
+            access_token=user1_tok,
+            content={},
+        )
+        self.assertEqual(channel.code, 200)
+
+        chunk = channel.json_body["chunk"]
+        self.assertIn(room_id, chunk)
+        thread_updates_threads = set(chunk[room_id].keys())
+
+        # Should get exactly 2 threads
+        self.assertEqual(len(thread_updates_threads), 2)
+
+        # Verify no overlap with second sync
+        self.assertEqual(
+            len(second_sync_threads & thread_updates_threads),
+            0,
+            "No thread should appear in both second sync and thread_updates responses",
+        )
+
+        # Verify no overlap with first sync (to=pos1 should exclude those)
+        self.assertEqual(
+            len(first_sync_threads & thread_updates_threads),
+            0,
+            "Threads from first sync should not appear in thread_updates (to=pos1 excludes them)",
+        )
+
+        # Verify no gaps - all new threads should be accounted for
+        all_new_threads = set(new_thread_ids)
+        combined_new_threads = second_sync_threads | thread_updates_threads
+        self.assertEqual(
+            combined_new_threads,
+            all_new_threads,
+            "Combined responses should include all new thread updates with no gaps",
+        )
+
     def test_threads_only_from_rooms_in_list(self) -> None:
         """
         Test that thread updates are only returned for rooms that are in the

tests/rest/client/test_thread_updates.py

@@ -0,0 +1,957 @@
+#
+# This file is licensed under the Affero General Public License (AGPL) version 3.
+#
+# Copyright (C) 2025 New Vector, Ltd
+#
+# This program is free software: you can redistribute it and/or modify
+# it under the terms of the GNU Affero General Public License as
+# published by the Free Software Foundation, either version 3 of the
+# License, or (at your option) any later version.
+#
+# See the GNU Affero General Public License for more details:
+# <https://www.gnu.org/licenses/agpl-3.0.html>.
+#
+import logging
+
+from twisted.test.proto_helpers import MemoryReactor
+
+import synapse.rest.admin
+from synapse.api.constants import RelationTypes
+from synapse.rest.client import login, relations, room
+from synapse.server import HomeServer
+from synapse.types import JsonDict
+from synapse.util.clock import Clock
+
+from tests import unittest
+
+logger = logging.getLogger(__name__)
+
+
+class ThreadUpdatesTestCase(unittest.HomeserverTestCase):
+    """
+    Test the /thread_updates companion endpoint (MSC4360).
+    """
+
+    servlets = [
+        synapse.rest.admin.register_servlets,
+        login.register_servlets,
+        room.register_servlets,
+        relations.register_servlets,
+    ]
+
+    def default_config(self) -> JsonDict:
+        config = super().default_config()
+        config["experimental_features"] = {"msc4360_enabled": True}
+        return config
+
+    def prepare(self, reactor: MemoryReactor, clock: Clock, hs: HomeServer) -> None:
+        self.store = hs.get_datastores().main
+
+    def test_no_updates_for_new_user(self) -> None:
+        """
+        Test that a user with no thread updates gets an empty response.
+        """
+        user1_id = self.register_user("user1", "pass")
+        user1_tok = self.login(user1_id, "pass")
+
+        # Request thread updates
+        channel = self.make_request(
+            "POST",
+            "/_matrix/client/unstable/io.element.msc4360/thread_updates?dir=b",
+            content={"include_roots": True},
+            access_token=user1_tok,
+        )
+        self.assertEqual(channel.code, 200, channel.json_body)
+
+        # Assert empty chunk and no next_batch
+        self.assertEqual(channel.json_body["chunk"], {})
+        self.assertNotIn("next_batch", channel.json_body)
+
+    def test_single_thread_update(self) -> None:
+        """
+        Test that a single thread with one reply appears in the response.
+        """
+        user1_id = self.register_user("user1", "pass")
+        user1_tok = self.login(user1_id, "pass")
+        room_id = self.helper.create_room_as(user1_id, tok=user1_tok)
+
+        # Create thread root
+        thread_root_resp = self.helper.send(room_id, body="Thread root", tok=user1_tok)
+        thread_root_id = thread_root_resp["event_id"]
+
+        # Add reply to thread
+        self.helper.send_event(
+            room_id,
+            type="m.room.message",
+            content={
+                "msgtype": "m.text",
+                "body": "Reply 1",
+                "m.relates_to": {
+                    "rel_type": RelationTypes.THREAD,
+                    "event_id": thread_root_id,
+                },
+            },
+            tok=user1_tok,
+        )
+
+        # Request thread updates
+        channel = self.make_request(
+            "POST",
+            "/_matrix/client/unstable/io.element.msc4360/thread_updates?dir=b",
+            access_token=user1_tok,
+            content={"include_roots": True},
+        )
+        self.assertEqual(channel.code, 200, channel.json_body)
+
+        # Assert thread is present
+        chunk = channel.json_body["chunk"]
+        self.assertIn(room_id, chunk)
+        self.assertIn(thread_root_id, chunk[room_id])
+
+        # Assert thread root is included
+        thread_update = chunk[room_id][thread_root_id]
+        self.assertIn("thread_root", thread_update)
+        self.assertEqual(thread_update["thread_root"]["event_id"], thread_root_id)
+
+        # Assert prev_batch is NOT present (only 1 update - the reply)
+        self.assertNotIn("prev_batch", thread_update)
+
+    def test_multiple_threads_single_room(self) -> None:
+        """
+        Test that multiple threads in the same room are grouped correctly.
+        """
+        user1_id = self.register_user("user1", "pass")
+        user1_tok = self.login(user1_id, "pass")
+        room_id = self.helper.create_room_as(user1_id, tok=user1_tok)
+
+        # Create two threads
+        thread1_root_id = self.helper.send(room_id, body="Thread 1", tok=user1_tok)[
+            "event_id"
+        ]
+        thread2_root_id = self.helper.send(room_id, body="Thread 2", tok=user1_tok)[
+            "event_id"
+        ]
+
+        # Add replies to both threads
+        self.helper.send_event(
+            room_id,
+            type="m.room.message",
+            content={
+                "msgtype": "m.text",
+                "body": "Reply to thread 1",
+                "m.relates_to": {
+                    "rel_type": RelationTypes.THREAD,
+                    "event_id": thread1_root_id,
+                },
+            },
+            tok=user1_tok,
+        )
+        self.helper.send_event(
+            room_id,
+            type="m.room.message",
+            content={
+                "msgtype": "m.text",
+                "body": "Reply to thread 2",
+                "m.relates_to": {
+                    "rel_type": RelationTypes.THREAD,
+                    "event_id": thread2_root_id,
+                },
+            },
+            tok=user1_tok,
+        )
+
+        # Request thread updates
+        channel = self.make_request(
+            "POST",
+            "/_matrix/client/unstable/io.element.msc4360/thread_updates?dir=b",
+            access_token=user1_tok,
+            content={"include_roots": True},
+        )
+        self.assertEqual(channel.code, 200, channel.json_body)
+
+        # Assert both threads are in the same room
+        chunk = channel.json_body["chunk"]
+        self.assertIn(room_id, chunk)
+        self.assertEqual(len(chunk), 1, "Should only have one room")
+        self.assertEqual(len(chunk[room_id]), 2, "Should have two threads")
+        self.assertIn(thread1_root_id, chunk[room_id])
+        self.assertIn(thread2_root_id, chunk[room_id])
+
+    def test_threads_across_multiple_rooms(self) -> None:
+        """
+        Test that threads from different rooms are grouped by room_id.
+        """
+        user1_id = self.register_user("user1", "pass")
+        user1_tok = self.login(user1_id, "pass")
+        room_a_id = self.helper.create_room_as(user1_id, tok=user1_tok)
+        room_b_id = self.helper.create_room_as(user1_id, tok=user1_tok)
+
+        # Create threads in both rooms
+        thread_a_root_id = self.helper.send(room_a_id, body="Thread A", tok=user1_tok)[
+            "event_id"
+        ]
+        thread_b_root_id = self.helper.send(room_b_id, body="Thread B", tok=user1_tok)[
+            "event_id"
+        ]
+
+        # Add replies
+        self.helper.send_event(
+            room_a_id,
+            type="m.room.message",
+            content={
+                "msgtype": "m.text",
+                "body": "Reply to A",
+                "m.relates_to": {
+                    "rel_type": RelationTypes.THREAD,
+                    "event_id": thread_a_root_id,
+                },
+            },
+            tok=user1_tok,
+        )
+        self.helper.send_event(
+            room_b_id,
+            type="m.room.message",
+            content={
+                "msgtype": "m.text",
+                "body": "Reply to B",
+                "m.relates_to": {
+                    "rel_type": RelationTypes.THREAD,
+                    "event_id": thread_b_root_id,
+                },
+            },
+            tok=user1_tok,
+        )
+
+        # Request thread updates
+        channel = self.make_request(
+            "POST",
+            "/_matrix/client/unstable/io.element.msc4360/thread_updates?dir=b",
+            access_token=user1_tok,
+            content={"include_roots": True},
+        )
+        self.assertEqual(channel.code, 200, channel.json_body)
+
+        # Assert both rooms are present with their threads
+        chunk = channel.json_body["chunk"]
+        self.assertEqual(len(chunk), 2, "Should have two rooms")
+        self.assertIn(room_a_id, chunk)
+        self.assertIn(room_b_id, chunk)
+        self.assertIn(thread_a_root_id, chunk[room_a_id])
+        self.assertIn(thread_b_root_id, chunk[room_b_id])
+
+    def test_pagination_with_from_token(self) -> None:
+        """
+        Test that pagination works using the next_batch token.
+        This verifies that multiple calls to /thread_updates return all thread
+        updates with no duplicates and no gaps.
+        """
+        user1_id = self.register_user("user1", "pass")
+        user1_tok = self.login(user1_id, "pass")
+        room_id = self.helper.create_room_as(user1_id, tok=user1_tok)
+
+        # Create many threads (more than default limit)
+        thread_ids = []
+        for i in range(5):
+            thread_root_id = self.helper.send(
+                room_id, body=f"Thread {i}", tok=user1_tok
+            )["event_id"]
+            thread_ids.append(thread_root_id)
+
+            # Add reply
+            self.helper.send_event(
+                room_id,
+                type="m.room.message",
+                content={
+                    "msgtype": "m.text",
+                    "body": f"Reply to thread {i}",
+                    "m.relates_to": {
+                        "rel_type": RelationTypes.THREAD,
+                        "event_id": thread_root_id,
+                    },
+                },
+                tok=user1_tok,
+            )
+
+        # Request first page with small limit
+        channel = self.make_request(
+            "POST",
+            "/_matrix/client/unstable/io.element.msc4360/thread_updates?dir=b&limit=2",
+            access_token=user1_tok,
+            content={"include_roots": True},
+        )
+        self.assertEqual(channel.code, 200, channel.json_body)
+
+        # Should have 2 threads and a next_batch token
+        first_page_threads = set(channel.json_body["chunk"][room_id].keys())
+        self.assertEqual(len(first_page_threads), 2)
+        self.assertIn("next_batch", channel.json_body)
+
+        next_batch = channel.json_body["next_batch"]
+
+        # Request second page
+        channel = self.make_request(
+            "POST",
+            f"/_matrix/client/unstable/io.element.msc4360/thread_updates?dir=b&limit=2&from={next_batch}",
+            access_token=user1_tok,
+            content={"include_roots": True},
+        )
+        self.assertEqual(channel.code, 200, channel.json_body)
+
+        second_page_threads = set(channel.json_body["chunk"][room_id].keys())
+        self.assertEqual(len(second_page_threads), 2)
+
+        # Verify no overlap
+        self.assertEqual(
+            len(first_page_threads & second_page_threads),
+            0,
+            "Pages should not have overlapping threads",
+        )
+
+        # Request third page to get the remaining thread
+        self.assertIn("next_batch", channel.json_body)
+        next_batch_2 = channel.json_body["next_batch"]
+
+        channel = self.make_request(
+            "POST",
+            f"/_matrix/client/unstable/io.element.msc4360/thread_updates?dir=b&limit=2&from={next_batch_2}",
+            access_token=user1_tok,
+            content={"include_roots": True},
+        )
+        self.assertEqual(channel.code, 200, channel.json_body)
+
+        third_page_threads = set(channel.json_body["chunk"][room_id].keys())
+        self.assertEqual(len(third_page_threads), 1)
+
+        # Verify no overlap between any pages
+        self.assertEqual(len(first_page_threads & third_page_threads), 0)
+        self.assertEqual(len(second_page_threads & third_page_threads), 0)
+
+        # Verify no gaps - all threads should be accounted for across all pages
+        all_threads = set(thread_ids)
+        combined_threads = first_page_threads | second_page_threads | third_page_threads
+        self.assertEqual(
+            combined_threads,
+            all_threads,
+            "Combined pages should include all thread updates with no gaps",
+        )
+
+    def test_invalid_dir_parameter(self) -> None:
+        """
+        Test that forward pagination (dir=f) is rejected with an error.
+        """
+        user1_id = self.register_user("user1", "pass")
+        user1_tok = self.login(user1_id, "pass")
+
+        # Request with forward direction should fail
+        channel = self.make_request(
+            "POST",
+            "/_matrix/client/unstable/io.element.msc4360/thread_updates?dir=f",
+            access_token=user1_tok,
+            content={"include_roots": True},
+        )
+        self.assertEqual(channel.code, 400)
+
+    def test_invalid_limit_parameter(self) -> None:
+        """
+        Test that invalid limit values are rejected.
+        """
+        user1_id = self.register_user("user1", "pass")
+        user1_tok = self.login(user1_id, "pass")
+
+        # Zero limit should fail
+        channel = self.make_request(
+            "POST",
+            "/_matrix/client/unstable/io.element.msc4360/thread_updates?dir=b&limit=0",
+            access_token=user1_tok,
+            content={"include_roots": True},
+        )
+        self.assertEqual(channel.code, 400)
+
+        # Negative limit should fail
+        channel = self.make_request(
+            "POST",
+            "/_matrix/client/unstable/io.element.msc4360/thread_updates?dir=b&limit=-5",
+            access_token=user1_tok,
+            content={"include_roots": True},
+        )
+        self.assertEqual(channel.code, 400)
+
+    def test_invalid_pagination_tokens(self) -> None:
+        """
+        Test that invalid from/to tokens are rejected with appropriate errors.
+        """
+        user1_id = self.register_user("user1", "pass")
+        user1_tok = self.login(user1_id, "pass")
+
+        # Invalid from token
+        channel = self.make_request(
+            "POST",
+            "/_matrix/client/unstable/io.element.msc4360/thread_updates?dir=b&from=invalid_token",
+            access_token=user1_tok,
+            content={"include_roots": True},
+        )
+        self.assertEqual(channel.code, 400)
+
+        # Invalid to token
+        channel = self.make_request(
+            "POST",
+            "/_matrix/client/unstable/io.element.msc4360/thread_updates?dir=b&to=invalid_token",
+            access_token=user1_tok,
+            content={"include_roots": True},
+        )
+        self.assertEqual(channel.code, 400)
+
+    def test_to_token_filtering(self) -> None:
+        """
+        Test that the to_token parameter correctly limits pagination to updates
+        newer than the to_token (since we paginate backwards from newest to oldest).
+        This also verifies the to_token boundary is exclusive - updates at exactly
+        the to_token position should not be included (as they were already returned
+        in a previous response that synced up to that position).
+        """
+        user1_id = self.register_user("user1", "pass")
+        user1_tok = self.login(user1_id, "pass")
+        room_id = self.helper.create_room_as(user1_id, tok=user1_tok)
+
+        # Create two thread roots
+        thread1_root_id = self.helper.send(room_id, body="Thread 1", tok=user1_tok)[
+            "event_id"
+        ]
+        thread2_root_id = self.helper.send(room_id, body="Thread 2", tok=user1_tok)[
+            "event_id"
+        ]
+
+        # Send replies to both threads
+        self.helper.send_event(
+            room_id,
+            type="m.room.message",
+            content={
+                "msgtype": "m.text",
+                "body": "Reply to thread 1",
+                "m.relates_to": {
+                    "rel_type": RelationTypes.THREAD,
+                    "event_id": thread1_root_id,
+                },
+            },
+            tok=user1_tok,
+        )
+        self.helper.send_event(
+            room_id,
+            type="m.room.message",
+            content={
+                "msgtype": "m.text",
+                "body": "Reply to thread 2",
+                "m.relates_to": {
+                    "rel_type": RelationTypes.THREAD,
+                    "event_id": thread2_root_id,
+                },
+            },
+            tok=user1_tok,
+        )
+
+        # Request with limit=1 to get only the latest thread update
+        channel = self.make_request(
+            "POST",
+            "/_matrix/client/unstable/io.element.msc4360/thread_updates?dir=b&limit=1",
+            access_token=user1_tok,
+            content={"include_roots": True},
+        )
+        self.assertEqual(channel.code, 200)
+        self.assertIn("next_batch", channel.json_body)
+
+        # next_batch points to before the update we just received
+        next_batch = channel.json_body["next_batch"]
+        first_response_threads = set(channel.json_body["chunk"][room_id].keys())
+
+        # Request again with to=next_batch (lower bound for backward pagination) and no
+        # limit.
+        # This should get only the same thread updates as before, not the additional
+        # update.
+        channel = self.make_request(
+            "POST",
+            f"/_matrix/client/unstable/io.element.msc4360/thread_updates?dir=b&to={next_batch}",
+            access_token=user1_tok,
+            content={"include_roots": True},
+        )
+        self.assertEqual(channel.code, 200)
+
+        chunk = channel.json_body["chunk"]
+        self.assertIn(room_id, chunk)
+        # Should have exactly one thread update
+        self.assertEqual(len(chunk[room_id]), 1)
+
+        second_response_threads = set(chunk[room_id].keys())
+
+        # Verify no overlap - the from parameter boundary should be exclusive
+        self.assertEqual(
+            first_response_threads,
+            second_response_threads,
+            "to parameter boundary should be exclusive - both responses should be identical",
+        )
+
+    def test_bundled_aggregations_on_thread_roots(self) -> None:
+        """
+        Test that thread root events include bundled aggregations with latest thread event.
+        """
+        user1_id = self.register_user("user1", "pass")
+        user1_tok = self.login(user1_id, "pass")
+        room_id = self.helper.create_room_as(user1_id, tok=user1_tok)
+
+        # Create thread root
+        thread_root_id = self.helper.send(room_id, body="Thread root", tok=user1_tok)[
+            "event_id"
+        ]
+
+        # Send replies to create bundled aggregation data
+        for i in range(2):
+            self.helper.send_event(
+                room_id,
+                type="m.room.message",
+                content={
+                    "msgtype": "m.text",
+                    "body": f"Reply {i + 1}",
+                    "m.relates_to": {
+                        "rel_type": RelationTypes.THREAD,
+                        "event_id": thread_root_id,
+                    },
+                },
+                tok=user1_tok,
+            )
+
+        # Request thread updates
+        channel = self.make_request(
+            "POST",
+            "/_matrix/client/unstable/io.element.msc4360/thread_updates?dir=b",
+            access_token=user1_tok,
+            content={"include_roots": True},
+        )
+        self.assertEqual(channel.code, 200)
+
+        # Check that thread root has bundled aggregations with latest event
+        chunk = channel.json_body["chunk"]
+        thread_update = chunk[room_id][thread_root_id]
+        thread_root_event = thread_update["thread_root"]
+
+        # Should have unsigned data with latest thread event content
+        self.assertIn("unsigned", thread_root_event)
+        self.assertIn("m.relations", thread_root_event["unsigned"])
+        relations = thread_root_event["unsigned"]["m.relations"]
+        self.assertIn(RelationTypes.THREAD, relations)
+
+        # Check latest event is present in bundled aggregations
+        thread_summary = relations[RelationTypes.THREAD]
+        self.assertIn("latest_event", thread_summary)
+        latest_event = thread_summary["latest_event"]
+        self.assertEqual(latest_event["content"]["body"], "Reply 2")
+
+    def test_only_joined_rooms(self) -> None:
+        """
+        Test that thread updates only include rooms where the user is currently joined.
+        """
+        user1_id = self.register_user("user1", "pass")
+        user1_tok = self.login(user1_id, "pass")
+        user2_id = self.register_user("user2", "pass")
+        user2_tok = self.login(user2_id, "pass")
+
+        # Create two rooms, user1 joins both
+        room1_id = self.helper.create_room_as(user1_id, tok=user1_tok)
+        room2_id = self.helper.create_room_as(user2_id, tok=user2_tok)
+        self.helper.join(room2_id, user1_id, tok=user1_tok)
+
+        # Create threads in both rooms
+        thread1_root_id = self.helper.send(room1_id, body="Thread 1", tok=user1_tok)[
+            "event_id"
+        ]
+        thread2_root_id = self.helper.send(room2_id, body="Thread 2", tok=user2_tok)[
+            "event_id"
+        ]
+
+        # Add replies to both threads
+        self.helper.send_event(
+            room1_id,
+            type="m.room.message",
+            content={
+                "msgtype": "m.text",
+                "body": "Reply to thread 1",
+                "m.relates_to": {
+                    "rel_type": RelationTypes.THREAD,
+                    "event_id": thread1_root_id,
+                },
+            },
+            tok=user1_tok,
+        )
+        self.helper.send_event(
+            room2_id,
+            type="m.room.message",
+            content={
+                "msgtype": "m.text",
+                "body": "Reply to thread 2",
+                "m.relates_to": {
+                    "rel_type": RelationTypes.THREAD,
+                    "event_id": thread2_root_id,
+                },
+            },
+            tok=user2_tok,
+        )
+
+        # User1 leaves room2
+        self.helper.leave(room2_id, user1_id, tok=user1_tok)
+
+        # Request thread updates for user1 - should only get room1
+        channel = self.make_request(
+            "POST",
+            "/_matrix/client/unstable/io.element.msc4360/thread_updates?dir=b",
+            access_token=user1_tok,
+            content={"include_roots": True},
+        )
+        self.assertEqual(channel.code, 200)
+
+        chunk = channel.json_body["chunk"]
+        # Should only have room1, not room2
+        self.assertIn(room1_id, chunk)
+        self.assertNotIn(room2_id, chunk)
+        self.assertIn(thread1_root_id, chunk[room1_id])
+
+    def test_room_filtering_with_lists(self) -> None:
+        """
+        Test that room filtering works correctly using the lists parameter.
+        This verifies that thread updates are only returned for rooms matching
+        the provided filters.
+        """
+        user1_id = self.register_user("user1", "pass")
+        user1_tok = self.login(user1_id, "pass")
+
+        # Create an encrypted room and an unencrypted room
+        encrypted_room_id = self.helper.create_room_as(
+            user1_id,
+            tok=user1_tok,
+            extra_content={
+                "initial_state": [
+                    {
+                        "type": "m.room.encryption",
+                        "state_key": "",
+                        "content": {"algorithm": "m.megolm.v1.aes-sha2"},
+                    }
+                ]
+            },
+        )
+        unencrypted_room_id = self.helper.create_room_as(user1_id, tok=user1_tok)
+
+        # Create threads in both rooms
+        enc_thread_root_id = self.helper.send(
+            encrypted_room_id, body="Encrypted thread", tok=user1_tok
+        )["event_id"]
+        unenc_thread_root_id = self.helper.send(
+            unencrypted_room_id, body="Unencrypted thread", tok=user1_tok
+        )["event_id"]
+
+        # Add replies to both threads
+        self.helper.send_event(
+            encrypted_room_id,
+            type="m.room.message",
+            content={
+                "msgtype": "m.text",
+                "body": "Reply in encrypted room",
+                "m.relates_to": {
+                    "rel_type": RelationTypes.THREAD,
+                    "event_id": enc_thread_root_id,
+                },
+            },
+            tok=user1_tok,
+        )
+        self.helper.send_event(
+            unencrypted_room_id,
+            type="m.room.message",
+            content={
+                "msgtype": "m.text",
+                "body": "Reply in unencrypted room",
+                "m.relates_to": {
+                    "rel_type": RelationTypes.THREAD,
+                    "event_id": unenc_thread_root_id,
+                },
+            },
+            tok=user1_tok,
+        )
+
+        # Request thread updates with filter for encrypted rooms only
+        channel = self.make_request(
+            "POST",
+            "/_matrix/client/unstable/io.element.msc4360/thread_updates?dir=b",
+            access_token=user1_tok,
+            content={
+                "lists": {
+                    "encrypted_list": {
+                        "ranges": [[0, 99]],
+                        "required_state": [["m.room.encryption", ""]],
+                        "timeline_limit": 10,
+                        "filters": {"is_encrypted": True},
+                    }
+                }
+            },
+        )
+        self.assertEqual(channel.code, 200, channel.json_body)
+
+        chunk = channel.json_body["chunk"]
+        # Should only include the encrypted room
+        self.assertIn(encrypted_room_id, chunk)
+        self.assertNotIn(unencrypted_room_id, chunk)
+        self.assertIn(enc_thread_root_id, chunk[encrypted_room_id])
+
+    def test_room_filtering_with_room_subscriptions(self) -> None:
+        """
+        Test that room filtering works correctly using the room_subscriptions parameter.
+        This verifies that thread updates are only returned for explicitly subscribed rooms.
+        """
+        user1_id = self.register_user("user1", "pass")
+        user1_tok = self.login(user1_id, "pass")
+
+        # Create three rooms
+        room1_id = self.helper.create_room_as(user1_id, tok=user1_tok)
+        room2_id = self.helper.create_room_as(user1_id, tok=user1_tok)
+        room3_id = self.helper.create_room_as(user1_id, tok=user1_tok)
+
+        # Create threads in all three rooms
+        thread1_root_id = self.helper.send(room1_id, body="Thread 1", tok=user1_tok)[
+            "event_id"
+        ]
+        thread2_root_id = self.helper.send(room2_id, body="Thread 2", tok=user1_tok)[
+            "event_id"
+        ]
+        thread3_root_id = self.helper.send(room3_id, body="Thread 3", tok=user1_tok)[
+            "event_id"
+        ]
+
+        # Add replies to all threads
+        for room_id, thread_root_id in [
+            (room1_id, thread1_root_id),
+            (room2_id, thread2_root_id),
+            (room3_id, thread3_root_id),
+        ]:
+            self.helper.send_event(
+                room_id,
+                type="m.room.message",
+                content={
+                    "msgtype": "m.text",
+                    "body": "Reply",
+                    "m.relates_to": {
+                        "rel_type": RelationTypes.THREAD,
+                        "event_id": thread_root_id,
+                    },
+                },
+                tok=user1_tok,
+            )
+
+        # Request thread updates with subscription to only room1 and room2
+        channel = self.make_request(
+            "POST",
+            "/_matrix/client/unstable/io.element.msc4360/thread_updates?dir=b",
+            access_token=user1_tok,
+            content={
+                "room_subscriptions": {
+                    room1_id: {
+                        "required_state": [["m.room.name", ""]],
+                        "timeline_limit": 10,
+                    },
+                    room2_id: {
+                        "required_state": [["m.room.name", ""]],
+                        "timeline_limit": 10,
+                    },
+                }
+            },
+        )
+        self.assertEqual(channel.code, 200, channel.json_body)
+
+        chunk = channel.json_body["chunk"]
+        # Should only include room1 and room2, not room3
+        self.assertIn(room1_id, chunk)
+        self.assertIn(room2_id, chunk)
+        self.assertNotIn(room3_id, chunk)
+        self.assertIn(thread1_root_id, chunk[room1_id])
+        self.assertIn(thread2_root_id, chunk[room2_id])
+
+    def test_room_filtering_with_lists_and_room_subscriptions(self) -> None:
+        """
+        Test that room filtering works correctly when both lists and room_subscriptions
+        are provided. The union of rooms from both should be included.
+        """
+        user1_id = self.register_user("user1", "pass")
+        user1_tok = self.login(user1_id, "pass")
+
+        # Create an encrypted room and two unencrypted rooms
+        encrypted_room_id = self.helper.create_room_as(
+            user1_id,
+            tok=user1_tok,
+            extra_content={
+                "initial_state": [
+                    {
+                        "type": "m.room.encryption",
+                        "state_key": "",
+                        "content": {"algorithm": "m.megolm.v1.aes-sha2"},
+                    }
+                ]
+            },
+        )
+        unencrypted_room1_id = self.helper.create_room_as(user1_id, tok=user1_tok)
+        unencrypted_room2_id = self.helper.create_room_as(user1_id, tok=user1_tok)
+
+        # Create threads in all three rooms
+        enc_thread_root_id = self.helper.send(
+            encrypted_room_id, body="Encrypted thread", tok=user1_tok
+        )["event_id"]
+        unenc1_thread_root_id = self.helper.send(
+            unencrypted_room1_id, body="Unencrypted thread 1", tok=user1_tok
+        )["event_id"]
+        unenc2_thread_root_id = self.helper.send(
+            unencrypted_room2_id, body="Unencrypted thread 2", tok=user1_tok
+        )["event_id"]
+
+        # Add replies to all threads
+        for room_id, thread_root_id in [
+            (encrypted_room_id, enc_thread_root_id),
+            (unencrypted_room1_id, unenc1_thread_root_id),
+            (unencrypted_room2_id, unenc2_thread_root_id),
+        ]:
+            self.helper.send_event(
+                room_id,
+                type="m.room.message",
+                content={
+                    "msgtype": "m.text",
+                    "body": "Reply",
+                    "m.relates_to": {
+                        "rel_type": RelationTypes.THREAD,
+                        "event_id": thread_root_id,
+                    },
+                },
+                tok=user1_tok,
+            )
+
+        # Request thread updates with:
+        # - lists: filter for encrypted rooms
+        # - room_subscriptions: explicitly subscribe to unencrypted_room1_id
+        # Expected: should get both encrypted_room_id (from list) and unencrypted_room1_id
+        # (from subscription), but NOT unencrypted_room2_id
+        channel = self.make_request(
+            "POST",
+            "/_matrix/client/unstable/io.element.msc4360/thread_updates?dir=b",
+            access_token=user1_tok,
+            content={
+                "lists": {
+                    "encrypted_list": {
+                        "ranges": [[0, 99]],
+                        "required_state": [["m.room.encryption", ""]],
+                        "timeline_limit": 10,
+                        "filters": {"is_encrypted": True},
+                    }
+                },
+                "room_subscriptions": {
+                    unencrypted_room1_id: {
+                        "required_state": [["m.room.name", ""]],
+                        "timeline_limit": 10,
+                    }
+                },
+            },
+        )
+        self.assertEqual(channel.code, 200, channel.json_body)
+
+        chunk = channel.json_body["chunk"]
+        # Should include encrypted_room_id (from list filter) and unencrypted_room1_id
+        # (from subscription), but not unencrypted_room2_id
+        self.assertIn(encrypted_room_id, chunk)
+        self.assertIn(unencrypted_room1_id, chunk)
+        self.assertNotIn(unencrypted_room2_id, chunk)
+        self.assertIn(enc_thread_root_id, chunk[encrypted_room_id])
+        self.assertIn(unenc1_thread_root_id, chunk[unencrypted_room1_id])
+
+    def test_threads_not_returned_after_leaving_room(self) -> None:
+        """
+        Test that thread updates are properly bounded when a user leaves a room.
+
+        Users should see thread updates that occurred up to the point they left,
+        but NOT updates that occurred after they left.
+        """
+        user1_id = self.register_user("user1", "pass")
+        user1_tok = self.login(user1_id, "pass")
+        user2_id = self.register_user("user2", "pass")
+        user2_tok = self.login(user2_id, "pass")
+
+        # Create room and both users join
+        room_id = self.helper.create_room_as(user1_id, tok=user1_tok)
+        self.helper.join(room_id, user2_id, tok=user2_tok)
+
+        # Create thread
+        res = self.helper.send(room_id, body="Thread root", tok=user1_tok)
+        thread_root = res["event_id"]
+
+        # Reply in thread while user2 is joined
+        self.helper.send_event(
+            room_id,
+            type="m.room.message",
+            content={
+                "msgtype": "m.text",
+                "body": "Reply 1 while user2 joined",
+                "m.relates_to": {
+                    "rel_type": RelationTypes.THREAD,
+                    "event_id": thread_root,
+                },
+            },
+            tok=user1_tok,
+        )
+
+        # User2 leaves the room
+        self.helper.leave(room_id, user2_id, tok=user2_tok)
+
+        # Another reply after user2 left
+        self.helper.send_event(
+            room_id,
+            type="m.room.message",
+            content={
+                "msgtype": "m.text",
+                "body": "Reply 2 after user2 left",
+                "m.relates_to": {
+                    "rel_type": RelationTypes.THREAD,
+                    "event_id": thread_root,
+                },
+            },
+            tok=user1_tok,
+        )
+
+        # User2 gets thread updates with an explicit room subscription
+        # (We need to explicitly subscribe to the room to include it after leaving;
+        # otherwise only joined rooms are returned)
+        channel = self.make_request(
+            "POST",
+            "/_matrix/client/unstable/io.element.msc4360/thread_updates?dir=b&limit=100",
+            {
+                "room_subscriptions": {
+                    room_id: {
+                        "required_state": [],
+                        "timeline_limit": 0,
+                    }
+                }
+            },
+            access_token=user2_tok,
+        )
+        self.assertEqual(channel.code, 200, channel.json_body)
+
+        # Assert: User2 SHOULD see Reply 1 (happened while joined) but NOT Reply 2 (after leaving)
+        chunk = channel.json_body["chunk"]
+        self.assertIn(
+            room_id,
+            chunk,
+            "Thread updates should include the room user2 left",
+        )
+        self.assertIn(
+            thread_root,
+            chunk[room_id],
+            "Thread root should be in the updates",
+        )
+
+        # Verify that only a single update was seen (Reply 1) by checking that there's
+        # no prev_batch token. If Reply 2 was also included, there would be multiple
+        # updates and a prev_batch token would be present.
+        thread_update = chunk[room_id][thread_root]
+        self.assertNotIn(
+            "prev_batch",
+            thread_update,
+            "No prev_batch should be present since only one update (Reply 1) is visible",
+        )