fix(cloud): serialize directory catch-up (#2500)

Co-authored-by: dadachann <185672915+dadachann@users.noreply.github.com>

src/langbot/pkg/api/http/controller/groups/user.py

@@ -453,13 +453,28 @@ class UserRouterGroup(group.RouterGroup):
                         )
                         break
                     except WorkspaceNotFoundError:
-                        if projection_service is None or attempt == 3:
+                        if projection_service is None:
                             raise
-                elif projection_service is None or attempt == 3:
+                elif projection_service is None:
                     break
 
+                if attempt == 3:
+                    break
                 await projection_service.sync_once()
                 account = await self.ap.user_service.get_user_by_uuid(launch['account_uuid'])
+
+            if access is None and projection_service is not None:
+                # The target event may be deeper than the bounded incremental
+                # page budget. One authoritative signed snapshot catches this
+                # process up without turning the callback into unbounded polling.
+                await projection_service.refresh_snapshot()
+                account = await self.ap.user_service.get_user_by_uuid(launch['account_uuid'])
+                if account is not None:
+                    self.ap.user_service._require_active_account(account)
+                    access = await self.ap.workspace_collaboration_service.resolve_account_workspace(
+                        account.uuid,
+                        launch['workspace_uuid'],
+                    )
             if account is None:
                 raise SpaceLaunchError('Launch Account is not projected into Core')
             if access is None:  # pragma: no cover - bounded loop resolves or raises.

src/langbot/pkg/cloud/directory_projection.py

@@ -125,10 +125,21 @@ class DirectoryProjectionService:
         # The database cursor remains the shared projection high-water mark,
         # while this cursor tracks what this process has actually observed.
         self._consumer_cursor: int | None = None
+        self._sync_lock = asyncio.Lock()
 
     async def initialize(self) -> None:
         """Block Cloud startup until one full signed snapshot is committed."""
 
+        async with self._sync_lock:
+            await self._refresh_snapshot()
+
+    async def refresh_snapshot(self) -> None:
+        """Refresh from one full signed snapshot within the sync single-flight."""
+
+        async with self._sync_lock:
+            await self._refresh_snapshot()
+
+    async def _refresh_snapshot(self) -> None:
         last_superseded: _DirectorySnapshotSuperseded | None = None
         for _attempt in range(5):
             snapshot = await self.provider.fetch_snapshot(self.instance_uuid)
@@ -159,9 +170,13 @@ class DirectoryProjectionService:
                 delay = min(max(delay * 2, self.sync_interval_seconds), self.max_staleness_seconds / 2)
 
     async def sync_once(self) -> None:
+        async with self._sync_lock:
+            await self._sync_once()
+
+    async def _sync_once(self) -> None:
         cursor = self._consumer_cursor
         if cursor is None:
-            await self.initialize()
+            await self._refresh_snapshot()
             return
         batch = await self.provider.fetch_events(
             self.instance_uuid,