Merge pull request #760 from AnishSarkar22/fix/connectors

fix: various issues with connectors & other issues

surfsense_backend/app/celery_app.py

@@ -79,6 +79,7 @@ celery_app = Celery(
         "app.tasks.celery_tasks.schedule_checker_task",
         "app.tasks.celery_tasks.blocknote_migration_tasks",
         "app.tasks.celery_tasks.document_reindex_tasks",
+        "app.tasks.celery_tasks.stale_notification_cleanup_task",
     ],
 )
 
@@ -121,4 +122,14 @@ celery_app.conf.beat_schedule = {
             "expires": 30,  # Task expires after 30 seconds if not picked up
         },
     },
+    # Cleanup stale connector indexing notifications every 5 minutes
+    # This detects tasks that crashed or timed out without proper cleanup
+    # and marks their notifications as failed so users don't see perpetual "syncing"
+    "cleanup-stale-indexing-notifications": {
+        "task": "cleanup_stale_indexing_notifications",
+        "schedule": crontab(minute="*/5"),  # Every 5 minutes
+        "options": {
+            "expires": 60,  # Task expires after 60 seconds if not picked up
+        },
+    },
 }

surfsense_backend/app/connectors/composio_gmail_connector.py

@@ -5,6 +5,8 @@ Provides Gmail specific methods for data retrieval and indexing via Composio.
 """
 
 import logging
+import time
+from collections.abc import Awaitable, Callable
 from datetime import UTC, datetime
 from typing import Any
 
@@ -26,6 +28,10 @@ from app.utils.document_converters import (
     generate_unique_identifier_hash,
 )
 
+# Heartbeat configuration
+HeartbeatCallbackType = Callable[[int], Awaitable[None]]
+HEARTBEAT_INTERVAL_SECONDS = 30
+
 logger = logging.getLogger(__name__)
 
 
@@ -427,6 +433,7 @@ async def index_composio_gmail(
     log_entry,
     update_last_indexed: bool = True,
     max_items: int = 1000,
+    on_heartbeat_callback: HeartbeatCallbackType | None = None,
 ) -> tuple[int, str]:
     """Index Gmail messages via Composio with pagination and incremental processing."""
     try:
@@ -471,8 +478,16 @@ async def index_composio_gmail(
         total_documents_skipped = 0
         total_messages_fetched = 0
         result_size_estimate = None  # Will be set from first API response
+        last_heartbeat_time = time.time()
 
         while total_messages_fetched < max_items:
+            # Send heartbeat periodically to indicate task is still alive
+            if on_heartbeat_callback:
+                current_time = time.time()
+                if current_time - last_heartbeat_time >= HEARTBEAT_INTERVAL_SECONDS:
+                    await on_heartbeat_callback(total_documents_indexed)
+                    last_heartbeat_time = current_time
+
             # Calculate how many messages to fetch in this batch
             remaining = max_items - total_messages_fetched
             current_batch_size = min(batch_size, remaining)

surfsense_backend/app/connectors/composio_google_calendar_connector.py

@@ -5,6 +5,8 @@ Provides Google Calendar specific methods for data retrieval and indexing via Co
 """
 
 import logging
+import time
+from collections.abc import Awaitable, Callable
 from datetime import UTC, datetime
 from typing import Any
 
@@ -29,6 +31,10 @@ from app.utils.document_converters import (
     generate_unique_identifier_hash,
 )
 
+# Heartbeat configuration
+HeartbeatCallbackType = Callable[[int], Awaitable[None]]
+HEARTBEAT_INTERVAL_SECONDS = 30
+
 logger = logging.getLogger(__name__)
 
 
@@ -191,6 +197,7 @@ async def index_composio_google_calendar(
     log_entry,
     update_last_indexed: bool = True,
     max_items: int = 2500,
+    on_heartbeat_callback: HeartbeatCallbackType | None = None,
 ) -> tuple[int, str]:
     """Index Google Calendar events via Composio."""
     try:
@@ -262,8 +269,15 @@ async def index_composio_google_calendar(
         duplicate_content_count = (
             0  # Track events skipped due to duplicate content_hash
         )
+        last_heartbeat_time = time.time()
 
         for event in events:
+            # Send heartbeat periodically to indicate task is still alive
+            if on_heartbeat_callback:
+                current_time = time.time()
+                if current_time - last_heartbeat_time >= HEARTBEAT_INTERVAL_SECONDS:
+                    await on_heartbeat_callback(documents_indexed)
+                    last_heartbeat_time = current_time
             try:
                 # Handle both standard Google API and potential Composio variations
                 event_id = event.get("id", "") or event.get("eventId", "")

surfsense_backend/app/connectors/composio_google_drive_connector.py

@@ -9,6 +9,8 @@ import json
 import logging
 import os
 import tempfile
+import time
+from collections.abc import Awaitable, Callable
 from datetime import UTC, datetime
 from pathlib import Path
 from typing import Any
@@ -29,6 +31,10 @@ from app.utils.document_converters import (
     generate_unique_identifier_hash,
 )
 
+# Heartbeat configuration
+HeartbeatCallbackType = Callable[[int], Awaitable[None]]
+HEARTBEAT_INTERVAL_SECONDS = 30
+
 logger = logging.getLogger(__name__)
 
 
@@ -552,7 +558,9 @@ def generate_indexing_settings_hash(
         "include_subfolders": indexing_options.get("include_subfolders", True),
         "max_files_per_folder": indexing_options.get("max_files_per_folder", 100),
     }
-    return hashlib.md5(json.dumps(settings, sort_keys=True).encode()).hexdigest()
+    return hashlib.md5(
+        json.dumps(settings, sort_keys=True).encode(), usedforsecurity=False
+    ).hexdigest()
 
 
 async def index_composio_google_drive(
@@ -565,6 +573,7 @@ async def index_composio_google_drive(
     log_entry,
     update_last_indexed: bool = True,
     max_items: int = 1000,
+    on_heartbeat_callback: HeartbeatCallbackType | None = None,
 ) -> tuple[int, int, str | None]:
     """Index Google Drive files via Composio with delta sync support.
 
@@ -652,6 +661,7 @@ async def index_composio_google_drive(
                 max_items=max_items,
                 task_logger=task_logger,
                 log_entry=log_entry,
+                on_heartbeat_callback=on_heartbeat_callback,
             )
         else:
             logger.info(
@@ -684,6 +694,7 @@ async def index_composio_google_drive(
                 max_items=max_items,
                 task_logger=task_logger,
                 log_entry=log_entry,
+                on_heartbeat_callback=on_heartbeat_callback,
             )
 
         # Get new page token for next sync (always update after successful sync)
@@ -765,6 +776,7 @@ async def _index_composio_drive_delta_sync(
     max_items: int,
     task_logger: TaskLoggingService,
     log_entry,
+    on_heartbeat_callback: HeartbeatCallbackType | None = None,
 ) -> tuple[int, int, list[str]]:
     """Index Google Drive files using delta sync (only changed files).
 
@@ -774,6 +786,7 @@ async def _index_composio_drive_delta_sync(
     documents_indexed = 0
     documents_skipped = 0
     processing_errors = []
+    last_heartbeat_time = time.time()
 
     # Fetch all changes with pagination
     all_changes = []
@@ -804,6 +817,13 @@ async def _index_composio_drive_delta_sync(
     logger.info(f"Processing {len(all_changes)} changes from delta sync")
 
     for change in all_changes[:max_items]:
+        # Send heartbeat periodically to indicate task is still alive
+        if on_heartbeat_callback:
+            current_time = time.time()
+            if current_time - last_heartbeat_time >= HEARTBEAT_INTERVAL_SECONDS:
+                await on_heartbeat_callback(documents_indexed)
+                last_heartbeat_time = current_time
+
         try:
             # Handle removed files
             is_removed = change.get("removed", False)
@@ -886,11 +906,13 @@ async def _index_composio_drive_full_scan(
     max_items: int,
     task_logger: TaskLoggingService,
     log_entry,
+    on_heartbeat_callback: HeartbeatCallbackType | None = None,
 ) -> tuple[int, int, list[str]]:
     """Index Google Drive files using full scan (first sync or when no delta token)."""
     documents_indexed = 0
     documents_skipped = 0
     processing_errors = []
+    last_heartbeat_time = time.time()
 
     all_files = []
 
@@ -1001,6 +1023,13 @@ async def _index_composio_drive_full_scan(
     )
 
     for file_info in all_files:
+        # Send heartbeat periodically to indicate task is still alive
+        if on_heartbeat_callback:
+            current_time = time.time()
+            if current_time - last_heartbeat_time >= HEARTBEAT_INTERVAL_SECONDS:
+                await on_heartbeat_callback(documents_indexed)
+                last_heartbeat_time = current_time
+
         try:
             # Handle both standard Google API and potential Composio variations
             file_id = file_info.get("id", "") or file_info.get("fileId", "")

surfsense_backend/app/connectors/discord_connector.py

@@ -61,6 +61,9 @@ class DiscordConnector(commands.Bot):
             self.token = None
         self._bot_task = None  # Holds the async bot task
         self._is_running = False  # Flag to track if the bot is running
+        self._start_called_event = (
+            asyncio.Event()
+        )  # Event to signal when start() is called
 
         # Event to confirm bot is ready
         @self.event
@@ -226,6 +229,9 @@ class DiscordConnector(commands.Bot):
                 )
                 return
 
+            # Signal that we're about to call start() - this allows _wait_until_ready() to proceed
+            self._start_called_event.set()
+
             await self.start(self.token)
             logger.info("Discord bot started successfully.")
         except discord.LoginFailure:
@@ -260,6 +266,9 @@ class DiscordConnector(commands.Bot):
         else:
             logger.info("Bot is not running or already disconnected.")
 
+        # Reset the start event so the connector can be reused
+        self._start_called_event.clear()
+
     def set_token(self, token: str) -> None:
         """
         Set the discord bot token (for backward compatibility).
@@ -277,10 +286,16 @@ class DiscordConnector(commands.Bot):
         """Helper to wait until the bot is connected and ready."""
         logger.info("Waiting for the bot to be ready...")
 
-        # Give the event loop a chance to switch to the bot's startup task.
-        # This allows self.start() to begin initializing the client.
-        # Terrible solution, but necessary to avoid blocking the event loop.
-        await asyncio.sleep(1)  # Yield control to the event loop
+        # Wait for start_bot() to actually call self.start()
+        # This ensures we don't call wait_until_ready() before the client is initialized
+        try:
+            await asyncio.wait_for(self._start_called_event.wait(), timeout=30.0)
+            logger.info("Bot start() has been called, now waiting for ready state...")
+        except TimeoutError:
+            logger.error("start_bot() did not call start() within 30 seconds")
+            raise RuntimeError(
+                "Discord client failed to initialize - start() was never called"
+            ) from None
 
         try:
             await asyncio.wait_for(self.wait_until_ready(), timeout=60.0)

surfsense_backend/app/connectors/google_calendar_connector.py

@@ -252,12 +252,16 @@ class GoogleCalendarConnector:
             if dt_start.tzinfo is None:
                 dt_start = dt_start.replace(hour=0, minute=0, second=0, tzinfo=pytz.UTC)
             else:
-                dt_start = dt_start.astimezone(pytz.UTC).replace(hour=0, minute=0, second=0)
+                dt_start = dt_start.astimezone(pytz.UTC).replace(
+                    hour=0, minute=0, second=0
+                )
 
             if dt_end.tzinfo is None:
                 dt_end = dt_end.replace(hour=23, minute=59, second=59, tzinfo=pytz.UTC)
             else:
-                dt_end = dt_end.astimezone(pytz.UTC).replace(hour=23, minute=59, second=59)
+                dt_end = dt_end.astimezone(pytz.UTC).replace(
+                    hour=23, minute=59, second=59
+                )
 
             if dt_start >= dt_end:
                 return [], (

surfsense_backend/app/connectors/google_drive/credentials.py

@@ -132,6 +132,15 @@ async def get_valid_credentials(
             await session.commit()
 
         except Exception as e:
+            error_str = str(e)
+            # Check if this is an invalid_grant error (token expired/revoked)
+            if (
+                "invalid_grant" in error_str.lower()
+                or "token has been expired or revoked" in error_str.lower()
+            ):
+                raise Exception(
+                    "Google Drive authentication failed. Please re-authenticate."
+                ) from e
             raise Exception(f"Failed to refresh Google OAuth credentials: {e!s}") from e
 
     return credentials

surfsense_backend/app/routes/airtable_add_connector_route.py

@@ -442,11 +442,24 @@ async def refresh_airtable_token(
 
         if token_response.status_code != 200:
             error_detail = token_response.text
+            error_code = ""
             try:
                 error_json = token_response.json()
                 error_detail = error_json.get("error_description", error_detail)
+                error_code = error_json.get("error", "")
             except Exception:
                 pass
+            # Check if this is a token expiration/revocation error
+            error_lower = (error_detail + error_code).lower()
+            if (
+                "invalid_grant" in error_lower
+                or "expired" in error_lower
+                or "revoked" in error_lower
+            ):
+                raise HTTPException(
+                    status_code=401,
+                    detail="Airtable authentication failed. Please re-authenticate.",
+                )
             raise HTTPException(
                 status_code=400, detail=f"Token refresh failed: {error_detail}"
             )

surfsense_backend/app/routes/clickup_add_connector_route.py

@@ -417,6 +417,17 @@ async def refresh_clickup_token(
                 error_detail = error_json.get("error", error_detail)
             except Exception:
                 pass
+            # Check if this is a token expiration/revocation error
+            error_lower = error_detail.lower()
+            if (
+                "invalid_grant" in error_lower
+                or "expired" in error_lower
+                or "revoked" in error_lower
+            ):
+                raise HTTPException(
+                    status_code=401,
+                    detail="ClickUp authentication failed. Please re-authenticate.",
+                )
             raise HTTPException(
                 status_code=400, detail=f"Token refresh failed: {error_detail}"
             )

surfsense_backend/app/routes/confluence_add_connector_route.py

@@ -428,13 +428,26 @@ async def refresh_confluence_token(
 
         if token_response.status_code != 200:
             error_detail = token_response.text
+            error_code = ""
             try:
                 error_json = token_response.json()
                 error_detail = error_json.get(
                     "error_description", error_json.get("error", error_detail)
                 )
+                error_code = error_json.get("error", "")
             except Exception:
                 pass
+            # Check if this is a token expiration/revocation error
+            error_lower = (error_detail + error_code).lower()
+            if (
+                "invalid_grant" in error_lower
+                or "expired" in error_lower
+                or "revoked" in error_lower
+            ):
+                raise HTTPException(
+                    status_code=401,
+                    detail="Confluence authentication failed. Please re-authenticate.",
+                )
             raise HTTPException(
                 status_code=400, detail=f"Token refresh failed: {error_detail}"
             )

surfsense_backend/app/routes/discord_add_connector_route.py

@@ -46,6 +46,11 @@ SCOPES = [
     "guilds.members.read",  # Read member information
 ]
 
+# Discord permission bits
+VIEW_CHANNEL = 1 << 10  # 1024
+READ_MESSAGE_HISTORY = 1 << 16  # 65536
+ADMINISTRATOR = 1 << 3  # 8
+
 # Initialize security utilities
 _state_manager = None
 _token_encryption = None
@@ -531,3 +536,296 @@ async def refresh_discord_token(
         raise HTTPException(
             status_code=500, detail=f"Failed to refresh Discord tokens: {e!s}"
         ) from e
+
+
+def _compute_channel_permissions(
+    base_permissions: int,
+    bot_role_ids: set[str],
+    bot_user_id: str | None,
+    channel_overwrites: list[dict],
+    guild_id: str,
+) -> int:
+    """
+    Compute effective permissions for a channel based on role permissions and overwrites.
+
+    Discord permission computation follows this order (per official docs):
+    1. Start with base permissions from roles
+    2. Apply @everyone role overwrites (deny, then allow)
+    3. Apply role-specific overwrites (deny, then allow)
+    4. Apply member-specific overwrites (deny, then allow)
+
+    Args:
+        base_permissions: Combined permissions from all bot roles
+        bot_role_ids: Set of role IDs the bot has
+        bot_user_id: The bot's user ID for member-specific overwrites
+        channel_overwrites: List of permission overwrites for the channel
+        guild_id: Guild ID (same as @everyone role ID)
+
+    Returns:
+        Computed permission integer
+    """
+    permissions = base_permissions
+
+    # Permission overwrites are applied in order: @everyone, roles, member
+    everyone_allow = 0
+    everyone_deny = 0
+    role_allow = 0
+    role_deny = 0
+    member_allow = 0
+    member_deny = 0
+
+    for overwrite in channel_overwrites:
+        overwrite_id = overwrite.get("id")
+        overwrite_type = overwrite.get("type")  # 0 = role, 1 = member
+        allow = int(overwrite.get("allow", 0))
+        deny = int(overwrite.get("deny", 0))
+
+        if overwrite_type == 0:  # Role overwrite
+            if overwrite_id == guild_id:  # @everyone role
+                everyone_allow = allow
+                everyone_deny = deny
+            elif overwrite_id in bot_role_ids:
+                role_allow |= allow
+                role_deny |= deny
+        elif overwrite_type == 1 and bot_user_id and overwrite_id == bot_user_id:
+            # Member-specific overwrite for the bot
+            member_allow = allow
+            member_deny = deny
+
+    # Apply in order per Discord docs:
+    # 1. @everyone deny, then allow
+    permissions &= ~everyone_deny
+    permissions |= everyone_allow
+    # 2. Role deny, then allow
+    permissions &= ~role_deny
+    permissions |= role_allow
+    # 3. Member deny, then allow (applied LAST, highest priority)
+    permissions &= ~member_deny
+    permissions |= member_allow
+
+    return permissions
+
+
+@router.get("/discord/connector/{connector_id}/channels", response_model=None)
+async def get_discord_channels(
+    connector_id: int,
+    session: AsyncSession = Depends(get_async_session),
+    user: User = Depends(current_active_user),
+):
+    """
+    Get list of Discord text channels for a connector with permission info.
+
+    Uses Discord's HTTP REST API directly instead of WebSocket bot connection.
+    Computes effective permissions to determine if bot can read message history.
+
+    Args:
+        connector_id: The Discord connector ID
+        session: Database session
+        user: Current authenticated user
+
+    Returns:
+        List of channels with id, name, type, position, category_id, and can_index fields
+    """
+    from sqlalchemy import select
+
+    try:
+        # Get connector and verify ownership
+        result = await session.execute(
+            select(SearchSourceConnector).where(
+                SearchSourceConnector.id == connector_id,
+                SearchSourceConnector.user_id == user.id,
+                SearchSourceConnector.connector_type
+                == SearchSourceConnectorType.DISCORD_CONNECTOR,
+            )
+        )
+        connector = result.scalar_one_or_none()
+
+        if not connector:
+            raise HTTPException(
+                status_code=404,
+                detail="Discord connector not found or access denied",
+            )
+
+        # Get credentials and decrypt bot token
+        credentials = DiscordAuthCredentialsBase.from_dict(connector.config)
+        token_encryption = get_token_encryption()
+        is_encrypted = connector.config.get("_token_encrypted", False)
+
+        bot_token = credentials.bot_token
+        if is_encrypted and bot_token:
+            try:
+                bot_token = token_encryption.decrypt_token(bot_token)
+            except Exception as e:
+                logger.error(f"Failed to decrypt bot token: {e!s}")
+                raise HTTPException(
+                    status_code=500, detail="Failed to decrypt stored bot token"
+                ) from e
+
+        if not bot_token:
+            raise HTTPException(
+                status_code=400,
+                detail="No bot token available. Please re-authenticate.",
+            )
+
+        # Get guild_id from connector config
+        guild_id = connector.config.get("guild_id")
+        if not guild_id:
+            raise HTTPException(
+                status_code=400,
+                detail="No guild_id associated with this connector. Please reconnect the Discord server.",
+            )
+
+        headers = {"Authorization": f"Bot {bot_token}"}
+
+        async with httpx.AsyncClient() as client:
+            # Fetch bot's user info to get bot user ID
+            bot_user_response = await client.get(
+                "https://discord.com/api/v10/users/@me",
+                headers=headers,
+                timeout=30.0,
+            )
+
+            if bot_user_response.status_code != 200:
+                logger.warning(
+                    f"Failed to fetch bot user info: {bot_user_response.text}"
+                )
+                bot_user_id = None
+            else:
+                bot_user_id = bot_user_response.json().get("id")
+
+            # Fetch guild info to get roles
+            guild_response = await client.get(
+                f"https://discord.com/api/v10/guilds/{guild_id}",
+                headers=headers,
+                timeout=30.0,
+            )
+
+            if guild_response.status_code != 200:
+                raise HTTPException(
+                    status_code=guild_response.status_code,
+                    detail="Failed to fetch guild information",
+                )
+
+            guild_data = guild_response.json()
+            guild_roles = {role["id"]: role for role in guild_data.get("roles", [])}
+
+            # Fetch bot's member info to get its roles
+            bot_member_response = await client.get(
+                f"https://discord.com/api/v10/guilds/{guild_id}/members/{bot_user_id}",
+                headers=headers,
+                timeout=30.0,
+            )
+
+            if bot_member_response.status_code != 200:
+                logger.warning(
+                    f"Failed to fetch bot member info: {bot_member_response.text}"
+                )
+                bot_role_ids = {guild_id}  # At minimum, bot has @everyone role
+                base_permissions = int(
+                    guild_roles.get(guild_id, {}).get("permissions", 0)
+                )
+            else:
+                bot_member_data = bot_member_response.json()
+                bot_role_ids = set(bot_member_data.get("roles", []))
+                bot_role_ids.add(guild_id)  # @everyone role is always included
+
+                # Compute base permissions from all bot roles
+                base_permissions = 0
+                for role_id in bot_role_ids:
+                    if role_id in guild_roles:
+                        role_perms = int(guild_roles[role_id].get("permissions", 0))
+                        base_permissions |= role_perms
+
+            # Check if bot has administrator permission (bypasses all checks)
+            is_admin = (base_permissions & ADMINISTRATOR) == ADMINISTRATOR
+
+            # Fetch channels
+            channels_response = await client.get(
+                f"https://discord.com/api/v10/guilds/{guild_id}/channels",
+                headers=headers,
+                timeout=30.0,
+            )
+
+        if channels_response.status_code == 403:
+            raise HTTPException(
+                status_code=403,
+                detail="Bot does not have permission to view channels in this server. Please ensure the bot has the 'View Channels' permission.",
+            )
+        elif channels_response.status_code == 404:
+            raise HTTPException(
+                status_code=404,
+                detail="Discord server not found. The bot may have been removed from the server.",
+            )
+        elif channels_response.status_code != 200:
+            error_detail = channels_response.text
+            try:
+                error_json = channels_response.json()
+                error_detail = error_json.get("message", error_detail)
+            except Exception:
+                pass
+            raise HTTPException(
+                status_code=channels_response.status_code,
+                detail=f"Failed to fetch Discord channels: {error_detail}",
+            )
+
+        channels_data = channels_response.json()
+
+        # Discord channel types:
+        # 0 = GUILD_TEXT, 2 = GUILD_VOICE, 4 = GUILD_CATEGORY, 5 = GUILD_ANNOUNCEMENT
+        # We want text channels (type 0) and announcement channels (type 5)
+        text_channel_types = {0, 5}
+
+        text_channels = []
+        for ch in channels_data:
+            if ch.get("type") in text_channel_types:
+                # Compute effective permissions for this channel
+                if is_admin:
+                    # Administrators bypass all permission checks
+                    can_index = True
+                else:
+                    channel_overwrites = ch.get("permission_overwrites", [])
+                    effective_perms = _compute_channel_permissions(
+                        base_permissions,
+                        bot_role_ids,
+                        bot_user_id,
+                        channel_overwrites,
+                        guild_id,
+                    )
+
+                    # Bot can index if it has both VIEW_CHANNEL and READ_MESSAGE_HISTORY
+                    has_view = (effective_perms & VIEW_CHANNEL) == VIEW_CHANNEL
+                    has_read_history = (
+                        effective_perms & READ_MESSAGE_HISTORY
+                    ) == READ_MESSAGE_HISTORY
+                    can_index = has_view and has_read_history
+
+                text_channels.append(
+                    {
+                        "id": ch["id"],
+                        "name": ch["name"],
+                        "type": "text" if ch["type"] == 0 else "announcement",
+                        "position": ch.get("position", 0),
+                        "category_id": ch.get("parent_id"),
+                        "can_index": can_index,
+                    }
+                )
+
+        # Sort by position
+        text_channels.sort(key=lambda x: x["position"])
+
+        logger.info(
+            f"Fetched {len(text_channels)} text channels for Discord connector {connector_id}"
+        )
+
+        return text_channels
+
+    except HTTPException:
+        raise
+    except Exception as e:
+        logger.error(
+            f"Failed to get Discord channels for connector {connector_id}: {e!s}",
+            exc_info=True,
+        )
+        raise HTTPException(
+            status_code=500, detail=f"Failed to get Discord channels: {e!s}"
+        ) from e

surfsense_backend/app/routes/jira_add_connector_route.py

@@ -446,13 +446,26 @@ async def refresh_jira_token(
 
         if token_response.status_code != 200:
             error_detail = token_response.text
+            error_code = ""
             try:
                 error_json = token_response.json()
                 error_detail = error_json.get(
                     "error_description", error_json.get("error", error_detail)
                 )
+                error_code = error_json.get("error", "")
             except Exception:
                 pass
+            # Check if this is a token expiration/revocation error
+            error_lower = (error_detail + error_code).lower()
+            if (
+                "invalid_grant" in error_lower
+                or "expired" in error_lower
+                or "revoked" in error_lower
+            ):
+                raise HTTPException(
+                    status_code=401,
+                    detail="Jira authentication failed. Please re-authenticate.",
+                )
             raise HTTPException(
                 status_code=400, detail=f"Token refresh failed: {error_detail}"
             )

surfsense_backend/app/routes/linear_add_connector_route.py

@@ -403,11 +403,24 @@ async def refresh_linear_token(
 
         if token_response.status_code != 200:
             error_detail = token_response.text
+            error_code = ""
             try:
                 error_json = token_response.json()
                 error_detail = error_json.get("error_description", error_detail)
+                error_code = error_json.get("error", "")
             except Exception:
                 pass
+            # Check if this is a token expiration/revocation error
+            error_lower = (error_detail + error_code).lower()
+            if (
+                "invalid_grant" in error_lower
+                or "expired" in error_lower
+                or "revoked" in error_lower
+            ):
+                raise HTTPException(
+                    status_code=401,
+                    detail="Linear authentication failed. Please re-authenticate.",
+                )
             raise HTTPException(
                 status_code=400, detail=f"Token refresh failed: {error_detail}"
             )

surfsense_backend/app/routes/notion_add_connector_route.py

@@ -407,11 +407,24 @@ async def refresh_notion_token(
 
         if token_response.status_code != 200:
             error_detail = token_response.text
+            error_code = ""
             try:
                 error_json = token_response.json()
                 error_detail = error_json.get("error_description", error_detail)
+                error_code = error_json.get("error", "")
             except Exception:
                 pass
+            # Check if this is a token expiration/revocation error
+            error_lower = (error_detail + error_code).lower()
+            if (
+                "invalid_grant" in error_lower
+                or "expired" in error_lower
+                or "revoked" in error_lower
+            ):
+                raise HTTPException(
+                    status_code=401,
+                    detail="Notion authentication failed. Please re-authenticate.",
+                )
             raise HTTPException(
                 status_code=400, detail=f"Token refresh failed: {error_detail}"
             )

surfsense_backend/app/routes/search_source_connectors_routes.py

@@ -19,10 +19,12 @@ Non-OAuth connectors (BookStack, GitHub, etc.) are limited to one per search spa
 """
 
 import logging
+import os
 from datetime import UTC, datetime, timedelta
 from typing import Any
 
 import pytz
+import redis
 from dateutil.parser import isoparse
 from fastapi import APIRouter, Body, Depends, HTTPException, Query
 from pydantic import BaseModel, Field, ValidationError
@@ -78,6 +80,27 @@ from app.utils.rbac import check_permission
 # Set up logging
 logger = logging.getLogger(__name__)
 
+# Redis client for heartbeat tracking
+_heartbeat_redis_client: redis.Redis | None = None
+
+# Redis key TTL - notification is stale if no heartbeat in this time
+HEARTBEAT_TTL_SECONDS = 120  # 2 minutes
+
+
+def get_heartbeat_redis_client() -> redis.Redis:
+    """Get or create Redis client for heartbeat tracking."""
+    global _heartbeat_redis_client
+    if _heartbeat_redis_client is None:
+        redis_url = os.getenv("CELERY_BROKER_URL", "redis://localhost:6379/0")
+        _heartbeat_redis_client = redis.from_url(redis_url, decode_responses=True)
+    return _heartbeat_redis_client
+
+
+def _get_heartbeat_key(notification_id: int) -> str:
+    """Generate Redis key for notification heartbeat."""
+    return f"indexing:heartbeat:{notification_id}"
+
+
 router = APIRouter()
 
 
@@ -1137,6 +1160,7 @@ async def run_slack_indexing(
         end_date=end_date,
         indexing_function=index_slack_messages,
         update_timestamp_func=_update_connector_timestamp_by_id,
+        supports_heartbeat_callback=True,
     )
 
 
@@ -1150,6 +1174,7 @@ async def _run_indexing_with_notifications(
     indexing_function,
     update_timestamp_func=None,
     supports_retry_callback: bool = False,
+    supports_heartbeat_callback: bool = False,
 ):
     """
     Generic helper to run indexing with real-time notifications.
@@ -1164,11 +1189,14 @@ async def _run_indexing_with_notifications(
         indexing_function: Async function that performs the indexing
         update_timestamp_func: Optional function to update connector timestamp
         supports_retry_callback: Whether the indexing function supports on_retry_callback
+        supports_heartbeat_callback: Whether the indexing function supports on_heartbeat_callback
     """
     from uuid import UUID
 
+    from celery.exceptions import SoftTimeLimitExceeded
+
     notification = None
-    # Track indexed count for retry notifications
+    # Track indexed count for retry notifications and heartbeat
     current_indexed_count = 0
 
     try:
@@ -1195,6 +1223,16 @@ async def _run_indexing_with_notifications(
                 )
             )
 
+            # Set initial Redis heartbeat for stale detection
+            if notification:
+                try:
+                    heartbeat_key = _get_heartbeat_key(notification.id)
+                    get_heartbeat_redis_client().setex(
+                        heartbeat_key, HEARTBEAT_TTL_SECONDS, "0"
+                    )
+                except Exception as e:
+                    logger.warning(f"Failed to set initial Redis heartbeat: {e}")
+
         # Update notification to fetching stage
         if notification:
             await NotificationService.connector_indexing.notify_indexing_progress(
@@ -1227,6 +1265,40 @@ async def _run_indexing_with_notifications(
                     # Don't let notification errors break the indexing
                     logger.warning(f"Failed to update retry notification: {e}")
 
+        # Create heartbeat callback for connectors that support it
+        # This updates the notification periodically during long-running indexing loops
+        # to prevent the task from appearing stuck if the worker crashes
+        async def on_heartbeat_callback(indexed_count: int) -> None:
+            """Callback to update notification during indexing (heartbeat)."""
+            nonlocal notification, current_indexed_count
+            current_indexed_count = indexed_count
+            if notification:
+                try:
+                    # Set Redis heartbeat key with TTL (fast, for stale detection)
+                    heartbeat_key = _get_heartbeat_key(notification.id)
+                    get_heartbeat_redis_client().setex(
+                        heartbeat_key, HEARTBEAT_TTL_SECONDS, str(indexed_count)
+                    )
+                except Exception as e:
+                    # Don't let Redis errors break the indexing
+                    logger.warning(f"Failed to set Redis heartbeat: {e}")
+
+                try:
+                    # Still update DB notification for progress display
+                    await session.refresh(notification)
+                    await (
+                        NotificationService.connector_indexing.notify_indexing_progress(
+                            session=session,
+                            notification=notification,
+                            indexed_count=indexed_count,
+                            stage="processing",
+                        )
+                    )
+                    await session.commit()
+                except Exception as e:
+                    # Don't let notification errors break the indexing
+                    logger.warning(f"Failed to update heartbeat notification: {e}")
+
         # Build kwargs for indexing function
         indexing_kwargs = {
             "session": session,
@@ -1242,6 +1314,10 @@ async def _run_indexing_with_notifications(
         if supports_retry_callback:
             indexing_kwargs["on_retry_callback"] = on_retry_callback
 
+        # Add heartbeat callback for connectors that support it
+        if supports_heartbeat_callback:
+            indexing_kwargs["on_heartbeat_callback"] = on_heartbeat_callback
+
         # Run the indexing function
         # Some indexers return (indexed, error), others return (indexed, skipped, error)
         result = await indexing_function(**indexing_kwargs)
@@ -1398,6 +1474,32 @@ async def _run_indexing_with_notifications(
                     await (
                         session.commit()
                     )  # Commit to ensure Electric SQL syncs the notification update
+    except SoftTimeLimitExceeded:
+        # Celery soft time limit was reached - task is about to be killed
+        # Gracefully save progress and mark as interrupted
+        logger.warning(
+            f"Soft time limit reached for connector {connector_id}. "
+            f"Saving partial progress: {current_indexed_count} items indexed."
+        )
+
+        if notification:
+            try:
+                await session.refresh(notification)
+                await NotificationService.connector_indexing.notify_indexing_completed(
+                    session=session,
+                    notification=notification,
+                    indexed_count=current_indexed_count,
+                    error_message="Time limit reached. Partial sync completed. Please run again for remaining items.",
+                    is_warning=True,  # Mark as warning since partial data was indexed
+                )
+                await session.commit()
+            except Exception as notif_error:
+                logger.error(
+                    f"Failed to update notification on soft timeout: {notif_error!s}"
+                )
+
+        # Re-raise so Celery knows the task was terminated
+        raise
     except Exception as e:
         logger.error(f"Error in indexing task: {e!s}", exc_info=True)
 
@@ -1409,12 +1511,20 @@ async def _run_indexing_with_notifications(
                 await NotificationService.connector_indexing.notify_indexing_completed(
                     session=session,
                     notification=notification,
-                    indexed_count=0,
+                    indexed_count=current_indexed_count,  # Use tracked count, not 0
                     error_message=str(e),
                     skipped_count=None,  # Unknown on exception
                 )
             except Exception as notif_error:
                 logger.error(f"Failed to update notification: {notif_error!s}")
+    finally:
+        # Clean up Redis heartbeat key when task completes (success or failure)
+        if notification:
+            try:
+                heartbeat_key = _get_heartbeat_key(notification.id)
+                get_heartbeat_redis_client().delete(heartbeat_key)
+            except Exception:
+                pass  # Ignore cleanup errors - key will expire anyway
 
 
 async def run_notion_indexing_with_new_session(
@@ -1439,6 +1549,7 @@ async def run_notion_indexing_with_new_session(
             indexing_function=index_notion_pages,
             update_timestamp_func=_update_connector_timestamp_by_id,
             supports_retry_callback=True,  # Notion connector supports retry notifications
+            supports_heartbeat_callback=True,  # Notion connector supports heartbeat notifications
         )
 
 
@@ -1471,6 +1582,7 @@ async def run_notion_indexing(
         indexing_function=index_notion_pages,
         update_timestamp_func=_update_connector_timestamp_by_id,
         supports_retry_callback=True,  # Notion connector supports retry notifications
+        supports_heartbeat_callback=True,  # Notion connector supports heartbeat notifications
     )
 
 
@@ -1521,6 +1633,7 @@ async def run_github_indexing(
         end_date=end_date,
         indexing_function=index_github_repos,
         update_timestamp_func=_update_connector_timestamp_by_id,
+        supports_heartbeat_callback=True,
     )
 
 
@@ -1571,6 +1684,7 @@ async def run_linear_indexing(
         end_date=end_date,
         indexing_function=index_linear_issues,
         update_timestamp_func=_update_connector_timestamp_by_id,
+        supports_heartbeat_callback=True,
     )
 
 
@@ -1620,6 +1734,7 @@ async def run_discord_indexing(
         end_date=end_date,
         indexing_function=index_discord_messages,
         update_timestamp_func=_update_connector_timestamp_by_id,
+        supports_heartbeat_callback=True,
     )
 
 
@@ -1670,6 +1785,7 @@ async def run_teams_indexing(
         end_date=end_date,
         indexing_function=index_teams_messages,
         update_timestamp_func=_update_connector_timestamp_by_id,
+        supports_heartbeat_callback=True,
     )
 
 
@@ -1720,6 +1836,7 @@ async def run_jira_indexing(
         end_date=end_date,
         indexing_function=index_jira_issues,
         update_timestamp_func=_update_connector_timestamp_by_id,
+        supports_heartbeat_callback=True,
     )
 
 
@@ -1772,6 +1889,7 @@ async def run_confluence_indexing(
         end_date=end_date,
         indexing_function=index_confluence_pages,
         update_timestamp_func=_update_connector_timestamp_by_id,
+        supports_heartbeat_callback=True,
     )
 
 
@@ -1822,6 +1940,7 @@ async def run_clickup_indexing(
         end_date=end_date,
         indexing_function=index_clickup_tasks,
         update_timestamp_func=_update_connector_timestamp_by_id,
+        supports_heartbeat_callback=True,
     )
 
 
@@ -1872,6 +1991,7 @@ async def run_airtable_indexing(
         end_date=end_date,
         indexing_function=index_airtable_records,
         update_timestamp_func=_update_connector_timestamp_by_id,
+        supports_heartbeat_callback=True,
     )
 
 
@@ -1924,6 +2044,7 @@ async def run_google_calendar_indexing(
         end_date=end_date,
         indexing_function=index_google_calendar_events,
         update_timestamp_func=_update_connector_timestamp_by_id,
+        supports_heartbeat_callback=True,
     )
 
 
@@ -1998,6 +2119,7 @@ async def run_google_gmail_indexing(
         end_date=end_date,
         indexing_function=gmail_indexing_wrapper,
         update_timestamp_func=_update_connector_timestamp_by_id,
+        supports_heartbeat_callback=True,
     )
 
 
@@ -2206,6 +2328,7 @@ async def run_luma_indexing(
         end_date=end_date,
         indexing_function=index_luma_events,
         update_timestamp_func=_update_connector_timestamp_by_id,
+        supports_heartbeat_callback=True,
     )
 
 
@@ -2257,6 +2380,7 @@ async def run_elasticsearch_indexing(
         end_date=end_date,
         indexing_function=index_elasticsearch_documents,
         update_timestamp_func=_update_connector_timestamp_by_id,
+        supports_heartbeat_callback=True,
     )
 
 
@@ -2306,6 +2430,7 @@ async def run_web_page_indexing(
         end_date=end_date,
         indexing_function=index_crawled_urls,
         update_timestamp_func=_update_connector_timestamp_by_id,
+        supports_heartbeat_callback=True,
     )
 
 
@@ -2360,6 +2485,7 @@ async def run_bookstack_indexing(
         end_date=end_date,
         indexing_function=index_bookstack_pages,
         update_timestamp_func=_update_connector_timestamp_by_id,
+        supports_heartbeat_callback=True,
     )
 
 
@@ -2412,6 +2538,7 @@ async def run_obsidian_indexing(
         end_date=end_date,
         indexing_function=index_obsidian_vault,
         update_timestamp_func=_update_connector_timestamp_by_id,
+        supports_heartbeat_callback=True,
     )
 
 
@@ -2465,6 +2592,7 @@ async def run_composio_indexing(
         end_date=end_date,
         indexing_function=index_composio_connector,
         update_timestamp_func=_update_connector_timestamp_by_id,
+        supports_heartbeat_callback=True,
     )
 
 

surfsense_backend/app/routes/slack_add_connector_route.py

@@ -6,6 +6,7 @@ Handles OAuth 2.0 authentication flow for Slack connector.
 
 import logging
 from datetime import UTC, datetime, timedelta
+from typing import Any
 from uuid import UUID
 
 import httpx
@@ -14,6 +15,7 @@ from fastapi.responses import RedirectResponse
 from pydantic import ValidationError
 from sqlalchemy.exc import IntegrityError
 from sqlalchemy.ext.asyncio import AsyncSession
+from sqlalchemy.future import select
 
 from app.config import config
 from app.db import (
@@ -418,6 +420,19 @@ async def refresh_slack_token(
                 error_detail = error_json.get("error", error_detail)
             except Exception:
                 pass
+            # Check if this is a token expiration/revocation error
+            error_lower = error_detail.lower()
+            if (
+                "invalid_grant" in error_lower
+                or "invalid_auth" in error_lower
+                or "token_revoked" in error_lower
+                or "expired" in error_lower
+                or "revoked" in error_lower
+            ):
+                raise HTTPException(
+                    status_code=401,
+                    detail="Slack authentication failed. Please re-authenticate.",
+                )
             raise HTTPException(
                 status_code=400, detail=f"Token refresh failed: {error_detail}"
             )
@@ -427,6 +442,20 @@ async def refresh_slack_token(
         # Slack OAuth v2 returns success status in the JSON
         if not token_json.get("ok", False):
             error_msg = token_json.get("error", "Unknown error")
+            # Check if this is a token expiration/revocation error
+            error_lower = error_msg.lower()
+            if (
+                "invalid_grant" in error_lower
+                or "invalid_auth" in error_lower
+                or "invalid_refresh_token" in error_lower
+                or "token_revoked" in error_lower
+                or "expired" in error_lower
+                or "revoked" in error_lower
+            ):
+                raise HTTPException(
+                    status_code=401,
+                    detail="Slack authentication failed. Please re-authenticate.",
+                )
             raise HTTPException(
                 status_code=400, detail=f"Slack OAuth refresh error: {error_msg}"
             )
@@ -490,3 +519,88 @@ async def refresh_slack_token(
         raise HTTPException(
             status_code=500, detail=f"Failed to refresh Slack token: {e!s}"
         ) from e
+
+
+@router.get("/slack/connector/{connector_id}/channels")
+async def get_slack_channels(
+    connector_id: int,
+    session: AsyncSession = Depends(get_async_session),
+    user: User = Depends(current_active_user),
+) -> list[dict[str, Any]]:
+    """
+    Get list of Slack channels with bot membership status.
+
+    This endpoint fetches all channels the bot can see and indicates
+    whether the bot is a member of each channel (required for accessing messages).
+
+    Args:
+        connector_id: The Slack connector ID
+        session: Database session
+        user: Current authenticated user
+
+    Returns:
+        List of channels with id, name, is_private, and is_member fields
+    """
+    try:
+        # Get the connector and verify ownership
+        result = await session.execute(
+            select(SearchSourceConnector).where(
+                SearchSourceConnector.id == connector_id,
+                SearchSourceConnector.user_id == user.id,
+                SearchSourceConnector.connector_type
+                == SearchSourceConnectorType.SLACK_CONNECTOR,
+            )
+        )
+        connector = result.scalar_one_or_none()
+
+        if not connector:
+            raise HTTPException(
+                status_code=404,
+                detail="Slack connector not found or access denied",
+            )
+
+        # Get credentials and decrypt bot token
+        credentials = SlackAuthCredentialsBase.from_dict(connector.config)
+        token_encryption = get_token_encryption()
+        is_encrypted = connector.config.get("_token_encrypted", False)
+
+        bot_token = credentials.bot_token
+        if is_encrypted and bot_token:
+            try:
+                bot_token = token_encryption.decrypt_token(bot_token)
+            except Exception as e:
+                logger.error(f"Failed to decrypt bot token: {e!s}")
+                raise HTTPException(
+                    status_code=500, detail="Failed to decrypt stored bot token"
+                ) from e
+
+        if not bot_token:
+            raise HTTPException(
+                status_code=400,
+                detail="No bot token available. Please re-authenticate.",
+            )
+
+        # Import SlackHistory here to avoid circular imports
+        from app.connectors.slack_history import SlackHistory
+
+        # Create Slack client with direct token (simple pattern for quick operations)
+        slack_client = SlackHistory(token=bot_token)
+
+        channels = await slack_client.get_all_channels(include_private=True)
+
+        logger.info(
+            f"Fetched {len(channels)} channels for Slack connector {connector_id}"
+        )
+
+        return channels
+
+    except HTTPException:
+        raise
+    except Exception as e:
+        logger.error(
+            f"Failed to get Slack channels for connector {connector_id}: {e!s}",
+            exc_info=True,
+        )
+        raise HTTPException(
+            status_code=500, detail=f"Failed to get Slack channels: {e!s}"
+        ) from e

surfsense_backend/app/routes/teams_add_connector_route.py

@@ -420,11 +420,24 @@ async def refresh_teams_token(
 
     if token_response.status_code != 200:
         error_detail = token_response.text
+        error_code = ""
         try:
             error_json = token_response.json()
             error_detail = error_json.get("error_description", error_detail)
+            error_code = error_json.get("error", "")
         except Exception:
             pass
+        # Check if this is a token expiration/revocation error
+        error_lower = (error_detail + error_code).lower()
+        if (
+            "invalid_grant" in error_lower
+            or "expired" in error_lower
+            or "revoked" in error_lower
+        ):
+            raise HTTPException(
+                status_code=401,
+                detail="Microsoft Teams authentication failed. Please re-authenticate.",
+            )
         raise HTTPException(
             status_code=400, detail=f"Token refresh failed: {error_detail}"
         )

surfsense_backend/app/tasks/celery_tasks/stale_notification_cleanup_task.py

@@ -0,0 +1,164 @@
+"""Celery task to detect and mark stale connector indexing notifications as failed.
+
+This task runs periodically (every 5 minutes by default) to find notifications
+that are stuck in "in_progress" status but don't have an active Redis heartbeat key.
+These are marked as "failed" to prevent the frontend from showing a perpetual "syncing" state.
+
+Detection mechanism:
+- Active indexing tasks set a Redis key with TTL (2 minutes) as a heartbeat
+- If the task crashes, the Redis key expires automatically
+- This cleanup task checks for in-progress notifications without a Redis heartbeat key
+- Such notifications are marked as failed with O(1) batch UPDATE
+"""
+
+import json
+import logging
+import os
+from datetime import UTC, datetime
+
+import redis
+from sqlalchemy import and_, text
+from sqlalchemy.ext.asyncio import async_sessionmaker, create_async_engine
+from sqlalchemy.future import select
+from sqlalchemy.pool import NullPool
+
+from app.celery_app import celery_app
+from app.config import config
+from app.db import Notification
+
+logger = logging.getLogger(__name__)
+
+# Redis client for checking heartbeats
+_redis_client: redis.Redis | None = None
+
+
+def get_redis_client() -> redis.Redis:
+    """Get or create Redis client for heartbeat checking."""
+    global _redis_client
+    if _redis_client is None:
+        redis_url = os.getenv("CELERY_BROKER_URL", "redis://localhost:6379/0")
+        _redis_client = redis.from_url(redis_url, decode_responses=True)
+    return _redis_client
+
+
+def _get_heartbeat_key(notification_id: int) -> str:
+    """Generate Redis key for notification heartbeat."""
+    return f"indexing:heartbeat:{notification_id}"
+
+
+def get_celery_session_maker():
+    """Create async session maker for Celery tasks."""
+    engine = create_async_engine(
+        config.DATABASE_URL,
+        poolclass=NullPool,
+        echo=False,
+    )
+    return async_sessionmaker(engine, expire_on_commit=False)
+
+
+@celery_app.task(name="cleanup_stale_indexing_notifications")
+def cleanup_stale_indexing_notifications_task():
+    """
+    Check for stale connector indexing notifications and mark them as failed.
+
+    This task finds notifications that:
+    - Have type = 'connector_indexing'
+    - Have metadata.status = 'in_progress'
+    - Do NOT have a corresponding Redis heartbeat key (meaning task crashed)
+
+    And marks them as failed with O(1) batch UPDATE.
+    """
+    import asyncio
+
+    loop = asyncio.new_event_loop()
+    asyncio.set_event_loop(loop)
+
+    try:
+        loop.run_until_complete(_cleanup_stale_notifications())
+    finally:
+        loop.close()
+
+
+async def _cleanup_stale_notifications():
+    """Find and mark stale connector indexing notifications as failed.
+
+    Uses Redis TTL-based detection:
+    1. Find all in-progress notifications
+    2. Check which ones are missing their Redis heartbeat key
+    3. Mark those as failed with O(1) batch UPDATE using JSONB || operator
+    """
+    async with get_celery_session_maker()() as session:
+        try:
+            # Find all in-progress connector indexing notifications
+            result = await session.execute(
+                select(Notification.id).where(
+                    and_(
+                        Notification.type == "connector_indexing",
+                        Notification.notification_metadata["status"].astext
+                        == "in_progress",
+                    )
+                )
+            )
+            in_progress_ids = [row[0] for row in result.fetchall()]
+
+            if not in_progress_ids:
+                logger.debug("No in-progress connector indexing notifications found")
+                return
+
+            # Check which ones are missing heartbeat keys in Redis
+            redis_client = get_redis_client()
+            stale_notification_ids = []
+
+            for notification_id in in_progress_ids:
+                heartbeat_key = _get_heartbeat_key(notification_id)
+                if not redis_client.exists(heartbeat_key):
+                    stale_notification_ids.append(notification_id)
+
+            if not stale_notification_ids:
+                logger.debug(
+                    f"All {len(in_progress_ids)} in-progress notifications have active Redis heartbeats"
+                )
+                return
+
+            logger.warning(
+                f"Found {len(stale_notification_ids)} stale connector indexing notifications "
+                f"(no Redis heartbeat key): {stale_notification_ids}"
+            )
+
+            # O(1) Batch UPDATE using JSONB || operator
+            # This merges the update data into existing notification_metadata
+            # Also updates title and message for proper UI display
+            error_message = (
+                "Something went wrong while syncing your content. Please retry."
+            )
+
+            update_data = {
+                "status": "failed",
+                "completed_at": datetime.now(UTC).isoformat(),
+                "error_message": error_message,
+                "sync_stage": "failed",
+            }
+
+            await session.execute(
+                text("""
+                    UPDATE notifications 
+                    SET metadata = metadata || CAST(:update_json AS jsonb),
+                        title = 'Failed: ' || COALESCE(metadata->>'connector_name', 'Connector'),
+                        message = :display_message
+                    WHERE id = ANY(:ids)
+                """),
+                {
+                    "update_json": json.dumps(update_data),
+                    "display_message": f"{error_message}",
+                    "ids": stale_notification_ids,
+                },
+            )
+
+            await session.commit()
+            logger.info(
+                f"Successfully marked {len(stale_notification_ids)} stale notifications as failed (batch UPDATE)"
+            )
+
+        except Exception as e:
+            logger.error(f"Error cleaning up stale notifications: {e!s}", exc_info=True)
+            await session.rollback()

surfsense_backend/app/tasks/composio_indexer.py

@@ -9,6 +9,7 @@ to avoid circular import issues with the connector_indexers package.
 """
 
 import logging
+from collections.abc import Awaitable, Callable
 from importlib import import_module
 
 from sqlalchemy.exc import SQLAlchemyError
@@ -22,6 +23,9 @@ from app.db import (
 from app.services.composio_service import INDEXABLE_TOOLKITS, TOOLKIT_TO_INDEXER
 from app.services.task_logging_service import TaskLoggingService
 
+# Type alias for heartbeat callback function
+HeartbeatCallbackType = Callable[[int], Awaitable[None]]
+
 # Set up logging
 logger = logging.getLogger(__name__)
 
@@ -86,6 +90,7 @@ async def index_composio_connector(
     end_date: str | None = None,
     update_last_indexed: bool = True,
     max_items: int = 1000,
+    on_heartbeat_callback: HeartbeatCallbackType | None = None,
 ) -> tuple[int, int, str | None]:
     """
     Index content from a Composio connector.
@@ -102,6 +107,7 @@ async def index_composio_connector(
         end_date: End date for filtering (YYYY-MM-DD format)
         update_last_indexed: Whether to update the last_indexed_at timestamp
         max_items: Maximum number of items to fetch
+        on_heartbeat_callback: Optional callback to report progress for heartbeat updates
 
     Returns:
         Tuple of (number_of_indexed_items, number_of_skipped_items, error_message or None)
@@ -180,6 +186,7 @@ async def index_composio_connector(
             "log_entry": log_entry,
             "update_last_indexed": update_last_indexed,
             "max_items": max_items,
+            "on_heartbeat_callback": on_heartbeat_callback,
         }
 
         # Add date params for toolkits that support them

surfsense_backend/app/tasks/connector_indexers/airtable_indexer.py

@@ -2,6 +2,9 @@
 Airtable connector indexer.
 """
 
+import time
+from collections.abc import Awaitable, Callable
+
 from sqlalchemy.exc import SQLAlchemyError
 from sqlalchemy.ext.asyncio import AsyncSession
 
@@ -27,6 +30,12 @@ from .base import (
     update_connector_last_indexed,
 )
 
+# Type hint for heartbeat callback
+HeartbeatCallbackType = Callable[[int], Awaitable[None]]
+
+# Heartbeat interval in seconds
+HEARTBEAT_INTERVAL_SECONDS = 30
+
 
 async def index_airtable_records(
     session: AsyncSession,
@@ -37,6 +46,7 @@ async def index_airtable_records(
     end_date: str | None = None,
     max_records: int = 2500,
     update_last_indexed: bool = True,
+    on_heartbeat_callback: HeartbeatCallbackType | None = None,
 ) -> tuple[int, str | None]:
     """
     Index Airtable records for a given connector.
@@ -50,6 +60,7 @@ async def index_airtable_records(
         end_date: End date for filtering records (YYYY-MM-DD)
         max_records: Maximum number of records to fetch per table
         update_last_indexed: Whether to update the last_indexed_at timestamp
+        on_heartbeat_callback: Optional callback to update notification during long-running indexing.
 
     Returns:
         Tuple of (number_of_documents_processed, error_message)
@@ -127,8 +138,20 @@ async def index_airtable_records(
 
             logger.info(f"Found {len(bases)} Airtable bases to process")
 
+            # Heartbeat tracking - update notification periodically to prevent appearing stuck
+            last_heartbeat_time = time.time()
+            total_documents_indexed = 0
+
             # Process each base
             for base in bases:
+                # Check if it's time for a heartbeat update
+                if (
+                    on_heartbeat_callback
+                    and (time.time() - last_heartbeat_time)
+                    >= HEARTBEAT_INTERVAL_SECONDS
+                ):
+                    await on_heartbeat_callback(total_documents_indexed)
+                    last_heartbeat_time = time.time()
                 base_id = base.get("id")
                 base_name = base.get("name", "Unknown Base")
 
@@ -204,6 +227,15 @@ async def index_airtable_records(
                     documents_skipped = 0
                     # Process each record
                     for record in records:
+                        # Check if it's time for a heartbeat update
+                        if (
+                            on_heartbeat_callback
+                            and (time.time() - last_heartbeat_time)
+                            >= HEARTBEAT_INTERVAL_SECONDS
+                        ):
+                            await on_heartbeat_callback(total_documents_indexed)
+                            last_heartbeat_time = time.time()
+
                         try:
                             # Generate markdown content
                             markdown_content = (

surfsense_backend/app/tasks/connector_indexers/bookstack_indexer.py

@@ -2,6 +2,8 @@
 BookStack connector indexer.
 """
 
+import time
+from collections.abc import Awaitable, Callable
 from datetime import datetime
 
 from sqlalchemy.exc import SQLAlchemyError
@@ -29,6 +31,12 @@ from .base import (
     update_connector_last_indexed,
 )
 
+# Type hint for heartbeat callback
+HeartbeatCallbackType = Callable[[int], Awaitable[None]]
+
+# Heartbeat interval in seconds
+HEARTBEAT_INTERVAL_SECONDS = 30
+
 
 async def index_bookstack_pages(
     session: AsyncSession,
@@ -38,6 +46,7 @@ async def index_bookstack_pages(
     start_date: str | None = None,
     end_date: str | None = None,
     update_last_indexed: bool = True,
+    on_heartbeat_callback: HeartbeatCallbackType | None = None,
 ) -> tuple[int, str | None]:
     """
     Index BookStack pages.
@@ -50,6 +59,7 @@ async def index_bookstack_pages(
         start_date: Start date for indexing (YYYY-MM-DD format)
         end_date: End date for indexing (YYYY-MM-DD format)
         update_last_indexed: Whether to update the last_indexed_at timestamp (default: True)
+        on_heartbeat_callback: Optional callback to update notification during long-running indexing.
 
     Returns:
         Tuple containing (number of documents indexed, error message or None)
@@ -179,7 +189,17 @@ async def index_bookstack_pages(
         skipped_pages = []
         documents_skipped = 0
 
+        # Heartbeat tracking - update notification periodically to prevent appearing stuck
+        last_heartbeat_time = time.time()
+
         for page in pages:
+            # Check if it's time for a heartbeat update
+            if (
+                on_heartbeat_callback
+                and (time.time() - last_heartbeat_time) >= HEARTBEAT_INTERVAL_SECONDS
+            ):
+                await on_heartbeat_callback(documents_indexed)
+                last_heartbeat_time = time.time()
             try:
                 page_id = page.get("id")
                 page_name = page.get("name", "")

surfsense_backend/app/tasks/connector_indexers/clickup_indexer.py

@@ -3,6 +3,8 @@ ClickUp connector indexer.
 """
 
 import contextlib
+import time
+from collections.abc import Awaitable, Callable
 from datetime import datetime
 
 from sqlalchemy.exc import SQLAlchemyError
@@ -29,6 +31,12 @@ from .base import (
     update_connector_last_indexed,
 )
 
+# Type hint for heartbeat callback
+HeartbeatCallbackType = Callable[[int], Awaitable[None]]
+
+# Heartbeat interval in seconds
+HEARTBEAT_INTERVAL_SECONDS = 30
+
 
 async def index_clickup_tasks(
     session: AsyncSession,
@@ -38,6 +46,7 @@ async def index_clickup_tasks(
     start_date: str | None = None,
     end_date: str | None = None,
     update_last_indexed: bool = True,
+    on_heartbeat_callback: HeartbeatCallbackType | None = None,
 ) -> tuple[int, str | None]:
     """
     Index tasks from ClickUp workspace.
@@ -50,6 +59,7 @@ async def index_clickup_tasks(
         start_date: Start date for filtering tasks (YYYY-MM-DD format)
         end_date: End date for filtering tasks (YYYY-MM-DD format)
         update_last_indexed: Whether to update the last_indexed_at timestamp
+        on_heartbeat_callback: Optional callback to update notification during long-running indexing.
 
     Returns:
         Tuple of (number of indexed tasks, error message if any)
@@ -132,6 +142,9 @@ async def index_clickup_tasks(
         documents_indexed = 0
         documents_skipped = 0
 
+        # Heartbeat tracking - update notification periodically to prevent appearing stuck
+        last_heartbeat_time = time.time()
+
         # Iterate workspaces and fetch tasks
         for workspace in workspaces:
             workspace_id = workspace.get("id")
@@ -170,6 +183,15 @@ async def index_clickup_tasks(
             )
 
             for task in tasks:
+                # Check if it's time for a heartbeat update
+                if (
+                    on_heartbeat_callback
+                    and (time.time() - last_heartbeat_time)
+                    >= HEARTBEAT_INTERVAL_SECONDS
+                ):
+                    await on_heartbeat_callback(documents_indexed)
+                    last_heartbeat_time = time.time()
+
                 try:
                     task_id = task.get("id")
                     task_name = task.get("name", "Untitled Task")

surfsense_backend/app/tasks/connector_indexers/confluence_indexer.py

@@ -3,6 +3,8 @@ Confluence connector indexer.
 """
 
 import contextlib
+import time
+from collections.abc import Awaitable, Callable
 from datetime import datetime
 
 from sqlalchemy.exc import SQLAlchemyError
@@ -30,6 +32,12 @@ from .base import (
     update_connector_last_indexed,
 )
 
+# Type hint for heartbeat callback
+HeartbeatCallbackType = Callable[[int], Awaitable[None]]
+
+# Heartbeat interval in seconds
+HEARTBEAT_INTERVAL_SECONDS = 30
+
 
 async def index_confluence_pages(
     session: AsyncSession,
@@ -39,6 +47,7 @@ async def index_confluence_pages(
     start_date: str | None = None,
     end_date: str | None = None,
     update_last_indexed: bool = True,
+    on_heartbeat_callback: HeartbeatCallbackType | None = None,
 ) -> tuple[int, str | None]:
     """
     Index Confluence pages and comments.
@@ -51,6 +60,7 @@ async def index_confluence_pages(
         start_date: Start date for indexing (YYYY-MM-DD format)
         end_date: End date for indexing (YYYY-MM-DD format)
         update_last_indexed: Whether to update the last_indexed_at timestamp (default: True)
+        on_heartbeat_callback: Optional callback to update notification during long-running indexing.
 
     Returns:
         Tuple containing (number of documents indexed, error message or None)
@@ -175,7 +185,17 @@ async def index_confluence_pages(
         skipped_pages = []
         documents_skipped = 0
 
+        # Heartbeat tracking - update notification periodically to prevent appearing stuck
+        last_heartbeat_time = time.time()
+
         for page in pages:
+            # Check if it's time for a heartbeat update
+            if (
+                on_heartbeat_callback
+                and (time.time() - last_heartbeat_time) >= HEARTBEAT_INTERVAL_SECONDS
+            ):
+                await on_heartbeat_callback(documents_indexed)
+                last_heartbeat_time = time.time()
             try:
                 page_id = page.get("id")
                 page_title = page.get("title", "")

surfsense_backend/app/tasks/connector_indexers/discord_indexer.py

@@ -3,6 +3,8 @@ Discord connector indexer.
 """
 
 import asyncio
+import time
+from collections.abc import Awaitable, Callable
 from datetime import UTC, datetime, timedelta
 
 from sqlalchemy.exc import SQLAlchemyError
@@ -28,6 +30,12 @@ from .base import (
     update_connector_last_indexed,
 )
 
+# Type hint for heartbeat callback
+HeartbeatCallbackType = Callable[[int], Awaitable[None]]
+
+# Heartbeat interval in seconds - update notification every 30 seconds
+HEARTBEAT_INTERVAL_SECONDS = 30
+
 
 async def index_discord_messages(
     session: AsyncSession,
@@ -37,6 +45,7 @@ async def index_discord_messages(
     start_date: str | None = None,
     end_date: str | None = None,
     update_last_indexed: bool = True,
+    on_heartbeat_callback: HeartbeatCallbackType | None = None,
 ) -> tuple[int, str | None]:
     """
     Index Discord messages from all accessible channels.
@@ -49,6 +58,8 @@ async def index_discord_messages(
         start_date: Start date for indexing (YYYY-MM-DD format)
         end_date: End date for indexing (YYYY-MM-DD format)
         update_last_indexed: Whether to update the last_indexed_at timestamp (default: True)
+        on_heartbeat_callback: Optional callback to update notification during long-running indexing.
+            Called periodically with (indexed_count) to prevent task appearing stuck.
 
     Returns:
         Tuple containing (number of documents indexed, error message or None)
@@ -281,6 +292,9 @@ async def index_discord_messages(
         documents_skipped = 0
         skipped_channels: list[str] = []
 
+        # Heartbeat tracking - update notification periodically to prevent appearing stuck
+        last_heartbeat_time = time.time()
+
         # Process each guild and channel
         await task_logger.log_task_progress(
             log_entry,
@@ -290,6 +304,14 @@ async def index_discord_messages(
 
         try:
             for guild in guilds:
+                # Check if it's time for a heartbeat update
+                if (
+                    on_heartbeat_callback
+                    and (time.time() - last_heartbeat_time)
+                    >= HEARTBEAT_INTERVAL_SECONDS
+                ):
+                    await on_heartbeat_callback(documents_indexed)
+                    last_heartbeat_time = time.time()
                 guild_id = guild["id"]
                 guild_name = guild["name"]
                 logger.info(f"Processing guild: {guild_name} ({guild_id})")

surfsense_backend/app/tasks/connector_indexers/elasticsearch_indexer.py

@@ -4,6 +4,8 @@ Elasticsearch indexer for SurfSense
 
 import json
 import logging
+import time
+from collections.abc import Awaitable, Callable
 from datetime import UTC, datetime
 from typing import Any
 
@@ -25,6 +27,12 @@ from .base import (
     get_current_timestamp,
 )
 
+# Type hint for heartbeat callback
+HeartbeatCallbackType = Callable[[int], Awaitable[None]]
+
+# Heartbeat interval in seconds
+HEARTBEAT_INTERVAL_SECONDS = 30
+
 logger = logging.getLogger(__name__)
 
 
@@ -36,6 +44,7 @@ async def index_elasticsearch_documents(
     start_date: str,
     end_date: str,
     update_last_indexed: bool = True,
+    on_heartbeat_callback: HeartbeatCallbackType | None = None,
 ) -> tuple[int, str | None]:
     """
     Index documents from Elasticsearch into SurfSense
@@ -48,6 +57,7 @@ async def index_elasticsearch_documents(
         start_date: Start date for indexing (not used for Elasticsearch, kept for compatibility)
         end_date: End date for indexing (not used for Elasticsearch, kept for compatibility)
         update_last_indexed: Whether to update the last indexed timestamp
+        on_heartbeat_callback: Optional callback to update notification during long-running indexing.
 
     Returns:
         Tuple of (number of documents processed, error message if any)
@@ -155,6 +165,9 @@ async def index_elasticsearch_documents(
 
         documents_processed = 0
 
+        # Heartbeat tracking - update notification periodically to prevent appearing stuck
+        last_heartbeat_time = time.time()
+
         try:
             await task_logger.log_task_progress(
                 log_entry,
@@ -172,6 +185,15 @@ async def index_elasticsearch_documents(
                 size=min(max_documents, 100),  # Scroll in batches
                 fields=config.get("ELASTICSEARCH_FIELDS"),
             ):
+                # Check if it's time for a heartbeat update
+                if (
+                    on_heartbeat_callback
+                    and (time.time() - last_heartbeat_time)
+                    >= HEARTBEAT_INTERVAL_SECONDS
+                ):
+                    await on_heartbeat_callback(documents_processed)
+                    last_heartbeat_time = time.time()
+
                 if documents_processed >= max_documents:
                     break
 

surfsense_backend/app/tasks/connector_indexers/github_indexer.py

@@ -5,6 +5,8 @@ This indexer processes entire repository digests in one pass, dramatically
 reducing LLM API calls compared to the previous file-by-file approach.
 """
 
+import time
+from collections.abc import Awaitable, Callable
 from datetime import UTC, datetime
 
 from sqlalchemy.exc import SQLAlchemyError
@@ -30,6 +32,12 @@ from .base import (
     logger,
 )
 
+# Type hint for heartbeat callback
+HeartbeatCallbackType = Callable[[int], Awaitable[None]]
+
+# Heartbeat interval in seconds - update notification every 30 seconds
+HEARTBEAT_INTERVAL_SECONDS = 30
+
 # Maximum tokens for a single digest before splitting
 # Most LLMs can handle 128k+ tokens now, but we'll be conservative
 MAX_DIGEST_CHARS = 500_000  # ~125k tokens
@@ -43,6 +51,7 @@ async def index_github_repos(
     start_date: str | None = None,  # Ignored - GitHub indexes full repo snapshots
     end_date: str | None = None,  # Ignored - GitHub indexes full repo snapshots
     update_last_indexed: bool = True,
+    on_heartbeat_callback: HeartbeatCallbackType | None = None,
 ) -> tuple[int, str | None]:
     """
     Index GitHub repositories using gitingest for efficient processing.
@@ -62,6 +71,7 @@ async def index_github_repos(
         start_date: Ignored - kept for API compatibility
         end_date: Ignored - kept for API compatibility
         update_last_indexed: Whether to update the last_indexed_at timestamp (default: True)
+        on_heartbeat_callback: Optional callback to update notification during long-running indexing.
 
     Returns:
         Tuple containing (number of documents indexed, error message or None)
@@ -168,7 +178,18 @@ async def index_github_repos(
             f"Starting gitingest indexing for {len(repo_full_names_to_index)} repositories."
         )
 
+        # Heartbeat tracking - update notification periodically to prevent appearing stuck
+        last_heartbeat_time = time.time()
+        documents_indexed = 0
+
         for repo_full_name in repo_full_names_to_index:
+            # Check if it's time for a heartbeat update
+            if (
+                on_heartbeat_callback
+                and (time.time() - last_heartbeat_time) >= HEARTBEAT_INTERVAL_SECONDS
+            ):
+                await on_heartbeat_callback(documents_indexed)
+                last_heartbeat_time = time.time()
             if not repo_full_name or not isinstance(repo_full_name, str):
                 logger.warning(f"Skipping invalid repository entry: {repo_full_name}")
                 continue

surfsense_backend/app/tasks/connector_indexers/google_calendar_indexer.py

@@ -2,10 +2,10 @@
 Google Calendar connector indexer.
 """
 
+import time
+from collections.abc import Awaitable, Callable
 from datetime import datetime, timedelta
 
-import pytz
-from dateutil.parser import isoparse
 from google.oauth2.credentials import Credentials
 from sqlalchemy.exc import SQLAlchemyError
 from sqlalchemy.ext.asyncio import AsyncSession
@@ -30,6 +30,12 @@ from .base import (
     update_connector_last_indexed,
 )
 
+# Type hint for heartbeat callback
+HeartbeatCallbackType = Callable[[int], Awaitable[None]]
+
+# Heartbeat interval in seconds
+HEARTBEAT_INTERVAL_SECONDS = 30
+
 
 async def index_google_calendar_events(
     session: AsyncSession,
@@ -39,6 +45,7 @@ async def index_google_calendar_events(
     start_date: str | None = None,
     end_date: str | None = None,
     update_last_indexed: bool = True,
+    on_heartbeat_callback: HeartbeatCallbackType | None = None,
 ) -> tuple[int, str | None]:
     """
     Index Google Calendar events.
@@ -52,6 +59,7 @@ async def index_google_calendar_events(
         end_date: End date for indexing (YYYY-MM-DD format). Can be in the future to index upcoming events.
                   Defaults to today if not provided.
         update_last_indexed: Whether to update the last_indexed_at timestamp (default: True)
+        on_heartbeat_callback: Optional callback to update notification during long-running indexing.
 
     Returns:
         Tuple containing (number of documents indexed, error message or None)
@@ -281,7 +289,17 @@ async def index_google_calendar_events(
             0  # Track events skipped due to duplicate content_hash
         )
 
+        # Heartbeat tracking - update notification periodically to prevent appearing stuck
+        last_heartbeat_time = time.time()
+
         for event in events:
+            # Check if it's time for a heartbeat update
+            if (
+                on_heartbeat_callback
+                and (time.time() - last_heartbeat_time) >= HEARTBEAT_INTERVAL_SECONDS
+            ):
+                await on_heartbeat_callback(documents_indexed)
+                last_heartbeat_time = time.time()
             try:
                 event_id = event.get("id")
                 event_summary = event.get("summary", "No Title")

surfsense_backend/app/tasks/connector_indexers/google_drive_indexer.py

@@ -1,6 +1,8 @@
 """Google Drive indexer using Surfsense file processors."""
 
 import logging
+import time
+from collections.abc import Awaitable, Callable
 
 from sqlalchemy.exc import SQLAlchemyError
 from sqlalchemy.ext.asyncio import AsyncSession
@@ -24,6 +26,12 @@ from app.tasks.connector_indexers.base import (
 )
 from app.utils.document_converters import generate_unique_identifier_hash
 
+# Type hint for heartbeat callback
+HeartbeatCallbackType = Callable[[int], Awaitable[None]]
+
+# Heartbeat interval in seconds
+HEARTBEAT_INTERVAL_SECONDS = 30
+
 logger = logging.getLogger(__name__)
 
 
@@ -38,6 +46,7 @@ async def index_google_drive_files(
     update_last_indexed: bool = True,
     max_files: int = 500,
     include_subfolders: bool = False,
+    on_heartbeat_callback: HeartbeatCallbackType | None = None,
 ) -> tuple[int, str | None]:
     """
     Index Google Drive files for a specific connector.
@@ -53,6 +62,7 @@ async def index_google_drive_files(
         update_last_indexed: Whether to update last_indexed_at timestamp
         max_files: Maximum number of files to index
         include_subfolders: Whether to recursively index files in subfolders
+        on_heartbeat_callback: Optional callback to update notification during long-running indexing.
 
     Returns:
         Tuple of (number_of_indexed_files, error_message)
@@ -147,6 +157,7 @@ async def index_google_drive_files(
                 log_entry=log_entry,
                 max_files=max_files,
                 include_subfolders=include_subfolders,
+                on_heartbeat_callback=on_heartbeat_callback,
             )
         else:
             logger.info(f"Using full scan for connector {connector_id}")
@@ -163,6 +174,7 @@ async def index_google_drive_files(
                 log_entry=log_entry,
                 max_files=max_files,
                 include_subfolders=include_subfolders,
+                on_heartbeat_callback=on_heartbeat_callback,
             )
 
         documents_indexed, documents_skipped = result
@@ -383,6 +395,7 @@ async def _index_full_scan(
     log_entry: any,
     max_files: int,
     include_subfolders: bool = False,
+    on_heartbeat_callback: HeartbeatCallbackType | None = None,
 ) -> tuple[int, int]:
     """Perform full scan indexing of a folder."""
     await task_logger.log_task_progress(
@@ -399,10 +412,20 @@ async def _index_full_scan(
     documents_skipped = 0
     files_processed = 0
 
+    # Heartbeat tracking - update notification periodically to prevent appearing stuck
+    last_heartbeat_time = time.time()
+
     # Queue of folders to process: (folder_id, folder_name)
     folders_to_process = [(folder_id, folder_name)]
 
     while folders_to_process and files_processed < max_files:
+        # Check if it's time for a heartbeat update
+        if (
+            on_heartbeat_callback
+            and (time.time() - last_heartbeat_time) >= HEARTBEAT_INTERVAL_SECONDS
+        ):
+            await on_heartbeat_callback(documents_indexed)
+            last_heartbeat_time = time.time()
         current_folder_id, current_folder_name = folders_to_process.pop(0)
         logger.info(f"Processing folder: {current_folder_name} ({current_folder_id})")
         page_token = None
@@ -485,6 +508,7 @@ async def _index_with_delta_sync(
     log_entry: any,
     max_files: int,
     include_subfolders: bool = False,
+    on_heartbeat_callback: HeartbeatCallbackType | None = None,
 ) -> tuple[int, int]:
     """Perform delta sync indexing using change tracking.
 
@@ -515,7 +539,17 @@ async def _index_with_delta_sync(
     documents_skipped = 0
     files_processed = 0
 
+    # Heartbeat tracking - update notification periodically to prevent appearing stuck
+    last_heartbeat_time = time.time()
+
     for change in changes:
+        # Check if it's time for a heartbeat update
+        if (
+            on_heartbeat_callback
+            and (time.time() - last_heartbeat_time) >= HEARTBEAT_INTERVAL_SECONDS
+        ):
+            await on_heartbeat_callback(documents_indexed)
+            last_heartbeat_time = time.time()
         if files_processed >= max_files:
             break
 

surfsense_backend/app/tasks/connector_indexers/google_gmail_indexer.py

@@ -2,6 +2,8 @@
 Google Gmail connector indexer.
 """
 
+import time
+from collections.abc import Awaitable, Callable
 from datetime import datetime
 
 from google.oauth2.credentials import Credentials
@@ -33,6 +35,12 @@ from .base import (
     update_connector_last_indexed,
 )
 
+# Type hint for heartbeat callback
+HeartbeatCallbackType = Callable[[int], Awaitable[None]]
+
+# Heartbeat interval in seconds
+HEARTBEAT_INTERVAL_SECONDS = 30
+
 
 async def index_google_gmail_messages(
     session: AsyncSession,
@@ -43,6 +51,7 @@ async def index_google_gmail_messages(
     end_date: str | None = None,
     update_last_indexed: bool = True,
     max_messages: int = 1000,
+    on_heartbeat_callback: HeartbeatCallbackType | None = None,
 ) -> tuple[int, str]:
     """
     Index Gmail messages for a specific connector.
@@ -56,6 +65,7 @@ async def index_google_gmail_messages(
         end_date: End date for filtering messages (YYYY-MM-DD format)
         update_last_indexed: Whether to update the last_indexed_at timestamp (default: True)
         max_messages: Maximum number of messages to fetch (default: 100)
+        on_heartbeat_callback: Optional callback to update notification during long-running indexing.
 
     Returns:
         Tuple of (number_of_indexed_messages, status_message)
@@ -212,7 +222,18 @@ async def index_google_gmail_messages(
         documents_indexed = 0
         skipped_messages = []
         documents_skipped = 0
+
+        # Heartbeat tracking - update notification periodically to prevent appearing stuck
+        last_heartbeat_time = time.time()
+
         for message in messages:
+            # Check if it's time for a heartbeat update
+            if (
+                on_heartbeat_callback
+                and (time.time() - last_heartbeat_time) >= HEARTBEAT_INTERVAL_SECONDS
+            ):
+                await on_heartbeat_callback(documents_indexed)
+                last_heartbeat_time = time.time()
             try:
                 # Extract message information
                 message_id = message.get("id", "")

surfsense_backend/app/tasks/connector_indexers/jira_indexer.py

@@ -3,6 +3,8 @@ Jira connector indexer.
 """
 
 import contextlib
+import time
+from collections.abc import Awaitable, Callable
 from datetime import datetime
 
 from sqlalchemy.exc import SQLAlchemyError
@@ -30,6 +32,12 @@ from .base import (
     update_connector_last_indexed,
 )
 
+# Type hint for heartbeat callback
+HeartbeatCallbackType = Callable[[int], Awaitable[None]]
+
+# Heartbeat interval in seconds - update notification every 30 seconds
+HEARTBEAT_INTERVAL_SECONDS = 30
+
 
 async def index_jira_issues(
     session: AsyncSession,
@@ -39,6 +47,7 @@ async def index_jira_issues(
     start_date: str | None = None,
     end_date: str | None = None,
     update_last_indexed: bool = True,
+    on_heartbeat_callback: HeartbeatCallbackType | None = None,
 ) -> tuple[int, str | None]:
     """
     Index Jira issues and comments.
@@ -51,6 +60,7 @@ async def index_jira_issues(
         start_date: Start date for indexing (YYYY-MM-DD format)
         end_date: End date for indexing (YYYY-MM-DD format)
         update_last_indexed: Whether to update the last_indexed_at timestamp (default: True)
+        on_heartbeat_callback: Optional callback to update notification during long-running indexing.
 
     Returns:
         Tuple containing (number of documents indexed, error message or None)
@@ -169,7 +179,17 @@ async def index_jira_issues(
         skipped_issues = []
         documents_skipped = 0
 
+        # Heartbeat tracking - update notification periodically to prevent appearing stuck
+        last_heartbeat_time = time.time()
+
         for issue in issues:
+            # Check if it's time for a heartbeat update
+            if (
+                on_heartbeat_callback
+                and (time.time() - last_heartbeat_time) >= HEARTBEAT_INTERVAL_SECONDS
+            ):
+                await on_heartbeat_callback(documents_indexed)
+                last_heartbeat_time = time.time()
             try:
                 issue_id = issue.get("key")
                 issue_identifier = issue.get("key", "")

surfsense_backend/app/tasks/connector_indexers/linear_indexer.py

@@ -2,6 +2,8 @@
 Linear connector indexer.
 """
 
+import time
+from collections.abc import Awaitable, Callable
 from datetime import datetime
 
 from sqlalchemy.exc import SQLAlchemyError
@@ -29,6 +31,12 @@ from .base import (
     update_connector_last_indexed,
 )
 
+# Type hint for heartbeat callback
+HeartbeatCallbackType = Callable[[int], Awaitable[None]]
+
+# Heartbeat interval in seconds - update notification every 30 seconds
+HEARTBEAT_INTERVAL_SECONDS = 30
+
 
 async def index_linear_issues(
     session: AsyncSession,
@@ -38,6 +46,7 @@ async def index_linear_issues(
     start_date: str | None = None,
     end_date: str | None = None,
     update_last_indexed: bool = True,
+    on_heartbeat_callback: HeartbeatCallbackType | None = None,
 ) -> tuple[int, str | None]:
     """
     Index Linear issues and comments.
@@ -50,6 +59,7 @@ async def index_linear_issues(
         start_date: Start date for indexing (YYYY-MM-DD format)
         end_date: End date for indexing (YYYY-MM-DD format)
         update_last_indexed: Whether to update the last_indexed_at timestamp (default: True)
+        on_heartbeat_callback: Optional callback to update notification during long-running indexing.
 
     Returns:
         Tuple containing (number of documents indexed, error message or None)
@@ -188,6 +198,9 @@ async def index_linear_issues(
         documents_skipped = 0
         skipped_issues = []
 
+        # Heartbeat tracking - update notification periodically to prevent appearing stuck
+        last_heartbeat_time = time.time()
+
         await task_logger.log_task_progress(
             log_entry,
             f"Starting to process {len(issues)} Linear issues",
@@ -196,6 +209,14 @@ async def index_linear_issues(
 
         # Process each issue
         for issue in issues:
+            # Check if it's time for a heartbeat update
+            if (
+                on_heartbeat_callback
+                and (time.time() - last_heartbeat_time) >= HEARTBEAT_INTERVAL_SECONDS
+            ):
+                await on_heartbeat_callback(documents_indexed)
+                last_heartbeat_time = time.time()
+
             try:
                 issue_id = issue.get("id", "")
                 issue_identifier = issue.get("identifier", "")

surfsense_backend/app/tasks/connector_indexers/luma_indexer.py

@@ -2,6 +2,8 @@
 Luma connector indexer.
 """
 
+import time
+from collections.abc import Awaitable, Callable
 from datetime import datetime, timedelta
 
 from sqlalchemy.exc import SQLAlchemyError
@@ -28,6 +30,12 @@ from .base import (
     update_connector_last_indexed,
 )
 
+# Type hint for heartbeat callback
+HeartbeatCallbackType = Callable[[int], Awaitable[None]]
+
+# Heartbeat interval in seconds
+HEARTBEAT_INTERVAL_SECONDS = 30
+
 
 async def index_luma_events(
     session: AsyncSession,
@@ -37,6 +45,7 @@ async def index_luma_events(
     start_date: str | None = None,
     end_date: str | None = None,
     update_last_indexed: bool = True,
+    on_heartbeat_callback: HeartbeatCallbackType | None = None,
 ) -> tuple[int, str | None]:
     """
     Index Luma events.
@@ -50,6 +59,7 @@ async def index_luma_events(
         end_date: End date for indexing (YYYY-MM-DD format). Can be in the future to index upcoming events.
                   Defaults to today if not provided.
         update_last_indexed: Whether to update the last_indexed_at timestamp (default: True)
+        on_heartbeat_callback: Optional callback to update notification during long-running indexing.
 
     Returns:
         Tuple containing (number of documents indexed, error message or None)
@@ -221,7 +231,17 @@ async def index_luma_events(
         documents_skipped = 0
         skipped_events = []
 
+        # Heartbeat tracking - update notification periodically to prevent appearing stuck
+        last_heartbeat_time = time.time()
+
         for event in events:
+            # Check if it's time for a heartbeat update
+            if (
+                on_heartbeat_callback
+                and (time.time() - last_heartbeat_time) >= HEARTBEAT_INTERVAL_SECONDS
+            ):
+                await on_heartbeat_callback(documents_indexed)
+                last_heartbeat_time = time.time()
             try:
                 # Luma event structure fields - events have nested 'event' field
                 event_data = event.get("event", {})

surfsense_backend/app/tasks/connector_indexers/notion_indexer.py

@@ -2,6 +2,7 @@
 Notion connector indexer.
 """
 
+import time
 from collections.abc import Awaitable, Callable
 from datetime import datetime
 
@@ -34,6 +35,13 @@ from .base import (
 # Signature: async callback(retry_reason, attempt, max_attempts, wait_seconds) -> None
 RetryCallbackType = Callable[[str, int, int, float], Awaitable[None]]
 
+# Type alias for heartbeat callback
+# Signature: async callback(indexed_count) -> None
+HeartbeatCallbackType = Callable[[int], Awaitable[None]]
+
+# Heartbeat interval in seconds - update notification every 30 seconds
+HEARTBEAT_INTERVAL_SECONDS = 30
+
 
 async def index_notion_pages(
     session: AsyncSession,
@@ -44,6 +52,7 @@ async def index_notion_pages(
     end_date: str | None = None,
     update_last_indexed: bool = True,
     on_retry_callback: RetryCallbackType | None = None,
+    on_heartbeat_callback: HeartbeatCallbackType | None = None,
 ) -> tuple[int, str | None]:
     """
     Index Notion pages from all accessible pages.
@@ -59,6 +68,8 @@ async def index_notion_pages(
         on_retry_callback: Optional callback for retry progress notifications.
             Signature: async callback(retry_reason, attempt, max_attempts, wait_seconds)
             retry_reason is one of: 'rate_limit', 'server_error', 'timeout'
+        on_heartbeat_callback: Optional callback to update notification during long-running indexing.
+            Called periodically with (indexed_count) to prevent task appearing stuck.
 
     Returns:
         Tuple containing (number of documents indexed, error message or None)
@@ -211,6 +222,9 @@ async def index_notion_pages(
         documents_skipped = 0
         skipped_pages = []
 
+        # Heartbeat tracking - update notification periodically to prevent appearing stuck
+        last_heartbeat_time = time.time()
+
         await task_logger.log_task_progress(
             log_entry,
             f"Starting to process {len(pages)} Notion pages",
@@ -219,6 +233,14 @@ async def index_notion_pages(
 
         # Process each page
         for page in pages:
+            # Check if it's time for a heartbeat update
+            if (
+                on_heartbeat_callback
+                and (time.time() - last_heartbeat_time) >= HEARTBEAT_INTERVAL_SECONDS
+            ):
+                await on_heartbeat_callback(documents_indexed)
+                last_heartbeat_time = time.time()
+
             try:
                 page_id = page.get("page_id")
                 page_title = page.get("title", f"Untitled page ({page_id})")

surfsense_backend/app/tasks/connector_indexers/obsidian_indexer.py

@@ -7,6 +7,8 @@ This connector is only available in self-hosted mode.
 
 import os
 import re
+import time
+from collections.abc import Awaitable, Callable
 from datetime import UTC, datetime
 from pathlib import Path
 
@@ -35,6 +37,12 @@ from .base import (
     update_connector_last_indexed,
 )
 
+# Type hint for heartbeat callback
+HeartbeatCallbackType = Callable[[int], Awaitable[None]]
+
+# Heartbeat interval in seconds
+HEARTBEAT_INTERVAL_SECONDS = 30
+
 
 def parse_frontmatter(content: str) -> tuple[dict | None, str]:
     """
@@ -152,6 +160,7 @@ async def index_obsidian_vault(
     start_date: str | None = None,
     end_date: str | None = None,
     update_last_indexed: bool = True,
+    on_heartbeat_callback: HeartbeatCallbackType | None = None,
 ) -> tuple[int, str | None]:
     """
     Index notes from a local Obsidian vault.
@@ -167,6 +176,7 @@ async def index_obsidian_vault(
         start_date: Start date for filtering (YYYY-MM-DD format) - optional
         end_date: End date for filtering (YYYY-MM-DD format) - optional
         update_last_indexed: Whether to update the last_indexed_at timestamp
+        on_heartbeat_callback: Optional callback to update notification during long-running indexing.
 
     Returns:
         Tuple containing (number of documents indexed, error message or None)
@@ -305,7 +315,17 @@ async def index_obsidian_vault(
         indexed_count = 0
         skipped_count = 0
 
+        # Heartbeat tracking - update notification periodically to prevent appearing stuck
+        last_heartbeat_time = time.time()
+
         for file_info in files:
+            # Check if it's time for a heartbeat update
+            if (
+                on_heartbeat_callback
+                and (time.time() - last_heartbeat_time) >= HEARTBEAT_INTERVAL_SECONDS
+            ):
+                await on_heartbeat_callback(indexed_count)
+                last_heartbeat_time = time.time()
             try:
                 file_path = file_info["path"]
                 relative_path = file_info["relative_path"]

surfsense_backend/app/tasks/connector_indexers/slack_indexer.py

@@ -2,6 +2,8 @@
 Slack connector indexer.
 """
 
+import time
+from collections.abc import Awaitable, Callable
 from datetime import datetime
 
 from slack_sdk.errors import SlackApiError
@@ -29,6 +31,12 @@ from .base import (
     update_connector_last_indexed,
 )
 
+# Type hint for heartbeat callback
+HeartbeatCallbackType = Callable[[int], Awaitable[None]]
+
+# Heartbeat interval in seconds - update notification every 30 seconds
+HEARTBEAT_INTERVAL_SECONDS = 30
+
 
 async def index_slack_messages(
     session: AsyncSession,
@@ -38,6 +46,7 @@ async def index_slack_messages(
     start_date: str | None = None,
     end_date: str | None = None,
     update_last_indexed: bool = True,
+    on_heartbeat_callback: HeartbeatCallbackType | None = None,
 ) -> tuple[int, str | None]:
     """
     Index Slack messages from all accessible channels.
@@ -50,6 +59,8 @@ async def index_slack_messages(
         start_date: Start date for indexing (YYYY-MM-DD format)
         end_date: End date for indexing (YYYY-MM-DD format)
         update_last_indexed: Whether to update the last_indexed_at timestamp (default: True)
+        on_heartbeat_callback: Optional callback to update notification during long-running indexing.
+            Called periodically with (indexed_count) to prevent task appearing stuck.
 
     Returns:
         Tuple containing (number of documents indexed, error message or None)
@@ -164,6 +175,9 @@ async def index_slack_messages(
         documents_skipped = 0
         skipped_channels = []
 
+        # Heartbeat tracking - update notification periodically to prevent appearing stuck
+        last_heartbeat_time = time.time()
+
         await task_logger.log_task_progress(
             log_entry,
             f"Starting to process {len(channels)} Slack channels",
@@ -172,6 +186,13 @@ async def index_slack_messages(
 
         # Process each channel
         for channel_obj in channels:
+            # Check if it's time for a heartbeat update
+            if (
+                on_heartbeat_callback
+                and (time.time() - last_heartbeat_time) >= HEARTBEAT_INTERVAL_SECONDS
+            ):
+                await on_heartbeat_callback(documents_indexed)
+                last_heartbeat_time = time.time()
             channel_id = channel_obj["id"]
             channel_name = channel_obj["name"]
             is_private = channel_obj["is_private"]

surfsense_backend/app/tasks/connector_indexers/teams_indexer.py

@@ -2,6 +2,8 @@
 Microsoft Teams connector indexer.
 """
 
+import time
+from collections.abc import Awaitable, Callable
 from datetime import UTC
 
 from sqlalchemy.exc import SQLAlchemyError
@@ -28,6 +30,12 @@ from .base import (
     update_connector_last_indexed,
 )
 
+# Type hint for heartbeat callback
+HeartbeatCallbackType = Callable[[int], Awaitable[None]]
+
+# Heartbeat interval in seconds - update notification every 30 seconds
+HEARTBEAT_INTERVAL_SECONDS = 30
+
 
 async def index_teams_messages(
     session: AsyncSession,
@@ -37,6 +45,7 @@ async def index_teams_messages(
     start_date: str | None = None,
     end_date: str | None = None,
     update_last_indexed: bool = True,
+    on_heartbeat_callback: HeartbeatCallbackType | None = None,
 ) -> tuple[int, str | None]:
     """
     Index Microsoft Teams messages from all accessible teams and channels.
@@ -49,6 +58,8 @@ async def index_teams_messages(
         start_date: Start date for indexing (YYYY-MM-DD format)
         end_date: End date for indexing (YYYY-MM-DD format)
         update_last_indexed: Whether to update the last_indexed_at timestamp (default: True)
+        on_heartbeat_callback: Optional callback to update notification during long-running indexing.
+            Called periodically with (indexed_count) to prevent task appearing stuck.
 
     Returns:
         Tuple containing (number of documents indexed, error message or None)
@@ -161,6 +172,9 @@ async def index_teams_messages(
         documents_skipped = 0
         skipped_channels = []
 
+        # Heartbeat tracking - update notification periodically to prevent appearing stuck
+        last_heartbeat_time = time.time()
+
         await task_logger.log_task_progress(
             log_entry,
             f"Starting to process {len(teams)} Teams",
@@ -185,6 +199,14 @@ async def index_teams_messages(
 
         # Process each team
         for team in teams:
+            # Check if it's time for a heartbeat update
+            if (
+                on_heartbeat_callback
+                and (time.time() - last_heartbeat_time) >= HEARTBEAT_INTERVAL_SECONDS
+            ):
+                await on_heartbeat_callback(documents_indexed)
+                last_heartbeat_time = time.time()
+
             team_id = team.get("id")
             team_name = team.get("displayName", "Unknown Team")
 

surfsense_backend/app/tasks/connector_indexers/webcrawler_indexer.py

@@ -2,6 +2,8 @@
 Webcrawler connector indexer.
 """
 
+import time
+from collections.abc import Awaitable, Callable
 from datetime import datetime
 
 from sqlalchemy.exc import SQLAlchemyError
@@ -29,6 +31,12 @@ from .base import (
     update_connector_last_indexed,
 )
 
+# Type hint for heartbeat callback
+HeartbeatCallbackType = Callable[[int], Awaitable[None]]
+
+# Heartbeat interval in seconds
+HEARTBEAT_INTERVAL_SECONDS = 30
+
 
 async def index_crawled_urls(
     session: AsyncSession,
@@ -38,6 +46,7 @@ async def index_crawled_urls(
     start_date: str | None = None,
     end_date: str | None = None,
     update_last_indexed: bool = True,
+    on_heartbeat_callback: HeartbeatCallbackType | None = None,
 ) -> tuple[int, str | None]:
     """
     Index web page URLs.
@@ -50,6 +59,7 @@ async def index_crawled_urls(
         start_date: Start date for filtering (YYYY-MM-DD format) - optional
         end_date: End date for filtering (YYYY-MM-DD format) - optional
         update_last_indexed: Whether to update the last_indexed_at timestamp (default: True)
+        on_heartbeat_callback: Optional callback to update notification during long-running indexing.
 
     Returns:
         Tuple containing (number of documents indexed, error message or None)
@@ -140,7 +150,17 @@ async def index_crawled_urls(
         documents_skipped = 0
         failed_urls = []
 
+        # Heartbeat tracking - update notification periodically to prevent appearing stuck
+        last_heartbeat_time = time.time()
+
         for idx, url in enumerate(urls, 1):
+            # Check if it's time for a heartbeat update
+            if (
+                on_heartbeat_callback
+                and (time.time() - last_heartbeat_time) >= HEARTBEAT_INTERVAL_SECONDS
+            ):
+                await on_heartbeat_callback(documents_indexed)
+                last_heartbeat_time = time.time()
             try:
                 logger.info(f"Processing URL {idx}/{len(urls)}: {url}")
 

surfsense_web/components/assistant-ui/connector-popup.tsx

@@ -240,6 +240,10 @@ export const ConnectorIndicator: FC = () => {
 							...editingConnector,
 							config: connectorConfig || editingConnector.config,
 							name: editingConnector.name,
+							// Sync last_indexed_at with live data from Electric SQL for real-time updates
+							last_indexed_at:
+								(connectors as SearchSourceConnector[]).find((c) => c.id === editingConnector.id)
+									?.last_indexed_at ?? editingConnector.last_indexed_at,
 						}}
 						startDate={startDate}
 						endDate={endDate}

surfsense_web/components/assistant-ui/connector-popup/components/date-range-selector.tsx

@@ -7,6 +7,7 @@ import { Button } from "@/components/ui/button";
 import { Calendar } from "@/components/ui/calendar";
 import { Label } from "@/components/ui/label";
 import { Popover, PopoverContent, PopoverTrigger } from "@/components/ui/popover";
+import { formatRelativeDate } from "@/lib/format-date";
 import { cn } from "@/lib/utils";
 
 interface DateRangeSelectorProps {
@@ -26,19 +27,10 @@ export const DateRangeSelector: FC<DateRangeSelectorProps> = ({
 	allowFutureDates = false,
 	lastIndexedAt,
 }) => {
-	// Get the placeholder text for start date based on whether connector was previously indexed
-	const getStartDatePlaceholder = () => {
-		if (lastIndexedAt) {
-			const date = new Date(lastIndexedAt);
-			const currentYear = new Date().getFullYear();
-			const indexedYear = date.getFullYear();
-			// Show year only if different from current year
-			const formatStr = indexedYear === currentYear ? "MMM d, HH:mm" : "MMM d, yyyy HH:mm";
-			const formattedDate = format(date, formatStr);
-			return `Since (${formattedDate})`;
-		}
-		return "Default (1 year ago)";
-	};
+	const startDatePlaceholder = lastIndexedAt
+		? `From ${formatRelativeDate(lastIndexedAt)}`
+		: "Default (1 year)";
+
 	const handleLast30Days = () => {
 		const today = new Date();
 		onStartDateChange(subDays(today, 30));
@@ -88,7 +80,7 @@ export const DateRangeSelector: FC<DateRangeSelectorProps> = ({
 								)}
 							>
 								<CalendarIcon className="mr-2 h-4 w-4" />
-								{startDate ? format(startDate, "PPP") : getStartDatePlaceholder()}
+								{startDate ? format(startDate, "PPP") : startDatePlaceholder}
 							</Button>
 						</PopoverTrigger>
 						<PopoverContent className="w-auto p-0 z-[100]" align="start">

surfsense_web/components/assistant-ui/connector-popup/connector-configs/components/discord-config.tsx

@@ -1,29 +1,188 @@
 "use client";
 
-import { Info } from "lucide-react";
-import type { FC } from "react";
+import { AlertCircle, CheckCircle2, Hash, Info, Megaphone, RefreshCw } from "lucide-react";
+import { type FC, useCallback, useEffect, useState } from "react";
+import { Button } from "@/components/ui/button";
+import { Spinner } from "@/components/ui/spinner";
+import { connectorsApiService, type DiscordChannel } from "@/lib/apis/connectors-api.service";
+import { cn } from "@/lib/utils";
 import type { ConnectorConfigProps } from "../index";
 
 export interface DiscordConfigProps extends ConnectorConfigProps {
 	onNameChange?: (name: string) => void;
 }
 
-export const DiscordConfig: FC<DiscordConfigProps> = () => {
+export const DiscordConfig: FC<DiscordConfigProps> = ({ connector }) => {
+	const [channels, setChannels] = useState<DiscordChannel[]>([]);
+	const [isLoading, setIsLoading] = useState(false);
+	const [error, setError] = useState<string | null>(null);
+	const [lastFetched, setLastFetched] = useState<Date | null>(null);
+
+	const fetchChannels = useCallback(async () => {
+		if (!connector?.id) return;
+
+		setIsLoading(true);
+		setError(null);
+
+		try {
+			const data = await connectorsApiService.getDiscordChannels(connector.id);
+			setChannels(data);
+			setLastFetched(new Date());
+		} catch (err) {
+			console.error("Failed to fetch Discord channels:", err);
+			setError(err instanceof Error ? err.message : "Failed to fetch channels");
+		} finally {
+			setIsLoading(false);
+		}
+	}, [connector?.id]);
+
+	// Fetch channels on mount
+	useEffect(() => {
+		fetchChannels();
+	}, [fetchChannels]);
+
+	// Auto-refresh when user returns to tab
+	useEffect(() => {
+		const handleVisibilityChange = () => {
+			if (document.visibilityState === "visible" && connector?.id) {
+				fetchChannels();
+			}
+		};
+
+		document.addEventListener("visibilitychange", handleVisibilityChange);
+		return () => document.removeEventListener("visibilitychange", handleVisibilityChange);
+	}, [connector?.id, fetchChannels]);
+
+	// Separate channels by indexing capability
+	const readyToIndex = channels.filter((ch) => ch.can_index);
+	const needsPermissions = channels.filter((ch) => !ch.can_index);
+
+	// Format last fetched time
+	const formatLastFetched = () => {
+		if (!lastFetched) return null;
+		const now = new Date();
+		const diffMs = now.getTime() - lastFetched.getTime();
+		const diffSecs = Math.floor(diffMs / 1000);
+		const diffMins = Math.floor(diffSecs / 60);
+
+		if (diffSecs < 60) return "just now";
+		if (diffMins === 1) return "1 minute ago";
+		if (diffMins < 60) return `${diffMins} minutes ago`;
+		return lastFetched.toLocaleTimeString();
+	};
+
 	return (
 		<div className="space-y-6">
+			{/* Info box */}
 			<div className="rounded-xl border border-border bg-primary/5 p-4 flex items-start gap-3">
 				<div className="flex h-8 w-8 items-center justify-center rounded-lg bg-primary/10 shrink-0 mt-0.5">
 					<Info className="size-4" />
 				</div>
 				<div className="text-xs sm:text-sm">
-					<p className="font-medium text-xs sm:text-sm">Add Bot to Servers</p>
 					<p className="text-muted-foreground mt-1 text-[10px] sm:text-sm">
-						Before indexing, make sure the Discord bot has been added to the servers (guilds) you
-						want to index. The bot can only access messages from servers it's been added to. Use the
-						OAuth authorization flow to add the bot to your servers.
+						The bot needs &quot;Read Message History&quot; permission to index channels. Ask a
+						server admin to grant this permission for channels shown below.
 					</p>
 				</div>
 			</div>
+
+			{/* Channels Section */}
+			<div className="space-y-3">
+				<div className="flex items-center justify-between">
+					<div className="flex items-center gap-3">
+						<h3 className="text-sm font-semibold">Channel Access</h3>
+					</div>
+					<div className="flex items-center gap-2">
+						{lastFetched && (
+							<span className="text-[10px] text-muted-foreground">{formatLastFetched()}</span>
+						)}
+						<Button
+							variant="secondary"
+							size="sm"
+							onClick={fetchChannels}
+							disabled={isLoading}
+							className="h-7 px-2.5 text-[11px] bg-slate-400/10 dark:bg-white/10 hover:bg-slate-400/20 dark:hover:bg-white/20 border-slate-400/20 dark:border-white/20"
+						>
+							<RefreshCw className={cn("mr-1.5 size-3", isLoading && "animate-spin")} />
+							Refresh
+						</Button>
+					</div>
+				</div>
+
+				{error && (
+					<div className="rounded-lg border border-destructive/50 bg-destructive/10 p-3 text-xs text-destructive">
+						{error}
+					</div>
+				)}
+
+				{isLoading && channels.length === 0 ? (
+					<div className="flex items-center justify-center py-8">
+						<Spinner size="sm" />
+						<span className="ml-2 text-sm text-muted-foreground">Loading channels</span>
+					</div>
+				) : channels.length === 0 && !error ? (
+					<div className="text-center py-8 text-sm text-muted-foreground">
+						No channels found. Make sure the bot has been added to your Discord server with proper
+						permissions.
+					</div>
+				) : (
+					<div className="rounded-xl bg-slate-400/5 dark:bg-white/5 overflow-hidden">
+						{/* Ready to index */}
+						{readyToIndex.length > 0 && (
+							<div className={cn("p-3", needsPermissions.length > 0 && "border-b border-border")}>
+								<div className="flex items-center gap-2 mb-2">
+									<CheckCircle2 className="size-3.5 text-emerald-500" />
+									<span className="text-[11px] font-medium">Ready to index</span>
+									<span className="text-[10px] text-muted-foreground">
+										{readyToIndex.length} {readyToIndex.length === 1 ? "channel" : "channels"}
+									</span>
+								</div>
+								<div className="flex flex-wrap gap-1.5">
+									{readyToIndex.map((channel) => (
+										<ChannelPill key={channel.id} channel={channel} />
+									))}
+								</div>
+							</div>
+						)}
+
+						{/* Needs permissions */}
+						{needsPermissions.length > 0 && (
+							<div className="p-3">
+								<div className="flex items-center gap-2 mb-2">
+									<AlertCircle className="size-3.5 text-amber-500" />
+									<span className="text-[11px] font-medium">Grant permissions to index</span>
+									<span className="text-[10px] text-muted-foreground">
+										{needsPermissions.length}{" "}
+										{needsPermissions.length === 1 ? "channel" : "channels"}
+									</span>
+								</div>
+								<div className="flex flex-wrap gap-1.5">
+									{needsPermissions.map((channel) => (
+										<ChannelPill key={channel.id} channel={channel} />
+									))}
+								</div>
+							</div>
+						)}
+					</div>
+				)}
+			</div>
+		</div>
+	);
+};
+
+interface ChannelPillProps {
+	channel: DiscordChannel;
+}
+
+const ChannelPill: FC<ChannelPillProps> = ({ channel }) => {
+	return (
+		<div className="inline-flex items-center gap-1 px-2 py-1 rounded-md text-[11px] font-medium bg-slate-400/10 dark:bg-white/10 hover:bg-slate-400/20 dark:hover:bg-white/20 transition-colors">
+			{channel.type === "announcement" ? (
+				<Megaphone className="size-2.5 text-muted-foreground" />
+			) : (
+				<Hash className="size-2.5 text-muted-foreground" />
+			)}
+			<span>{channel.name}</span>
 		</div>
 	);
 };

surfsense_web/components/assistant-ui/connector-popup/connector-configs/components/slack-config.tsx

@@ -1,16 +1,79 @@
 "use client";
 
-import { Info } from "lucide-react";
-import type { FC } from "react";
+import { AlertCircle, CheckCircle2, Hash, Info, Lock, RefreshCw } from "lucide-react";
+import { type FC, useCallback, useEffect, useState } from "react";
+import { Button } from "@/components/ui/button";
+import { Spinner } from "@/components/ui/spinner";
+import { connectorsApiService, type SlackChannel } from "@/lib/apis/connectors-api.service";
+import { cn } from "@/lib/utils";
 import type { ConnectorConfigProps } from "../index";
 
 export interface SlackConfigProps extends ConnectorConfigProps {
 	onNameChange?: (name: string) => void;
 }
 
-export const SlackConfig: FC<SlackConfigProps> = () => {
+export const SlackConfig: FC<SlackConfigProps> = ({ connector }) => {
+	const [channels, setChannels] = useState<SlackChannel[]>([]);
+	const [isLoading, setIsLoading] = useState(false);
+	const [error, setError] = useState<string | null>(null);
+	const [lastFetched, setLastFetched] = useState<Date | null>(null);
+
+	const fetchChannels = useCallback(async () => {
+		if (!connector?.id) return;
+
+		setIsLoading(true);
+		setError(null);
+
+		try {
+			const data = await connectorsApiService.getSlackChannels(connector.id);
+			setChannels(data);
+			setLastFetched(new Date());
+		} catch (err) {
+			console.error("Failed to fetch Slack channels:", err);
+			setError(err instanceof Error ? err.message : "Failed to fetch channels");
+		} finally {
+			setIsLoading(false);
+		}
+	}, [connector?.id]);
+
+	// Fetch channels on mount
+	useEffect(() => {
+		fetchChannels();
+	}, [fetchChannels]);
+
+	// Auto-refresh when user returns to tab
+	useEffect(() => {
+		const handleVisibilityChange = () => {
+			if (document.visibilityState === "visible" && connector?.id) {
+				fetchChannels();
+			}
+		};
+
+		document.addEventListener("visibilitychange", handleVisibilityChange);
+		return () => document.removeEventListener("visibilitychange", handleVisibilityChange);
+	}, [connector?.id, fetchChannels]);
+
+	// Separate channels by bot membership
+	const channelsWithBot = channels.filter((ch) => ch.is_member);
+	const channelsWithoutBot = channels.filter((ch) => !ch.is_member);
+
+	// Format last fetched time
+	const formatLastFetched = () => {
+		if (!lastFetched) return null;
+		const now = new Date();
+		const diffMs = now.getTime() - lastFetched.getTime();
+		const diffSecs = Math.floor(diffMs / 1000);
+		const diffMins = Math.floor(diffSecs / 60);
+
+		if (diffSecs < 60) return "just now";
+		if (diffMins === 1) return "1 minute ago";
+		if (diffMins < 60) return `${diffMins} minutes ago`;
+		return lastFetched.toLocaleTimeString();
+	};
+
 	return (
 		<div className="space-y-6">
+			{/* Info box */}
 			<div className="rounded-xl border border-border bg-primary/5 p-4 flex items-start gap-3">
 				<div className="flex h-8 w-8 items-center justify-center rounded-lg bg-primary/10 shrink-0 mt-0.5">
 					<Info className="size-4" />
@@ -25,6 +88,103 @@ export const SlackConfig: FC<SlackConfigProps> = () => {
 					</p>
 				</div>
 			</div>
+
+			{/* Channels Section */}
+			<div className="space-y-3">
+				<div className="flex items-center justify-between">
+					<div className="flex items-center gap-3">
+						<h3 className="text-sm font-semibold">Channel Access</h3>
+					</div>
+					<div className="flex items-center gap-2">
+						{lastFetched && (
+							<span className="text-[10px] text-muted-foreground">{formatLastFetched()}</span>
+						)}
+						<Button
+							variant="secondary"
+							size="sm"
+							onClick={fetchChannels}
+							disabled={isLoading}
+							className="h-7 px-2.5 text-[11px] bg-slate-400/10 dark:bg-white/10 hover:bg-slate-400/20 dark:hover:bg-white/20 border-slate-400/20 dark:border-white/20"
+						>
+							<RefreshCw className={cn("mr-1.5 size-3", isLoading && "animate-spin")} />
+							Refresh
+						</Button>
+					</div>
+				</div>
+
+				{error && (
+					<div className="rounded-lg border border-destructive/50 bg-destructive/10 p-3 text-xs text-destructive">
+						{error}
+					</div>
+				)}
+
+				{isLoading && channels.length === 0 ? (
+					<div className="flex items-center justify-center py-8">
+						<Spinner size="sm" />
+						<span className="ml-2 text-sm text-muted-foreground">Loading channels</span>
+					</div>
+				) : channels.length === 0 && !error ? (
+					<div className="text-center py-8 text-sm text-muted-foreground">
+						No channels found. Make sure the bot has been added to your Slack workspace.
+					</div>
+				) : (
+					<div className="rounded-xl bg-slate-400/5 dark:bg-white/5 overflow-hidden">
+						{/* Channels with bot access */}
+						{channelsWithBot.length > 0 && (
+							<div className={cn("p-3", channelsWithoutBot.length > 0 && "border-b border-border")}>
+								<div className="flex items-center gap-2 mb-2">
+									<CheckCircle2 className="size-3.5 text-emerald-500" />
+									<span className="text-[11px] font-medium">Ready to index</span>
+									<span className="text-[10px] text-muted-foreground">
+										{channelsWithBot.length} {channelsWithBot.length === 1 ? "channel" : "channels"}
+									</span>
+								</div>
+								<div className="flex flex-wrap gap-1.5">
+									{channelsWithBot.map((channel) => (
+										<ChannelPill key={channel.id} channel={channel} />
+									))}
+								</div>
+							</div>
+						)}
+
+						{/* Channels without bot access */}
+						{channelsWithoutBot.length > 0 && (
+							<div className="p-3">
+								<div className="flex items-center gap-2 mb-2">
+									<AlertCircle className="size-3.5 text-amber-500" />
+									<span className="text-[11px] font-medium">Add bot to index</span>
+									<span className="text-[10px] text-muted-foreground">
+										{channelsWithoutBot.length}{" "}
+										{channelsWithoutBot.length === 1 ? "channel" : "channels"}
+									</span>
+								</div>
+								<div className="flex flex-wrap gap-1.5">
+									{channelsWithoutBot.map((channel) => (
+										<ChannelPill key={channel.id} channel={channel} />
+									))}
+								</div>
+							</div>
+						)}
+					</div>
+				)}
+			</div>
+		</div>
+	);
+};
+
+interface ChannelPillProps {
+	channel: SlackChannel;
+}
+
+const ChannelPill: FC<ChannelPillProps> = ({ channel }) => {
+	return (
+		<div className="inline-flex items-center gap-1 px-2 py-1 rounded-md text-[11px] font-medium bg-slate-400/10 dark:bg-white/10 hover:bg-slate-400/20 dark:hover:bg-white/20 transition-colors">
+			{channel.is_private ? (
+				<Lock className="size-2.5 text-muted-foreground" />
+			) : (
+				<Hash className="size-2.5 text-muted-foreground" />
+			)}
+			<span>{channel.name}</span>
 		</div>
 	);
 };

surfsense_web/components/assistant-ui/connector-popup/hooks/use-indexing-connectors.ts

@@ -5,6 +5,44 @@ import type { SearchSourceConnector } from "@/contracts/types/connector.types";
 import type { InboxItem } from "@/contracts/types/inbox.types";
 import { isConnectorIndexingMetadata } from "@/contracts/types/inbox.types";
 
+/**
+ * Timeout thresholds for stuck task detection
+ *
+ * These align with the backend Celery configuration:
+ * - HARD_TIMEOUT: 8 hours (task_time_limit=28800 in Celery)
+ *   Any task running longer than this is definitely dead.
+ *
+ * - STALE_THRESHOLD: 15 minutes without notification updates
+ *   If heartbeats are being sent every 30s, missing 15+ minutes of updates
+ *   indicates the task has likely crashed or the worker is down.
+ */
+const HARD_TIMEOUT_MS = 8 * 60 * 60 * 1000; // 8 hours in milliseconds
+const STALE_THRESHOLD_MS = 15 * 60 * 1000; // 15 minutes in milliseconds
+
+/**
+ * Check if a notification is stale (no updates for too long)
+ * @param updatedAt - ISO timestamp of last notification update
+ * @returns true if the notification hasn't been updated recently
+ */
+function isNotificationStale(updatedAt: string | null | undefined): boolean {
+	if (!updatedAt) return false;
+	const lastUpdate = new Date(updatedAt).getTime();
+	const now = Date.now();
+	return now - lastUpdate > STALE_THRESHOLD_MS;
+}
+
+/**
+ * Check if a task has exceeded the hard timeout (definitely dead)
+ * @param startedAt - ISO timestamp when the task started
+ * @returns true if the task has been running longer than the hard limit
+ */
+function isTaskTimedOut(startedAt: string | null | undefined): boolean {
+	if (!startedAt) return false;
+	const startTime = new Date(startedAt).getTime();
+	const now = Date.now();
+	return now - startTime > HARD_TIMEOUT_MS;
+}
+
 /**
  * Hook to track which connectors are currently indexing using local state.
  *
@@ -13,6 +51,8 @@ import { isConnectorIndexingMetadata } from "@/contracts/types/inbox.types";
  * 2. Detecting in_progress notifications from Electric SQL to restore state after remounts
  * 3. Clearing indexing state when notifications become completed or failed
  * 4. Clearing indexing state when Electric SQL detects last_indexed_at changed
+ * 5. Detecting stale/stuck tasks that haven't updated in 15+ minutes
+ * 6. Detecting hard timeout (8h) - tasks that definitely cannot still be running
  *
  * The actual `last_indexed_at` value comes from Electric SQL/PGlite, not local state.
  */
@@ -57,6 +97,7 @@ export function useIndexingConnectors(
 
 	// Detect notification status changes and update indexing state accordingly
 	// This restores spinner state after component remounts and handles all status transitions
+	// Also detects stale/stuck tasks that haven't been updated in a while
 	useEffect(() => {
 		if (!inboxItems || inboxItems.length === 0) return;
 
@@ -71,11 +112,26 @@ export function useIndexingConnectors(
 				const metadata = isConnectorIndexingMetadata(item.metadata) ? item.metadata : null;
 				if (!metadata) continue;
 
-				// If status is "in_progress", add connector to indexing set
+				// If status is "in_progress", check if it's actually still running
 				if (metadata.status === "in_progress") {
-					if (!newIndexingIds.has(metadata.connector_id)) {
-						newIndexingIds.add(metadata.connector_id);
-						hasChanges = true;
+					// Check for hard timeout (8h) - task is definitely dead
+					const timedOut = isTaskTimedOut(metadata.started_at);
+
+					// Check for stale notification (15min without updates) - task likely crashed
+					const stale = isNotificationStale(item.updated_at);
+
+					if (timedOut || stale) {
+						// Task is stuck - don't show as indexing
+						if (newIndexingIds.has(metadata.connector_id)) {
+							newIndexingIds.delete(metadata.connector_id);
+							hasChanges = true;
+						}
+					} else {
+						// Task appears to be genuinely running
+						if (!newIndexingIds.has(metadata.connector_id)) {
+							newIndexingIds.add(metadata.connector_id);
+							hasChanges = true;
+						}
 					}
 				}
 				// If status is "completed" or "failed", remove connector from indexing set

surfsense_web/components/assistant-ui/connector-popup/views/connector-accounts-list-view.tsx

@@ -1,6 +1,5 @@
 "use client";
 
-import { differenceInDays, differenceInMinutes, format, isToday, isYesterday } from "date-fns";
 import { ArrowLeft, Plus, Server } from "lucide-react";
 import type { FC } from "react";
 import { Button } from "@/components/ui/button";
@@ -8,6 +7,7 @@ import { Spinner } from "@/components/ui/spinner";
 import { EnumConnectorName } from "@/contracts/enums/connector";
 import { getConnectorIcon } from "@/contracts/enums/connectorIcons";
 import type { SearchSourceConnector } from "@/contracts/types/connector.types";
+import { formatRelativeDate } from "@/lib/format-date";
 import { cn } from "@/lib/utils";
 import { useConnectorStatus } from "../hooks/use-connector-status";
 import { getConnectorDisplayName } from "../tabs/all-connectors-tab";
@@ -32,38 +32,6 @@ function isIndexableConnector(connectorType: string): boolean {
 	return !nonIndexableTypes.includes(connectorType);
 }
 
-/**
- * Format last indexed date with contextual messages
- */
-function formatLastIndexedDate(dateString: string): string {
-	const date = new Date(dateString);
-	const now = new Date();
-	const minutesAgo = differenceInMinutes(now, date);
-	const daysAgo = differenceInDays(now, date);
-
-	if (minutesAgo < 1) {
-		return "Just now";
-	}
-
-	if (minutesAgo < 60) {
-		return `${minutesAgo} ${minutesAgo === 1 ? "minute" : "minutes"} ago`;
-	}
-
-	if (isToday(date)) {
-		return `Today at ${format(date, "h:mm a")}`;
-	}
-
-	if (isYesterday(date)) {
-		return `Yesterday at ${format(date, "h:mm a")}`;
-	}
-
-	if (daysAgo < 7) {
-		return `${daysAgo} ${daysAgo === 1 ? "day" : "days"} ago`;
-	}
-
-	return format(date, "MMM d, yyyy");
-}
-
 export const ConnectorAccountsListView: FC<ConnectorAccountsListViewProps> = ({
 	connectorType,
 	connectorTitle,
@@ -215,7 +183,7 @@ export const ConnectorAccountsListView: FC<ConnectorAccountsListViewProps> = ({
 											<p className="text-[10px] text-muted-foreground mt-1 whitespace-nowrap truncate">
 												{isIndexableConnector(connector.connector_type)
 													? connector.last_indexed_at
-														? `Last indexed: ${formatLastIndexedDate(connector.last_indexed_at)}`
+														? `Last indexed: ${formatRelativeDate(connector.last_indexed_at)}`
 														: "Never indexed"
 													: "Active"}
 											</p>

surfsense_web/components/layout/ui/sidebar/Sidebar.tsx

@@ -1,6 +1,6 @@
 "use client";
 
-import { FolderOpen, MessageSquare, PenSquare } from "lucide-react";
+import { FolderOpen, PenSquare } from "lucide-react";
 import { useTranslations } from "next-intl";
 import { Button } from "@/components/ui/button";
 import { Tooltip, TooltipContent, TooltipTrigger } from "@/components/ui/tooltip";
@@ -122,33 +122,15 @@ export function Sidebar({
 
 			{/* Chat sections - fills available space */}
 			{isCollapsed ? (
-				<div className="flex-1 flex flex-col items-center gap-2 py-2 w-[60px]">
-					{(chats.length > 0 || sharedChats.length > 0) && (
-						<Tooltip>
-							<TooltipTrigger asChild>
-								<Button
-									variant="ghost"
-									size="icon"
-									className="h-10 w-10"
-									onClick={() => onToggleCollapse?.()}
-								>
-									<MessageSquare className="h-4 w-4" />
-									<span className="sr-only">{t("chats")}</span>
-								</Button>
-							</TooltipTrigger>
-							<TooltipContent side="right">
-								{t("chats")} ({chats.length + sharedChats.length})
-							</TooltipContent>
-						</Tooltip>
-					)}
-				</div>
+				<div className="flex-1 w-[60px]" />
 			) : (
 				<div className="flex-1 flex flex-col gap-1 py-2 w-[240px] min-h-0 overflow-hidden">
-					{/* Shared Chats Section - takes half the space */}
+					{/* Shared Chats Section - takes only space needed, max 50% */}
 					<SidebarSection
 						title={t("shared_chats")}
 						defaultOpen={true}
-						fillHeight={true}
+						fillHeight={false}
+						className="shrink-0 max-h-[50%] flex flex-col"
 						action={
 							onViewAllSharedChats ? (
 								<Tooltip>
@@ -170,9 +152,9 @@ export function Sidebar({
 						}
 					>
 						{sharedChats.length > 0 ? (
-							<div className="relative flex-1 min-h-0">
+							<div className="relative min-h-0 flex-1">
 								<div
-									className={`flex flex-col gap-0.5 h-full overflow-y-auto scrollbar-thin scrollbar-thumb-muted-foreground/20 scrollbar-track-transparent ${sharedChats.length > 4 ? "pb-8" : ""}`}
+									className={`flex flex-col gap-0.5 max-h-full overflow-y-auto scrollbar-thin scrollbar-thumb-muted-foreground/20 scrollbar-track-transparent ${sharedChats.length > 4 ? "pb-8" : ""}`}
 								>
 									{sharedChats.slice(0, 20).map((chat) => (
 										<ChatListItem
@@ -196,7 +178,7 @@ export function Sidebar({
 						)}
 					</SidebarSection>
 
-					{/* Private Chats Section - takes half the space */}
+					{/* Private Chats Section - fills remaining space */}
 					<SidebarSection
 						title={t("chats")}
 						defaultOpen={true}

surfsense_web/components/layout/ui/sidebar/SidebarSection.tsx

@@ -30,7 +30,12 @@ export function SidebarSection({
 		<Collapsible
 			open={isOpen}
 			onOpenChange={setIsOpen}
-			className={cn("overflow-hidden", fillHeight && "flex flex-col flex-1 min-h-0", className)}
+			className={cn(
+				"overflow-hidden",
+				fillHeight && "flex flex-col min-h-0",
+				fillHeight && isOpen && "flex-1",
+				className
+			)}
 		>
 			<div className="flex items-center group/section shrink-0">
 				<CollapsibleTrigger className="flex flex-1 items-center gap-1.5 px-2 py-1.5 text-xs font-medium text-muted-foreground hover:text-foreground transition-colors min-w-0">
@@ -56,12 +61,8 @@ export function SidebarSection({
 				)}
 			</div>
 
-			<CollapsibleContent
-				className={cn("overflow-hidden", fillHeight && "flex-1 flex flex-col min-h-0")}
-			>
-				<div
-					className={cn("px-2 pb-2", fillHeight && "flex-1 flex flex-col min-h-0 overflow-hidden")}
-				>
+			<CollapsibleContent className={cn("overflow-hidden flex-1 flex flex-col min-h-0")}>
+				<div className={cn("px-2 pb-2 flex-1 flex flex-col min-h-0 overflow-hidden")}>
 					{children}
 				</div>
 			</CollapsibleContent>

surfsense_web/contracts/types/connector.types.ts

@@ -203,6 +203,46 @@ export const listGoogleDriveFoldersResponse = z.object({
 	items: z.array(googleDriveItem),
 });
 
+/**
+ * Slack channel with bot membership status
+ */
+export const slackChannel = z.object({
+	id: z.string(),
+	name: z.string(),
+	is_private: z.boolean(),
+	is_member: z.boolean(),
+});
+
+/**
+ * List Slack channels
+ */
+export const listSlackChannelsRequest = z.object({
+	connector_id: z.number(),
+});
+
+export const listSlackChannelsResponse = z.array(slackChannel);
+
+/**
+ * Discord channel with indexing permission info
+ */
+export const discordChannel = z.object({
+	id: z.string(),
+	name: z.string(),
+	type: z.enum(["text", "announcement"]),
+	position: z.number(),
+	category_id: z.string().nullable().optional(),
+	can_index: z.boolean(),
+});
+
+/**
+ * List Discord channels
+ */
+export const listDiscordChannelsRequest = z.object({
+	connector_id: z.number(),
+});
+
+export const listDiscordChannelsResponse = z.array(discordChannel);
+
 // Inferred types
 export type SearchSourceConnectorType = z.infer<typeof searchSourceConnectorTypeEnum>;
 export type SearchSourceConnector = z.infer<typeof searchSourceConnector>;
@@ -223,3 +263,9 @@ export type ListGitHubRepositoriesResponse = z.infer<typeof listGitHubRepositori
 export type ListGoogleDriveFoldersRequest = z.infer<typeof listGoogleDriveFoldersRequest>;
 export type ListGoogleDriveFoldersResponse = z.infer<typeof listGoogleDriveFoldersResponse>;
 export type GoogleDriveItem = z.infer<typeof googleDriveItem>;
+export type SlackChannel = z.infer<typeof slackChannel>;
+export type ListSlackChannelsRequest = z.infer<typeof listSlackChannelsRequest>;
+export type ListSlackChannelsResponse = z.infer<typeof listSlackChannelsResponse>;
+export type DiscordChannel = z.infer<typeof discordChannel>;
+export type ListDiscordChannelsRequest = z.infer<typeof listDiscordChannelsRequest>;
+export type ListDiscordChannelsResponse = z.infer<typeof listDiscordChannelsResponse>;

surfsense_web/lib/apis/connectors-api.service.ts

@@ -5,6 +5,7 @@ import {
 	type DeleteConnectorRequest,
 	deleteConnectorRequest,
 	deleteConnectorResponse,
+	type DiscordChannel,
 	type GetConnectorRequest,
 	type GetConnectorsRequest,
 	getConnectorRequest,
@@ -16,10 +17,13 @@ import {
 	indexConnectorResponse,
 	type ListGitHubRepositoriesRequest,
 	type ListGoogleDriveFoldersRequest,
+	listDiscordChannelsResponse,
 	listGitHubRepositoriesRequest,
 	listGitHubRepositoriesResponse,
 	listGoogleDriveFoldersRequest,
 	listGoogleDriveFoldersResponse,
+	listSlackChannelsResponse,
+	type SlackChannel,
 	type UpdateConnectorRequest,
 	updateConnectorRequest,
 	updateConnectorResponse,
@@ -335,6 +339,36 @@ class ConnectorsApiService {
 			}
 		);
 	};
+
+	// =============================================================================
+	// Slack Connector Methods
+	// =============================================================================
+
+	/**
+	 * Get Slack channels with bot membership status
+	 */
+	getSlackChannels = async (connectorId: number) => {
+		return baseApiService.get(
+			`/api/v1/slack/connector/${connectorId}/channels`,
+			listSlackChannelsResponse
+		);
+	};
+
+	// =============================================================================
+	// Discord Connector Methods
+	// =============================================================================
+
+	/**
+	 * Get Discord text channels for a connector
+	 */
+	getDiscordChannels = async (connectorId: number) => {
+		return baseApiService.get(
+			`/api/v1/discord/connector/${connectorId}/channels`,
+			listDiscordChannelsResponse
+		);
+	};
 }
 
+export type { SlackChannel, DiscordChannel };
+
 export const connectorsApiService = new ConnectorsApiService();

surfsense_web/lib/electric/client.ts

@@ -55,7 +55,8 @@ const pendingSyncs = new Map<string, Promise<SyncHandle>>();
 // Version for sync state - increment this to force fresh sync when Electric config changes
 // v2: user-specific database architecture
 // v3: consistent cutoff date for sync+queries, visibility refresh support
-const SYNC_VERSION = 3;
+// v4: heartbeat-based stale notification detection with updated_at tracking
+const SYNC_VERSION = 4;
 
 // Database name prefix for identifying SurfSense databases
 const DB_PREFIX = "surfsense-";

surfsense_web/lib/format-date.ts

@@ -0,0 +1,24 @@
+import { differenceInDays, differenceInMinutes, format, isToday, isYesterday } from "date-fns";
+
+/**
+ * Format a date string as a human-readable relative time
+ * - < 1 min: "Just now"
+ * - < 60 min: "15m ago"
+ * - Today: "Today, 2:30 PM"
+ * - Yesterday: "Yesterday, 2:30 PM"
+ * - < 7 days: "3d ago"
+ * - Older: "Jan 15, 2026"
+ */
+export function formatRelativeDate(dateString: string): string {
+	const date = new Date(dateString);
+	const now = new Date();
+	const minutesAgo = differenceInMinutes(now, date);
+	const daysAgo = differenceInDays(now, date);
+
+	if (minutesAgo < 1) return "Just now";
+	if (minutesAgo < 60) return `${minutesAgo}m ago`;
+	if (isToday(date)) return `Today, ${format(date, "h:mm a")}`;
+	if (isYesterday(date)) return `Yesterday, ${format(date, "h:mm a")}`;
+	if (daysAgo < 7) return `${daysAgo}d ago`;
+	return format(date, "MMM d, yyyy");
+}