chore: only pause after repeated failure on cloud (#6673)

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: Wenxi Onyx <wenxi@onyx.app>

.github/workflows/pr-external-dependency-unit-tests.yml

@@ -29,6 +29,9 @@ env:
   CONFLUENCE_ACCESS_TOKEN: ${{ secrets.CONFLUENCE_ACCESS_TOKEN }}
   CONFLUENCE_ACCESS_TOKEN_SCOPED: ${{ secrets.CONFLUENCE_ACCESS_TOKEN_SCOPED }}
 
+  # Jira
+  JIRA_ADMIN_API_TOKEN: ${{ secrets.JIRA_ADMIN_API_TOKEN }}
+
   # LLMs
   OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
   ANTHROPIC_API_KEY: ${{ secrets.ANTHROPIC_API_KEY }}

backend/ee/onyx/configs/app_configs.py

@@ -41,6 +41,10 @@ CONFLUENCE_ANONYMOUS_ACCESS_IS_PUBLIC = (
 JIRA_PERMISSION_DOC_SYNC_FREQUENCY = int(
     os.environ.get("JIRA_PERMISSION_DOC_SYNC_FREQUENCY") or 30 * 60
 )
+# In seconds, default is 30 minutes
+JIRA_PERMISSION_GROUP_SYNC_FREQUENCY = int(
+    os.environ.get("JIRA_PERMISSION_GROUP_SYNC_FREQUENCY") or 30 * 60
+)
 
 
 #####

backend/ee/onyx/external_permissions/confluence/group_sync.py

@@ -3,12 +3,15 @@ from collections.abc import Generator
 from ee.onyx.db.external_perm import ExternalUserGroup
 from ee.onyx.external_permissions.confluence.constants import ALL_CONF_EMAILS_GROUP_NAME
 from onyx.background.error_logging import emit_background_error
+from onyx.configs.app_configs import CONFLUENCE_USE_ONYX_USERS_FOR_GROUP_SYNC
 from onyx.connectors.confluence.onyx_confluence import (
     get_user_email_from_username__server,
 )
 from onyx.connectors.confluence.onyx_confluence import OnyxConfluence
 from onyx.connectors.credentials_provider import OnyxDBCredentialsProvider
+from onyx.db.engine.sql_engine import get_session_with_current_tenant
 from onyx.db.models import ConnectorCredentialPair
+from onyx.db.users import get_all_users
 from onyx.utils.logger import setup_logger
 
 logger = setup_logger()
@@ -19,7 +22,7 @@ def _build_group_member_email_map(
 ) -> dict[str, set[str]]:
     group_member_emails: dict[str, set[str]] = {}
     for user in confluence_client.paginated_cql_user_retrieval():
-        logger.debug(f"Processing groups for user: {user}")
+        logger.info(f"Processing groups for user: {user}")
 
         email = user.email
         if not email:
@@ -31,6 +34,8 @@ def _build_group_member_email_map(
                     confluence_client=confluence_client,
                     user_name=user_name,
                 )
+            else:
+                logger.error(f"user result missing username field: {user}")
 
         if not email:
             # If we still don't have an email, skip this user
@@ -64,6 +69,92 @@ def _build_group_member_email_map(
     return group_member_emails
 
 
+def _build_group_member_email_map_from_onyx_users(
+    confluence_client: OnyxConfluence,
+) -> dict[str, set[str]]:
+    """Hacky, but it's the only way to do this as long as the
+    Confluence APIs are broken.
+
+    This is fixed in Confluence Data Center 10.1.0, so first choice
+    is to tell users to upgrade to 10.1.0.
+    https://jira.atlassian.com/browse/CONFSERVER-95999
+    """
+    with get_session_with_current_tenant() as db_session:
+        # don't include external since they are handled by the "through confluence"
+        # user fetching mechanism
+        user_emails = [
+            user.email for user in get_all_users(db_session, include_external=False)
+        ]
+
+    def _infer_username_from_email(email: str) -> str:
+        return email.split("@")[0]
+
+    group_member_emails: dict[str, set[str]] = {}
+    for email in user_emails:
+        logger.info(f"Processing groups for user with email: {email}")
+        try:
+            user_name = _infer_username_from_email(email)
+            response = confluence_client.get_user_details_by_username(user_name)
+            user_key = response.get("userKey")
+            if not user_key:
+                logger.error(f"User key not found for user with email {email}")
+                continue
+
+            all_users_groups: set[str] = set()
+            for group in confluence_client.paginated_groups_by_user_retrieval(user_key):
+                # group name uniqueness is enforced by Confluence, so we can use it as a group ID
+                group_id = group["name"]
+                group_member_emails.setdefault(group_id, set()).add(email)
+                all_users_groups.add(group_id)
+
+            if not all_users_groups:
+                msg = f"No groups found for user with email: {email}"
+                logger.error(msg)
+            else:
+                logger.info(
+                    f"Found groups {all_users_groups} for user with email {email}"
+                )
+        except Exception:
+            logger.exception(f"Error getting user details for user with email {email}")
+
+    return group_member_emails
+
+
+def _build_final_group_to_member_email_map(
+    confluence_client: OnyxConfluence,
+    cc_pair_id: int,
+    # if set, will infer confluence usernames from onyx users in addition to using the
+    # confluence users API. This is a hacky workaround for the fact that the Confluence
+    # users API is broken before Confluence Data Center 10.1.0.
+    use_onyx_users: bool = CONFLUENCE_USE_ONYX_USERS_FOR_GROUP_SYNC,
+) -> dict[str, set[str]]:
+    group_to_member_email_map = _build_group_member_email_map(
+        confluence_client=confluence_client,
+        cc_pair_id=cc_pair_id,
+    )
+    group_to_member_email_map_from_onyx_users = (
+        (
+            _build_group_member_email_map_from_onyx_users(
+                confluence_client=confluence_client,
+            )
+        )
+        if use_onyx_users
+        else {}
+    )
+
+    all_group_ids = set(group_to_member_email_map.keys()) | set(
+        group_to_member_email_map_from_onyx_users.keys()
+    )
+    final_group_to_member_email_map = {}
+    for group_id in all_group_ids:
+        group_member_emails = group_to_member_email_map.get(
+            group_id, set()
+        ) | group_to_member_email_map_from_onyx_users.get(group_id, set())
+        final_group_to_member_email_map[group_id] = group_member_emails
+
+    return final_group_to_member_email_map
+
+
 def confluence_group_sync(
     tenant_id: str,
     cc_pair: ConnectorCredentialPair,
@@ -87,13 +178,12 @@ def confluence_group_sync(
     confluence_client._probe_connection(**probe_kwargs)
     confluence_client._initialize_connection(**final_kwargs)
 
-    group_member_email_map = _build_group_member_email_map(
-        confluence_client=confluence_client,
-        cc_pair_id=cc_pair.id,
+    group_to_member_email_map = _build_final_group_to_member_email_map(
+        confluence_client, cc_pair.id
     )
 
     all_found_emails = set()
-    for group_id, group_member_emails in group_member_email_map.items():
+    for group_id, group_member_emails in group_to_member_email_map.items():
         yield (
             ExternalUserGroup(
                 id=group_id,

backend/ee/onyx/external_permissions/jira/group_sync.py

@@ -0,0 +1,136 @@
+from collections.abc import Generator
+
+from jira import JIRA
+
+from ee.onyx.db.external_perm import ExternalUserGroup
+from onyx.connectors.jira.utils import build_jira_client
+from onyx.db.models import ConnectorCredentialPair
+from onyx.utils.logger import setup_logger
+
+logger = setup_logger()
+
+
+def _get_jira_group_members_email(
+    jira_client: JIRA,
+    group_name: str,
+) -> list[str]:
+    """Get all member emails for a Jira group.
+
+    Filters out app accounts (bots, integrations) and only returns real user emails.
+    """
+    emails: list[str] = []
+
+    try:
+        # group_members returns an OrderedDict of account_id -> member_info
+        members = jira_client.group_members(group=group_name)
+
+        if not members:
+            logger.warning(f"No members found for group {group_name}")
+            return emails
+
+        for account_id, member_info in members.items():
+            # member_info is a dict with keys like 'fullname', 'email', 'active'
+            email = member_info.get("email")
+
+            # Skip "hidden" emails - these are typically app accounts
+            if email and email != "hidden":
+                emails.append(email)
+            else:
+                # For cloud, we might need to fetch user details separately
+                try:
+                    user = jira_client.user(id=account_id)
+
+                    # Skip app accounts (bots, integrations, etc.)
+                    if hasattr(user, "accountType") and user.accountType == "app":
+                        logger.info(
+                            f"Skipping app account {account_id} for group {group_name}"
+                        )
+                        continue
+
+                    if hasattr(user, "emailAddress") and user.emailAddress:
+                        emails.append(user.emailAddress)
+                    else:
+                        logger.warning(f"User {account_id} has no email address")
+                except Exception as e:
+                    logger.warning(
+                        f"Could not fetch email for user {account_id} in group {group_name}: {e}"
+                    )
+
+    except Exception as e:
+        logger.error(f"Error fetching members for group {group_name}: {e}")
+
+    return emails
+
+
+def _build_group_member_email_map(
+    jira_client: JIRA,
+) -> dict[str, set[str]]:
+    """Build a map of group names to member emails."""
+    group_member_emails: dict[str, set[str]] = {}
+
+    try:
+        # Get all groups from Jira - returns a list of group name strings
+        group_names = jira_client.groups()
+
+        if not group_names:
+            logger.warning("No groups found in Jira")
+            return group_member_emails
+
+        logger.info(f"Found {len(group_names)} groups in Jira")
+
+        for group_name in group_names:
+            if not group_name:
+                continue
+
+            member_emails = _get_jira_group_members_email(
+                jira_client=jira_client,
+                group_name=group_name,
+            )
+
+            if member_emails:
+                group_member_emails[group_name] = set(member_emails)
+                logger.debug(
+                    f"Found {len(member_emails)} members for group {group_name}"
+                )
+            else:
+                logger.debug(f"No members found for group {group_name}")
+
+    except Exception as e:
+        logger.error(f"Error building group member email map: {e}")
+
+    return group_member_emails
+
+
+def jira_group_sync(
+    tenant_id: str,
+    cc_pair: ConnectorCredentialPair,
+) -> Generator[ExternalUserGroup, None, None]:
+    """
+    Sync Jira groups and their members.
+
+    This function fetches all groups from Jira and yields ExternalUserGroup
+    objects containing the group ID and member emails.
+    """
+    jira_base_url = cc_pair.connector.connector_specific_config.get("jira_base_url", "")
+    scoped_token = cc_pair.connector.connector_specific_config.get(
+        "scoped_token", False
+    )
+
+    if not jira_base_url:
+        raise ValueError("No jira_base_url found in connector config")
+
+    jira_client = build_jira_client(
+        credentials=cc_pair.credential.credential_json,
+        jira_base=jira_base_url,
+        scoped_token=scoped_token,
+    )
+
+    group_member_email_map = _build_group_member_email_map(jira_client=jira_client)
+    if not group_member_email_map:
+        raise ValueError(f"No groups with members found for cc_pair_id={cc_pair.id}")
+
+    for group_id, group_member_emails in group_member_email_map.items():
+        yield ExternalUserGroup(
+            id=group_id,
+            user_emails=list(group_member_emails),
+        )

backend/ee/onyx/external_permissions/jira/page_access.py

@@ -16,6 +16,10 @@ HolderMap = dict[str, list[Holder]]
 logger = setup_logger()
 
 
+def _get_role_id(holder: Holder) -> str | None:
+    return holder.get("value") or holder.get("parameter")
+
+
 def _build_holder_map(permissions: list[dict]) -> dict[str, list[Holder]]:
     """
     A "Holder" in JIRA is a person / entity who "holds" the corresponding permission.
@@ -110,80 +114,137 @@ def _get_user_emails(user_holders: list[Holder]) -> list[str]:
     return emails
 
 
-def _get_user_emails_from_project_roles(
+def _get_user_emails_and_groups_from_project_roles(
     jira_client: JIRA,
     jira_project: str,
     project_role_holders: list[Holder],
-) -> list[str]:
-    # NOTE (@raunakab) a `parallel_yield` may be helpful here...?
+) -> tuple[list[str], list[str]]:
+    """
+    Get user emails and group names from project roles.
+    Returns a tuple of (emails, group_names).
+    """
+    # Get role IDs - Cloud uses "value", Data Center uses "parameter"
+    role_ids = []
+    for holder in project_role_holders:
+        role_id = _get_role_id(holder)
+        if role_id:
+            role_ids.append(role_id)
+        else:
+            logger.warning(f"No value or parameter in projectRole holder: {holder}")
+
     roles = [
-        jira_client.project_role(project=jira_project, id=project_role_holder["value"])
-        for project_role_holder in project_role_holders
-        if "value" in project_role_holder
+        jira_client.project_role(project=jira_project, id=role_id)
+        for role_id in role_ids
     ]
 
     emails = []
+    groups = []
 
     for role in roles:
         if not hasattr(role, "actors"):
+            logger.warning(f"Project role {role} has no actors attribute")
             continue
 
         for actor in role.actors:
-            if not hasattr(actor, "actorUser") or not hasattr(
-                actor.actorUser, "accountId"
-            ):
+            # Handle group actors
+            if hasattr(actor, "actorGroup"):
+                group_name = getattr(actor.actorGroup, "name", None) or getattr(
+                    actor.actorGroup, "displayName", None
+                )
+                if group_name:
+                    groups.append(group_name)
                 continue
 
-            user = jira_client.user(id=actor.actorUser.accountId)
-            if not hasattr(user, "accountType") or user.accountType != "atlassian":
+            # Handle user actors
+            if hasattr(actor, "actorUser"):
+                account_id = getattr(actor.actorUser, "accountId", None)
+                if not account_id:
+                    logger.error(f"No accountId in actorUser: {actor.actorUser}")
+                    continue
+
+                user = jira_client.user(id=account_id)
+                if not hasattr(user, "accountType") or user.accountType != "atlassian":
+                    logger.info(
+                        f"Skipping user {account_id} because it is not an atlassian user"
+                    )
+                    continue
+
+                if not hasattr(user, "emailAddress"):
+                    msg = f"User's email address was not able to be retrieved;  {actor.actorUser.accountId=}"
+                    if hasattr(user, "displayName"):
+                        msg += f" {actor.displayName=}"
+                    logger.warning(msg)
+                    continue
+
+                emails.append(user.emailAddress)
                 continue
 
-            if not hasattr(user, "emailAddress"):
-                msg = f"User's email address was not able to be retrieved;  {actor.actorUser.accountId=}"
-                if hasattr(user, "displayName"):
-                    msg += f" {actor.displayName=}"
-                logger.warn(msg)
-                continue
+            logger.debug(f"Skipping actor type: {actor}")
 
-            emails.append(user.emailAddress)
-
-    return emails
+    return emails, groups
 
 
 def _build_external_access_from_holder_map(
     jira_client: JIRA, jira_project: str, holder_map: HolderMap
 ) -> ExternalAccess:
     """
-    # Note:
-        If the `holder_map` contains an instance of "anyone", then this is a public JIRA project.
-        Otherwise, we fetch the "projectRole"s (i.e., the user-groups in JIRA speak), and the user emails.
-    """
+    Build ExternalAccess from the holder map.
 
+    Holder types handled:
+        - "anyone": Public project, anyone can access
+        - "applicationRole": All users with a Jira license can access (treated as public)
+        - "user": Specific users with access
+        - "projectRole": Project roles containing users and/or groups
+        - "group": Groups directly assigned in the permission scheme
+    """
+    # Public access - anyone can view
     if "anyone" in holder_map:
         return ExternalAccess(
             external_user_emails=set(), external_user_group_ids=set(), is_public=True
         )
 
+    # applicationRole means all users with a Jira license can access - treat as public
+    if "applicationRole" in holder_map:
+        return ExternalAccess(
+            external_user_emails=set(), external_user_group_ids=set(), is_public=True
+        )
+
+    # Get emails from explicit user holders
     user_emails = (
         _get_user_emails(user_holders=holder_map["user"])
         if "user" in holder_map
         else []
     )
-    project_role_user_emails = (
-        _get_user_emails_from_project_roles(
-            jira_client=jira_client,
-            jira_project=jira_project,
-            project_role_holders=holder_map["projectRole"],
+
+    # Get emails and groups from project roles
+    project_role_user_emails: list[str] = []
+    project_role_groups: list[str] = []
+    if "projectRole" in holder_map:
+        project_role_user_emails, project_role_groups = (
+            _get_user_emails_and_groups_from_project_roles(
+                jira_client=jira_client,
+                jira_project=jira_project,
+                project_role_holders=holder_map["projectRole"],
+            )
         )
-        if "projectRole" in holder_map
-        else []
-    )
+
+    # Get groups directly assigned in permission scheme (common in Data Center)
+    # Format: {'type': 'group', 'parameter': 'group-name', 'expand': 'group'}
+    direct_groups: list[str] = []
+    if "group" in holder_map:
+        for group_holder in holder_map["group"]:
+            group_name = _get_role_id(group_holder)
+            if group_name:
+                direct_groups.append(group_name)
+            else:
+                logger.error(f"No parameter/value in group holder: {group_holder}")
 
     external_user_emails = set(user_emails + project_role_user_emails)
+    external_user_group_ids = set(project_role_groups + direct_groups)
 
     return ExternalAccess(
         external_user_emails=external_user_emails,
-        external_user_group_ids=set(),
+        external_user_group_ids=external_user_group_ids,
         is_public=False,
     )
 
@@ -197,9 +258,11 @@ def get_project_permissions(
     )
 
     if not hasattr(project_permissions, "permissions"):
+        logger.error(f"Project {jira_project} has no permissions attribute")
         return None
 
     if not isinstance(project_permissions.permissions, list):
+        logger.error(f"Project {jira_project} permissions is not a list")
         return None
 
     holder_map = _build_holder_map(permissions=project_permissions.permissions)

backend/ee/onyx/external_permissions/sharepoint/permission_utils.py

@@ -15,6 +15,7 @@ from ee.onyx.db.external_perm import ExternalUserGroup
 from onyx.access.models import ExternalAccess
 from onyx.access.utils import build_ext_group_name_for_onyx
 from onyx.configs.constants import DocumentSource
+from onyx.connectors.sharepoint.connector import SHARED_DOCUMENTS_MAP_REVERSE
 from onyx.connectors.sharepoint.connector import sleep_and_retry
 from onyx.utils.logger import setup_logger
 
@@ -511,8 +512,8 @@ def get_external_access_from_sharepoint(
                 f"Failed to get SharePoint list item ID for item {drive_item.id}"
             )
 
-        if drive_name == "Shared Documents":
-            drive_name = "Documents"
+        if drive_name in SHARED_DOCUMENTS_MAP_REVERSE:
+            drive_name = SHARED_DOCUMENTS_MAP_REVERSE[drive_name]
 
         item = client_context.web.lists.get_by_title(drive_name).items.get_by_id(
             item_id

backend/ee/onyx/external_permissions/sync_params.py

@@ -11,6 +11,7 @@ from ee.onyx.configs.app_configs import GITHUB_PERMISSION_DOC_SYNC_FREQUENCY
 from ee.onyx.configs.app_configs import GITHUB_PERMISSION_GROUP_SYNC_FREQUENCY
 from ee.onyx.configs.app_configs import GOOGLE_DRIVE_PERMISSION_GROUP_SYNC_FREQUENCY
 from ee.onyx.configs.app_configs import JIRA_PERMISSION_DOC_SYNC_FREQUENCY
+from ee.onyx.configs.app_configs import JIRA_PERMISSION_GROUP_SYNC_FREQUENCY
 from ee.onyx.configs.app_configs import SHAREPOINT_PERMISSION_DOC_SYNC_FREQUENCY
 from ee.onyx.configs.app_configs import SHAREPOINT_PERMISSION_GROUP_SYNC_FREQUENCY
 from ee.onyx.configs.app_configs import SLACK_PERMISSION_DOC_SYNC_FREQUENCY
@@ -23,6 +24,7 @@ from ee.onyx.external_permissions.gmail.doc_sync import gmail_doc_sync
 from ee.onyx.external_permissions.google_drive.doc_sync import gdrive_doc_sync
 from ee.onyx.external_permissions.google_drive.group_sync import gdrive_group_sync
 from ee.onyx.external_permissions.jira.doc_sync import jira_doc_sync
+from ee.onyx.external_permissions.jira.group_sync import jira_group_sync
 from ee.onyx.external_permissions.perm_sync_types import CensoringFuncType
 from ee.onyx.external_permissions.perm_sync_types import DocSyncFuncType
 from ee.onyx.external_permissions.perm_sync_types import FetchAllDocumentsFunction
@@ -110,6 +112,11 @@ _SOURCE_TO_SYNC_CONFIG: dict[DocumentSource, SyncConfig] = {
             doc_sync_func=jira_doc_sync,
             initial_index_should_sync=True,
         ),
+        group_sync_config=GroupSyncConfig(
+            group_sync_frequency=JIRA_PERMISSION_GROUP_SYNC_FREQUENCY,
+            group_sync_func=jira_group_sync,
+            group_sync_is_cc_pair_agnostic=True,
+        ),
     ),
     # Groups are not needed for Slack.
     # All channel access is done at the individual user level.

backend/onyx/background/celery/tasks/docprocessing/tasks.py

@@ -156,8 +156,6 @@ def validate_active_indexing_attempts(
     """
     logger.info("Validating active indexing attempts")
 
-    heartbeat_timeout_seconds = HEARTBEAT_TIMEOUT_SECONDS
-
     with get_session_with_current_tenant() as db_session:
 
         # Find all active indexing attempts
@@ -175,6 +173,9 @@ def validate_active_indexing_attempts(
         for attempt in active_attempts:
             lock_beat.reacquire()
 
+            # Initialize timeout for each attempt to prevent state pollution
+            heartbeat_timeout_seconds = HEARTBEAT_TIMEOUT_SECONDS
+
             # Double-check the attempt still exists and has the same status
             fresh_attempt = get_index_attempt(db_session, attempt.id)
             if not fresh_attempt or fresh_attempt.status.is_terminal():
@@ -815,10 +816,14 @@ def check_for_indexing(self: Task, *, tenant_id: str) -> int | None:
             secondary_cc_pair_ids: list[int] = []
             secondary_search_settings = get_secondary_search_settings(db_session)
             if secondary_search_settings:
-                # Include paused CC pairs during embedding swap
+                # For ACTIVE_ONLY, we skip paused connectors
+                include_paused = (
+                    secondary_search_settings.switchover_type
+                    != SwitchoverType.ACTIVE_ONLY
+                )
                 standard_cc_pair_ids = (
                     fetch_indexable_standard_connector_credential_pair_ids(
-                        db_session, active_cc_pairs_only=False
+                        db_session, active_cc_pairs_only=not include_paused
                     )
                 )
                 user_file_cc_pair_ids = (

backend/onyx/configs/app_configs.py

@@ -528,6 +528,10 @@ CONFLUENCE_TIMEZONE_OFFSET = float(
     os.environ.get("CONFLUENCE_TIMEZONE_OFFSET", get_current_tz_offset())
 )
 
+CONFLUENCE_USE_ONYX_USERS_FOR_GROUP_SYNC = (
+    os.environ.get("CONFLUENCE_USE_ONYX_USERS_FOR_GROUP_SYNC", "").lower() == "true"
+)
+
 GOOGLE_DRIVE_CONNECTOR_SIZE_THRESHOLD = int(
     os.environ.get("GOOGLE_DRIVE_CONNECTOR_SIZE_THRESHOLD", 10 * 1024 * 1024)
 )

backend/onyx/connectors/cross_connector_utils/miscellaneous_utils.py

@@ -134,7 +134,7 @@ def process_onyx_metadata(
     metadata: dict[str, Any],
 ) -> tuple[OnyxMetadata, dict[str, Any]]:
     """
-    Users may set Onyx metadata and custom tags in text files. https://docs.onyx.app/admin/connectors/official/file
+    Users may set Onyx metadata and custom tags in text files. https://docs.onyx.app/admins/connectors/official/file
     Any unrecognized fields are treated as custom tags.
     """
     p_owner_names = metadata.get("primary_owners")

backend/onyx/connectors/file/connector.py

@@ -155,7 +155,7 @@ def _process_file(
         content_type=file_type,
     )
 
-    # Each file may have file-specific ONYX_METADATA https://docs.onyx.app/admin/connectors/official/file
+    # Each file may have file-specific ONYX_METADATA https://docs.onyx.app/admins/connectors/official/file
     # If so, we should add it to any metadata processed so far
     if extraction_result.metadata:
         logger.debug(

backend/onyx/connectors/google_utils/shared_constants.py

@@ -44,7 +44,7 @@ USER_FIELDS = "nextPageToken, users(primaryEmail)"
 MISSING_SCOPES_ERROR_STR = "client not authorized for any of the scopes requested"
 
 # Documentation and error messages
-SCOPE_DOC_URL = "https://docs.onyx.app/admin/connectors/official/google_drive/overview"
+SCOPE_DOC_URL = "https://docs.onyx.app/admins/connectors/official/google_drive/overview"
 ONYX_SCOPE_INSTRUCTIONS = (
     "You have upgraded Onyx without updating the Google Auth scopes. "
     f"Please refer to the documentation to learn how to update the scopes: {SCOPE_DOC_URL}"

backend/onyx/connectors/sharepoint/connector.py

@@ -68,6 +68,13 @@ logger = setup_logger()
 SLIM_BATCH_SIZE = 1000
 
 
+SHARED_DOCUMENTS_MAP = {
+    "Documents": "Shared Documents",
+    "Dokumente": "Freigegebene Dokumente",
+    "Documentos": "Documentos compartidos",
+}
+SHARED_DOCUMENTS_MAP_REVERSE = {v: k for k, v in SHARED_DOCUMENTS_MAP.items()}
+
 ASPX_EXTENSION = ".aspx"
 
 
@@ -778,7 +785,10 @@ class SharepointConnector(
                 drive
                 for drive in drives
                 if (drive.name and drive.name.lower() == drive_name.lower())
-                or (drive.name == "Documents" and drive_name == "Shared Documents")
+                or (
+                    drive.name in SHARED_DOCUMENTS_MAP
+                    and SHARED_DOCUMENTS_MAP[drive.name] == drive_name
+                )
             ]
             drive = drives[0] if len(drives) > 0 else None
             if drive is None:
@@ -885,10 +895,12 @@ class SharepointConnector(
                     for drive in drives
                     if drive.name == site_descriptor.drive_name
                     or (
-                        drive.name == "Documents"
-                        and site_descriptor.drive_name == "Shared Documents"
+                        drive.name in SHARED_DOCUMENTS_MAP
+                        and SHARED_DOCUMENTS_MAP[drive.name]
+                        == site_descriptor.drive_name
                     )
-                ]
+                ]  # NOTE: right now we only support english, german and spanish drive names
+                # add to SHARED_DOCUMENTS_MAP if you want to support more languages
                 if not drives:
                     logger.warning(f"Drive '{site_descriptor.drive_name}' not found")
                     return []
@@ -914,9 +926,11 @@ class SharepointConnector(
                     )
 
                     # Use "Shared Documents" as the library name for the default "Documents" drive
+                    # NOTE: right now we only support english, german and spanish drive names
+                    # add to SHARED_DOCUMENTS_MAP if you want to support more languages
                     drive_name = (
-                        "Shared Documents"
-                        if drive.name == "Documents"
+                        SHARED_DOCUMENTS_MAP[drive.name]
+                        if drive.name in SHARED_DOCUMENTS_MAP
                         else cast(str, drive.name)
                     )
 
@@ -1455,10 +1469,8 @@ class SharepointConnector(
                 # Clear current drive and continue to next
                 checkpoint.current_drive_name = None
                 return checkpoint
-            current_drive_name = (
-                "Shared Documents"
-                if current_drive_name == "Documents"
-                else current_drive_name
+            current_drive_name = SHARED_DOCUMENTS_MAP.get(
+                current_drive_name, current_drive_name
             )
             for driveitem in driveitems:
                 driveitem_extension = get_file_ext(driveitem.name)

backend/onyx/db/connector_credential_pair.py

@@ -17,8 +17,10 @@ from sqlalchemy.orm import joinedload
 from sqlalchemy.orm import selectinload
 from sqlalchemy.orm import Session
 
+from onyx.configs.app_configs import AUTH_TYPE
 from onyx.configs.app_configs import DISABLE_AUTH
 from onyx.configs.app_configs import USER_FILE_INDEXING_LIMIT
+from onyx.configs.constants import AuthType
 from onyx.configs.constants import DocumentSource
 from onyx.db.connector import fetch_connector_by_id
 from onyx.db.credentials import fetch_credential_by_id
@@ -445,10 +447,12 @@ def set_cc_pair_repeated_error_state(
     values: dict = {"in_repeated_error_state": in_repeated_error_state}
 
     # When entering repeated error state, also pause the connector
-    # to prevent continued indexing retry attempts.
+    # to prevent continued indexing retry attempts burning through embedding credits.
     # However, don't pause if there's an active manual indexing trigger,
     # which indicates the user wants to retry immediately.
-    if in_repeated_error_state:
+    # NOTE: only for Cloud, since most self-hosted users use self-hosted embedding
+    # models. Also, they are more prone to repeated failures -> eventual success.
+    if in_repeated_error_state and AUTH_TYPE == AuthType.CLOUD:
         cc_pair = get_connector_credential_pair_from_id(
             db_session=db_session,
             cc_pair_id=cc_pair_id,

backend/onyx/db/document.py

@@ -686,11 +686,6 @@ def delete_documents_complete__no_commit(
         document_ids=document_ids,
     )
 
-    delete_chunk_stats_by_connector_credential_pair__no_commit(
-        db_session=db_session,
-        document_ids=document_ids,
-    )
-
     delete_documents_by_connector_credential_pair__no_commit(db_session, document_ids)
     delete_document_feedback_for_documents__no_commit(
         document_ids=document_ids, db_session=db_session

backend/onyx/db/slack_channel_config.py

@@ -2,6 +2,7 @@ from collections.abc import Sequence
 from typing import Any
 
 from sqlalchemy import select
+from sqlalchemy.orm import joinedload
 from sqlalchemy.orm import Session
 
 from onyx.configs.chat_configs import MAX_CHUNKS_FED_TO_CHAT
@@ -269,7 +270,9 @@ def fetch_slack_channel_config_for_channel_or_default(
     # attempt to find channel-specific config first
     if channel_name is not None:
         sc_config = db_session.scalar(
-            select(SlackChannelConfig).where(
+            select(SlackChannelConfig)
+            .options(joinedload(SlackChannelConfig.persona))
+            .where(
                 SlackChannelConfig.slack_bot_id == slack_bot_id,
                 SlackChannelConfig.channel_config["channel_name"].astext
                 == channel_name,
@@ -283,7 +286,9 @@ def fetch_slack_channel_config_for_channel_or_default(
 
     # if none found, see if there is a default
     default_sc = db_session.scalar(
-        select(SlackChannelConfig).where(
+        select(SlackChannelConfig)
+        .options(joinedload(SlackChannelConfig.persona))
+        .where(
             SlackChannelConfig.slack_bot_id == slack_bot_id,
             SlackChannelConfig.is_default == True,  # noqa: E712
         )

backend/onyx/db/tag.py

@@ -4,6 +4,7 @@ from sqlalchemy import and_
 from sqlalchemy import delete
 from sqlalchemy import or_
 from sqlalchemy import select
+from sqlalchemy.dialects.postgresql import insert as pg_insert
 from sqlalchemy.orm import Session
 
 from onyx.configs.constants import DocumentSource
@@ -43,17 +44,26 @@ def create_or_add_document_tag(
     if not document:
         raise ValueError("Invalid Document, cannot attach Tags")
 
+    # Use upsert to avoid race condition when multiple workers try to create the same tag
+    insert_stmt = pg_insert(Tag).values(
+        tag_key=tag_key,
+        tag_value=tag_value,
+        source=source,
+        is_list=False,
+    )
+    insert_stmt = insert_stmt.on_conflict_do_nothing(
+        index_elements=["tag_key", "tag_value", "source", "is_list"]
+    )
+    db_session.execute(insert_stmt)
+
+    # Now fetch the tag (either just inserted or already existed)
     tag_stmt = select(Tag).where(
         Tag.tag_key == tag_key,
         Tag.tag_value == tag_value,
         Tag.source == source,
         Tag.is_list.is_(False),
     )
-    tag = db_session.execute(tag_stmt).scalar_one_or_none()
-
-    if not tag:
-        tag = Tag(tag_key=tag_key, tag_value=tag_value, source=source, is_list=False)
-        db_session.add(tag)
+    tag = db_session.execute(tag_stmt).scalar_one()
 
     if tag not in document.tags:
         document.tags.append(tag)
@@ -79,31 +89,27 @@ def create_or_add_document_tag_list(
     if not document:
         raise ValueError("Invalid Document, cannot attach Tags")
 
-    existing_tags_stmt = select(Tag).where(
+    # Use upsert to avoid race condition when multiple workers try to create the same tags
+    for tag_value in valid_tag_values:
+        insert_stmt = pg_insert(Tag).values(
+            tag_key=tag_key,
+            tag_value=tag_value,
+            source=source,
+            is_list=True,
+        )
+        insert_stmt = insert_stmt.on_conflict_do_nothing(
+            index_elements=["tag_key", "tag_value", "source", "is_list"]
+        )
+        db_session.execute(insert_stmt)
+
+    # Now fetch all tags (either just inserted or already existed)
+    all_tags_stmt = select(Tag).where(
         Tag.tag_key == tag_key,
         Tag.tag_value.in_(valid_tag_values),
         Tag.source == source,
         Tag.is_list.is_(True),
     )
-    existing_tags = list(db_session.execute(existing_tags_stmt).scalars().all())
-    existing_tag_values = {tag.tag_value for tag in existing_tags}
-
-    new_tags = []
-    for tag_value in valid_tag_values:
-        if tag_value not in existing_tag_values:
-            new_tag = Tag(
-                tag_key=tag_key, tag_value=tag_value, source=source, is_list=True
-            )
-            db_session.add(new_tag)
-            new_tags.append(new_tag)
-            existing_tag_values.add(tag_value)
-
-    if new_tags:
-        logger.debug(
-            f"Created new tags: {', '.join([f'{tag.tag_key}:{tag.tag_value}' for tag in new_tags])}"
-        )
-
-    all_tags = existing_tags + new_tags
+    all_tags = list(db_session.execute(all_tags_stmt).scalars().all())
 
     for tag in all_tags:
         if tag not in document.tags:

backend/onyx/llm/llm_provider_options.py

@@ -416,7 +416,7 @@ def fetch_available_well_known_llms() -> list[WellKnownLLMProviderDescriptor]:
                     name=VERTEX_LOCATION_KWARG,
                     display_name="Location",
                     description="The location of the Vertex AI model. Please refer to the "
-                    "[Vertex AI configuration docs](https://docs.onyx.app/admin/ai_models/google_ai) for all possible values.",
+                    "[Vertex AI configuration docs](https://docs.onyx.app/admins/ai_models/google_ai) for all possible values.",
                     is_required=False,
                     is_secret=False,
                     key_type=CustomConfigKeyType.TEXT_INPUT,

backend/onyx/llm/utils.py

@@ -910,22 +910,32 @@ def is_true_openai_model(model_provider: str, model_name: str) -> bool:
     """
 
     # NOTE: not using the OPENAI_PROVIDER_NAME constant here due to circular import issues
-    if model_provider != "openai":
+    if model_provider != "openai" and model_provider != "litellm_proxy":
         return False
 
+    model_map = get_model_map()
+
+    def _check_if_model_name_is_openai_provider(model_name: str) -> bool:
+        return (
+            model_name in model_map
+            and model_map[model_name].get("litellm_provider") == "openai"
+        )
+
     try:
-        model_map = get_model_map()
+
         # Check if any model exists in litellm's registry with openai prefix
         # If it's registered as "openai/model-name", it's a real OpenAI model
         if f"openai/{model_name}" in model_map:
             return True
 
-        if (
-            model_name in model_map
-            and model_map[model_name].get("litellm_provider") == "openai"
-        ):
+        if _check_if_model_name_is_openai_provider(model_name):
             return True
 
+        if model_name.startswith("azure/"):
+            model_name_with_azure_removed = "/".join(model_name.split("/")[1:])
+            if _check_if_model_name_is_openai_provider(model_name_with_azure_removed):
+                return True
+
         return False
 
     except Exception:

backend/onyx/natural_language_processing/search_nlp_models.py

@@ -1066,6 +1066,17 @@ class InformationContentClassificationModel:
         self,
         queries: list[str],
     ) -> list[ContentClassificationPrediction]:
+        if os.environ.get("DISABLE_MODEL_SERVER", "").lower() == "true":
+            logger.info(
+                "DISABLE_MODEL_SERVER is set, returning default classifications"
+            )
+            return [
+                ContentClassificationPrediction(
+                    predicted_label=1, content_boost_factor=1.0
+                )
+                for _ in queries
+            ]
+
         response = requests.post(self.content_server_endpoint, json=queries)
         response.raise_for_status()
 
@@ -1092,6 +1103,14 @@ class ConnectorClassificationModel:
         query: str,
         available_connectors: list[str],
     ) -> list[str]:
+        # Check if model server is disabled
+        if os.environ.get("DISABLE_MODEL_SERVER", "").lower() == "true":
+            logger.info(
+                "DISABLE_MODEL_SERVER is set, returning all available connectors"
+            )
+            # Return all available connectors when model server is disabled
+            return available_connectors
+
         connector_classification_request = ConnectorClassificationRequest(
             available_connectors=available_connectors,
             query=query,

backend/onyx/onyxbot/slack/handlers/handle_regular_answer.py

@@ -117,12 +117,14 @@ def handle_regular_answer(
     # This way slack flow always has a persona
     persona = slack_channel_config.persona
     if not persona:
+        logger.warning("No persona found for channel config, using default persona")
         with get_session_with_current_tenant() as db_session:
             persona = get_persona_by_id(DEFAULT_PERSONA_ID, user, db_session)
             document_set_names = [
                 document_set.name for document_set in persona.document_sets
             ]
     else:
+        logger.info(f"Using persona {persona.name} for channel config")
         document_set_names = [
             document_set.name for document_set in persona.document_sets
         ]

backend/onyx/onyxbot/slack/listener.py

@@ -1104,7 +1104,7 @@ def _get_socket_client(
     slack_bot_tokens: SlackBotTokens, tenant_id: str, slack_bot_id: int
 ) -> TenantSocketModeClient:
     # For more info on how to set this up, checkout the docs:
-    # https://docs.onyx.app/admin/getting_started/slack_bot_setup
+    # https://docs.onyx.app/admins/getting_started/slack_bot_setup
 
     # use the retry handlers built into the slack sdk
     connection_error_retry_handler = ConnectionErrorRetryHandler()

backend/onyx/server/manage/llm/api.py

@@ -351,12 +351,19 @@ def get_vision_capable_providers(
 
         # Only include providers with at least one vision-capable model
         if vision_models:
-            provider_dict = LLMProviderView.from_model(provider).model_dump()
-            provider_dict["vision_models"] = vision_models
+            provider_view = LLMProviderView.from_model(provider)
+            _mask_provider_api_key(provider_view)
+
+            vision_providers.append(
+                VisionProviderResponse(
+                    **provider_view.model_dump(),
+                    vision_models=vision_models,
+                )
+            )
+
             logger.info(
                 f"Vision provider: {provider.provider} with models: {vision_models}"
             )
-            vision_providers.append(VisionProviderResponse(**provider_dict))
 
     logger.info(f"Found {len(vision_providers)} vision-capable providers")
     return vision_providers

backend/onyx/utils/gpu_utils.py

@@ -1,3 +1,4 @@
+import os
 from functools import lru_cache
 
 import requests
@@ -13,16 +14,14 @@ logger = setup_logger()
 
 
 def _get_gpu_status_from_model_server(indexing: bool) -> bool:
+    if os.environ.get("DISABLE_MODEL_SERVER", "").lower() == "true":
+        logger.info("DISABLE_MODEL_SERVER is set, assuming no GPU available")
+        return False
     if indexing:
         model_server_url = f"{INDEXING_MODEL_SERVER_HOST}:{INDEXING_MODEL_SERVER_PORT}"
     else:
         model_server_url = f"{MODEL_SERVER_HOST}:{MODEL_SERVER_PORT}"
 
-    # If model server is disabled, return False (no GPU available)
-    if model_server_url in ["disabled", "disabled:9000"]:
-        logger.info("Model server is disabled, assuming no GPU available")
-        return False
-
     if "http" not in model_server_url:
         model_server_url = f"http://{model_server_url}"
 

backend/requirements/default.txt

@@ -85,7 +85,7 @@ botocore==1.39.11
     #   s3transfer
 botocore-stubs==1.40.74
     # via boto3-stubs
-braintrust==0.2.6
+braintrust==0.3.9
     # via braintrust-langchain
 braintrust-langchain==0.0.4
 bytecode==0.17.0
@@ -1150,6 +1150,7 @@ werkzeug==3.1.1
 wrapt==1.17.3
     # via
     #   aiobotocore
+    #   braintrust
     #   ddtrace
     #   deprecated
     #   langfuse

backend/shared_configs/configs.py

@@ -10,14 +10,24 @@ SLACK_CHANNEL_ID = "channel_id"
 # Default to True (skip warmup) if not set, otherwise respect the value
 SKIP_WARM_UP = os.environ.get("SKIP_WARM_UP", "true").lower() == "true"
 
-MODEL_SERVER_HOST = os.environ.get("MODEL_SERVER_HOST") or "localhost"
-MODEL_SERVER_ALLOWED_HOST = os.environ.get("MODEL_SERVER_HOST") or "0.0.0.0"
+# Check if model server is disabled
+DISABLE_MODEL_SERVER = os.environ.get("DISABLE_MODEL_SERVER", "").lower() == "true"
+
+# If model server is disabled, use "disabled" as host to trigger proper handling
+if DISABLE_MODEL_SERVER:
+    MODEL_SERVER_HOST = "disabled"
+    MODEL_SERVER_ALLOWED_HOST = "disabled"
+    INDEXING_MODEL_SERVER_HOST = "disabled"
+else:
+    MODEL_SERVER_HOST = os.environ.get("MODEL_SERVER_HOST") or "localhost"
+    MODEL_SERVER_ALLOWED_HOST = os.environ.get("MODEL_SERVER_HOST") or "0.0.0.0"
+    INDEXING_MODEL_SERVER_HOST = (
+        os.environ.get("INDEXING_MODEL_SERVER_HOST") or MODEL_SERVER_HOST
+    )
+
 MODEL_SERVER_PORT = int(os.environ.get("MODEL_SERVER_PORT") or "9000")
 # Model server for indexing should use a separate one to not allow indexing to introduce delay
 # for inference
-INDEXING_MODEL_SERVER_HOST = (
-    os.environ.get("INDEXING_MODEL_SERVER_HOST") or MODEL_SERVER_HOST
-)
 INDEXING_MODEL_SERVER_PORT = int(
     os.environ.get("INDEXING_MODEL_SERVER_PORT") or MODEL_SERVER_PORT
 )

backend/supervisord.conf

@@ -195,7 +195,7 @@ stopasgroup=true
 # Listens for Slack messages and responds with answers
 # for all channels that the OnyxBot has been added to.
 # If not setup, this will just fail 5 times and then stop.
-# More details on setup here: https://docs.onyx.app/admin/getting_started/slack_bot_setup
+# More details on setup here: https://docs.onyx.app/admins/getting_started/slack_bot_setup
 [program:slack_bot]
 command=python onyx/onyxbot/slack/listener.py
 stdout_logfile=/var/log/slack_bot.log

backend/tests/external_dependency_unit/connectors/jira/conftest.py

@@ -0,0 +1,29 @@
+import os
+from typing import Any
+
+import pytest
+
+
+@pytest.fixture
+def jira_connector_config() -> dict[str, Any]:
+    jira_base_url = os.environ.get("JIRA_BASE_URL", "https://danswerai.atlassian.net")
+
+    return {
+        "jira_base_url": jira_base_url,
+        "project_key": "",  # Empty to sync all projects
+        "scoped_token": False,
+    }
+
+
+@pytest.fixture
+def jira_credential_json() -> dict[str, Any]:
+    user_email = os.environ.get("JIRA_ADMIN_USER_EMAIL", "chris@onyx.app")
+    api_token = os.environ.get("JIRA_ADMIN_API_TOKEN")
+
+    assert user_email, "JIRA_ADMIN_USER_EMAIL environment variable is required"
+    assert api_token, "JIRA_ADMIN_API_TOKEN environment variable is required"
+
+    return {
+        "jira_user_email": user_email,
+        "jira_api_token": api_token,
+    }

backend/tests/external_dependency_unit/connectors/jira/test_jira_doc_sync.py

@@ -0,0 +1,228 @@
+from typing import Any
+
+from pydantic import BaseModel
+from sqlalchemy.orm import Session
+
+from ee.onyx.external_permissions.jira.doc_sync import jira_doc_sync
+from onyx.access.models import DocExternalAccess
+from onyx.configs.constants import DocumentSource
+from onyx.connectors.models import InputType
+from onyx.db.enums import AccessType
+from onyx.db.enums import ConnectorCredentialPairStatus
+from onyx.db.models import Connector
+from onyx.db.models import ConnectorCredentialPair
+from onyx.db.models import Credential
+from onyx.db.utils import DocumentRow
+from onyx.db.utils import SortOrder
+from onyx.utils.variable_functionality import global_version
+
+
+# In order to get these tests to run, use the credentials from Bitwarden.
+# Search up "ENV vars for local and Github tests", and find the Jira relevant key-value pairs.
+# Required env vars: JIRA_USER_EMAIL, JIRA_API_TOKEN
+
+
+class DocExternalAccessSet(BaseModel):
+    """A version of DocExternalAccess that uses sets for comparison."""
+
+    doc_id: str
+    external_user_emails: set[str]
+    external_user_group_ids: set[str]
+    is_public: bool
+
+    @classmethod
+    def from_doc_external_access(
+        cls, doc_external_access: DocExternalAccess
+    ) -> "DocExternalAccessSet":
+        return cls(
+            doc_id=doc_external_access.doc_id,
+            external_user_emails=doc_external_access.external_access.external_user_emails,
+            external_user_group_ids=doc_external_access.external_access.external_user_group_ids,
+            is_public=doc_external_access.external_access.is_public,
+        )
+
+
+def test_jira_doc_sync(
+    db_session: Session,
+    jira_connector_config: dict[str, Any],
+    jira_credential_json: dict[str, Any],
+) -> None:
+    """Test that Jira doc sync returns documents with correct permissions.
+
+    This test uses the AS project which has applicationRole permission,
+    meaning all documents should be marked as public.
+    """
+    # NOTE: must set EE on or else the connector will skip the perm syncing
+    global_version.set_ee()
+
+    try:
+        # Use AS project specifically for this test
+        connector_config = {
+            **jira_connector_config,
+            "project_key": "AS",  # DailyConnectorTestProject
+        }
+
+        connector = Connector(
+            name="Test Jira Doc Sync Connector",
+            source=DocumentSource.JIRA,
+            input_type=InputType.POLL,
+            connector_specific_config=connector_config,
+            refresh_freq=None,
+            prune_freq=None,
+            indexing_start=None,
+        )
+        db_session.add(connector)
+        db_session.flush()
+
+        credential = Credential(
+            source=DocumentSource.JIRA,
+            credential_json=jira_credential_json,
+        )
+        db_session.add(credential)
+        db_session.flush()
+
+        cc_pair = ConnectorCredentialPair(
+            connector_id=connector.id,
+            credential_id=credential.id,
+            name="Test Jira Doc Sync CC Pair",
+            status=ConnectorCredentialPairStatus.ACTIVE,
+            access_type=AccessType.SYNC,
+            auto_sync_options=None,
+        )
+        db_session.add(cc_pair)
+        db_session.flush()
+        db_session.refresh(cc_pair)
+
+        # Mock functions - we don't have existing docs in the test DB
+        def fetch_all_existing_docs_fn(
+            sort_order: SortOrder | None = None,
+        ) -> list[DocumentRow]:
+            return []
+
+        def fetch_all_existing_docs_ids_fn() -> list[str]:
+            return []
+
+        doc_sync_iter = jira_doc_sync(
+            cc_pair=cc_pair,
+            fetch_all_existing_docs_fn=fetch_all_existing_docs_fn,
+            fetch_all_existing_docs_ids_fn=fetch_all_existing_docs_ids_fn,
+        )
+
+        # Expected documents from the danswerai.atlassian.net Jira instance
+        # The AS project has applicationRole permission, so all docs should be public
+        _EXPECTED_JIRA_DOCS = [
+            DocExternalAccessSet(
+                doc_id="https://danswerai.atlassian.net/browse/AS-3",
+                external_user_emails=set(),
+                external_user_group_ids=set(),
+                is_public=True,
+            ),
+            DocExternalAccessSet(
+                doc_id="https://danswerai.atlassian.net/browse/AS-4",
+                external_user_emails=set(),
+                external_user_group_ids=set(),
+                is_public=True,
+            ),
+        ]
+
+        expected_docs = {doc.doc_id: doc for doc in _EXPECTED_JIRA_DOCS}
+        actual_docs = {
+            doc.doc_id: DocExternalAccessSet.from_doc_external_access(doc)
+            for doc in doc_sync_iter
+        }
+        assert expected_docs == actual_docs, (
+            f"Expected docs: {expected_docs}\n" f"Actual docs: {actual_docs}"
+        )
+    finally:
+        db_session.rollback()
+
+
+def test_jira_doc_sync_with_specific_permissions(
+    db_session: Session,
+    jira_connector_config: dict[str, Any],
+    jira_credential_json: dict[str, Any],
+) -> None:
+    """Test that Jira doc sync returns documents with specific permissions.
+
+    This test uses a project that has specific user permissions to verify
+    that specific users are correctly extracted.
+    """
+    # NOTE: must set EE on or else the connector will skip the perm syncing
+    global_version.set_ee()
+
+    try:
+        # Use SUP project which has specific user permissions
+        connector_config = {
+            **jira_connector_config,
+            "project_key": "SUP",
+        }
+
+        connector = Connector(
+            name="Test Jira Doc Sync with Groups Connector",
+            source=DocumentSource.JIRA,
+            input_type=InputType.POLL,
+            connector_specific_config=connector_config,
+            refresh_freq=None,
+            prune_freq=None,
+            indexing_start=None,
+        )
+        db_session.add(connector)
+        db_session.flush()
+
+        credential = Credential(
+            source=DocumentSource.JIRA,
+            credential_json=jira_credential_json,
+        )
+        db_session.add(credential)
+        db_session.flush()
+
+        cc_pair = ConnectorCredentialPair(
+            connector_id=connector.id,
+            credential_id=credential.id,
+            name="Test Jira Doc Sync with Groups CC Pair",
+            status=ConnectorCredentialPairStatus.ACTIVE,
+            access_type=AccessType.SYNC,
+            auto_sync_options=None,
+        )
+        db_session.add(cc_pair)
+        db_session.flush()
+        db_session.refresh(cc_pair)
+
+        # Mock functions
+        def fetch_all_existing_docs_fn(
+            sort_order: SortOrder | None = None,
+        ) -> list[DocumentRow]:
+            return []
+
+        def fetch_all_existing_docs_ids_fn() -> list[str]:
+            return []
+
+        doc_sync_iter = jira_doc_sync(
+            cc_pair=cc_pair,
+            fetch_all_existing_docs_fn=fetch_all_existing_docs_fn,
+            fetch_all_existing_docs_ids_fn=fetch_all_existing_docs_ids_fn,
+        )
+
+        docs = list(doc_sync_iter)
+
+        # SUP project should have user-specific permissions (not public)
+        assert len(docs) > 0, "Expected at least one document from SUP project"
+
+        _EXPECTED_USER_EMAILS = set(
+            ["yuhong@onyx.app", "chris@onyx.app", "founders@onyx.app"]
+        )
+        _EXPECTED_USER_GROUP_IDS = set(["jira-users-danswerai"])
+
+        for doc in docs:
+            assert doc.doc_id.startswith("https://danswerai.atlassian.net/browse/SUP-")
+            # SUP project has specific users assigned, not applicationRole
+            assert (
+                not doc.external_access.is_public
+            ), f"Document {doc.doc_id} should not be public"
+            # Should have user emails
+            assert doc.external_access.external_user_emails == _EXPECTED_USER_EMAILS
+            assert (
+                doc.external_access.external_user_group_ids == _EXPECTED_USER_GROUP_IDS
+            )
+    finally:
+        db_session.rollback()

backend/tests/external_dependency_unit/connectors/jira/test_jira_group_sync.py

@@ -0,0 +1,133 @@
+from typing import Any
+
+from sqlalchemy.orm import Session
+
+from ee.onyx.external_permissions.jira.group_sync import jira_group_sync
+from onyx.configs.constants import DocumentSource
+from onyx.connectors.models import InputType
+from onyx.db.enums import AccessType
+from onyx.db.enums import ConnectorCredentialPairStatus
+from onyx.db.models import Connector
+from onyx.db.models import ConnectorCredentialPair
+from onyx.db.models import Credential
+from shared_configs.contextvars import get_current_tenant_id
+from tests.daily.connectors.confluence.models import ExternalUserGroupSet
+
+
+# In order to get these tests to run, use the credentials from Bitwarden.
+# Search up "ENV vars for local and Github tests", and find the Jira relevant key-value pairs.
+# Required env vars: JIRA_USER_EMAIL, JIRA_API_TOKEN
+
+# Expected groups from the danswerai.atlassian.net Jira instance
+# Note: These groups are shared with Confluence since they're both Atlassian products
+# App accounts (bots, integrations) are filtered out
+_EXPECTED_JIRA_GROUPS = [
+    ExternalUserGroupSet(
+        id="Yuhong Only No Chris Allowed",
+        user_emails={"yuhong@onyx.app"},
+        gives_anyone_access=False,
+    ),
+    ExternalUserGroupSet(
+        id="confluence-admins-danswerai",
+        user_emails={"chris@onyx.app", "yuhong@onyx.app"},
+        gives_anyone_access=False,
+    ),
+    ExternalUserGroupSet(
+        id="confluence-user-access-admins-danswerai",
+        user_emails={"hagen@danswer.ai"},
+        gives_anyone_access=False,
+    ),
+    ExternalUserGroupSet(
+        id="confluence-users-danswerai",
+        user_emails={
+            "chris@onyx.app",
+            "founders@onyx.app",
+            "hagen@danswer.ai",
+            "pablo@onyx.app",
+            "yuhong@onyx.app",
+        },
+        gives_anyone_access=False,
+    ),
+    ExternalUserGroupSet(
+        id="jira-admins-danswerai",
+        user_emails={"founders@onyx.app", "hagen@danswer.ai", "pablo@onyx.app"},
+        gives_anyone_access=False,
+    ),
+    ExternalUserGroupSet(
+        id="jira-user-access-admins-danswerai",
+        user_emails={"hagen@danswer.ai"},
+        gives_anyone_access=False,
+    ),
+    ExternalUserGroupSet(
+        id="jira-users-danswerai",
+        user_emails={
+            "chris@onyx.app",
+            "founders@onyx.app",
+            "hagen@danswer.ai",
+            "pablo@onyx.app",
+        },
+        gives_anyone_access=False,
+    ),
+    ExternalUserGroupSet(
+        id="org-admins",
+        user_emails={
+            "chris@onyx.app",
+            "founders@onyx.app",
+            "yuhong@onyx.app",
+        },
+        gives_anyone_access=False,
+    ),
+]
+
+
+def test_jira_group_sync(
+    db_session: Session,
+    jira_connector_config: dict[str, Any],
+    jira_credential_json: dict[str, Any],
+) -> None:
+    try:
+        connector = Connector(
+            name="Test Jira Connector",
+            source=DocumentSource.JIRA,
+            input_type=InputType.POLL,
+            connector_specific_config=jira_connector_config,
+            refresh_freq=None,
+            prune_freq=None,
+            indexing_start=None,
+        )
+        db_session.add(connector)
+        db_session.flush()
+
+        credential = Credential(
+            source=DocumentSource.JIRA,
+            credential_json=jira_credential_json,
+        )
+        db_session.add(credential)
+        db_session.flush()
+
+        cc_pair = ConnectorCredentialPair(
+            connector_id=connector.id,
+            credential_id=credential.id,
+            name="Test Jira CC Pair",
+            status=ConnectorCredentialPairStatus.ACTIVE,
+            access_type=AccessType.SYNC,
+            auto_sync_options=None,
+        )
+        db_session.add(cc_pair)
+        db_session.flush()
+        db_session.refresh(cc_pair)
+
+        tenant_id = get_current_tenant_id()
+        group_sync_iter = jira_group_sync(
+            tenant_id=tenant_id,
+            cc_pair=cc_pair,
+        )
+
+        expected_groups = {group.id: group for group in _EXPECTED_JIRA_GROUPS}
+        actual_groups = {
+            group.id: ExternalUserGroupSet.from_model(external_user_group=group)
+            for group in group_sync_iter
+        }
+        assert expected_groups == actual_groups
+    finally:
+        db_session.rollback()

backend/tests/external_dependency_unit/db/test_tag_race_condition.py

@@ -0,0 +1,283 @@
+"""
+Test suite for tag creation race condition handling.
+
+Tests that concurrent tag creation operations don't fail due to
+UniqueViolation errors, which would occur if the upsert logic
+isn't properly implemented.
+"""
+
+from concurrent.futures import as_completed
+from concurrent.futures import Future
+from concurrent.futures import ThreadPoolExecutor
+from typing import Union
+from uuid import uuid4
+
+from sqlalchemy import select
+from sqlalchemy.orm import Session
+
+from onyx.configs.constants import DocumentSource
+from onyx.db.engine.sql_engine import get_session_with_current_tenant
+from onyx.db.models import Document
+from onyx.db.models import Tag
+from onyx.db.tag import create_or_add_document_tag
+from onyx.db.tag import create_or_add_document_tag_list
+
+
+def _create_test_document(db_session: Session, doc_id: str) -> Document:
+    """Create a minimal test document."""
+    document = Document(
+        id=doc_id,
+        semantic_id=f"semantic_{doc_id}",
+        boost=0,
+        hidden=False,
+        from_ingestion_api=False,
+    )
+    db_session.add(document)
+    db_session.commit()
+    return document
+
+
+class TestTagRaceCondition:
+    """Tests for tag creation race condition handling."""
+
+    def test_concurrent_tag_creation_single_tag(
+        self, db_session: Session, tenant_context: None
+    ) -> None:
+        """
+        Test that multiple concurrent calls to create_or_add_document_tag
+        with the same tag key/value all succeed without UniqueViolation errors.
+
+        This simulates the race condition that occurs when multiple workers
+        try to create the same tag simultaneously during document indexing.
+        """
+        # Create multiple test documents that will all get the same tag
+        num_documents = 20
+        doc_ids = [f"test_doc_race_{uuid4().hex[:8]}" for _ in range(num_documents)]
+
+        for doc_id in doc_ids:
+            _create_test_document(db_session, doc_id)
+
+        # Use a unique tag key/value for this test run to avoid interference
+        test_tag_key = f"test_key_{uuid4().hex[:8]}"
+        test_tag_value = f"test_value_{uuid4().hex[:8]}"
+        test_source = DocumentSource.FILE
+
+        errors: list[Exception] = []
+        results: list[Tag | None] = []
+
+        def create_tag_for_document(doc_id: str) -> Tag | None:
+            """Worker function that creates a tag for a document using its own session."""
+            with get_session_with_current_tenant() as session:
+                return create_or_add_document_tag(
+                    tag_key=test_tag_key,
+                    tag_value=test_tag_value,
+                    source=test_source,
+                    document_id=doc_id,
+                    db_session=session,
+                )
+
+        # Run all tag creations concurrently with high parallelism
+        with ThreadPoolExecutor(max_workers=num_documents) as executor:
+            futures = {
+                executor.submit(create_tag_for_document, doc_id): doc_id
+                for doc_id in doc_ids
+            }
+
+            for future in as_completed(futures):
+                doc_id = futures[future]
+                try:
+                    result = future.result()
+                    results.append(result)
+                except Exception as e:
+                    errors.append(e)
+
+        # All operations should succeed without errors
+        assert len(errors) == 0, f"Got {len(errors)} errors: {errors}"
+        assert len(results) == num_documents
+
+        # All results should be valid Tag objects
+        for result in results:
+            assert result is not None
+            assert result.tag_key == test_tag_key
+            assert result.tag_value == test_tag_value
+            assert result.source == test_source
+
+        # Verify only ONE tag was created in the database (not num_documents tags)
+        with get_session_with_current_tenant() as session:
+            tag_count = (
+                session.execute(
+                    select(Tag).where(
+                        Tag.tag_key == test_tag_key,
+                        Tag.tag_value == test_tag_value,
+                        Tag.source == test_source,
+                    )
+                )
+                .scalars()
+                .all()
+            )
+
+        assert len(tag_count) == 1, f"Expected 1 tag, found {len(tag_count)}"
+
+    def test_concurrent_tag_list_creation(
+        self, db_session: Session, tenant_context: None
+    ) -> None:
+        """
+        Test that multiple concurrent calls to create_or_add_document_tag_list
+        with the same tag values all succeed without UniqueViolation errors.
+        """
+        # Create multiple test documents
+        num_documents = 20
+        doc_ids = [
+            f"test_doc_list_race_{uuid4().hex[:8]}" for _ in range(num_documents)
+        ]
+
+        for doc_id in doc_ids:
+            _create_test_document(db_session, doc_id)
+
+        # Use unique tag key/values for this test run
+        test_tag_key = f"test_list_key_{uuid4().hex[:8]}"
+        test_tag_values = [f"value_{i}_{uuid4().hex[:4]}" for i in range(5)]
+        test_source = DocumentSource.FILE
+
+        errors: list[Exception] = []
+        results: list[list[Tag]] = []
+
+        def create_tag_list_for_document(doc_id: str) -> list[Tag]:
+            """Worker function that creates tag list for a document using its own session."""
+            with get_session_with_current_tenant() as session:
+                return create_or_add_document_tag_list(
+                    tag_key=test_tag_key,
+                    tag_values=test_tag_values,
+                    source=test_source,
+                    document_id=doc_id,
+                    db_session=session,
+                )
+
+        # Run all tag creations concurrently
+        with ThreadPoolExecutor(max_workers=num_documents) as executor:
+            futures = {
+                executor.submit(create_tag_list_for_document, doc_id): doc_id
+                for doc_id in doc_ids
+            }
+
+            for future in as_completed(futures):
+                doc_id = futures[future]
+                try:
+                    result = future.result()
+                    results.append(result)
+                except Exception as e:
+                    errors.append(e)
+
+        # All operations should succeed without errors
+        assert len(errors) == 0, f"Got {len(errors)} errors: {errors}"
+        assert len(results) == num_documents
+
+        # Each result should have all the expected tags
+        for result in results:
+            assert len(result) == len(test_tag_values)
+            result_values = {tag.tag_value for tag in result}
+            assert result_values == set(test_tag_values)
+
+        # Verify exactly len(test_tag_values) tags were created (one per value)
+        with get_session_with_current_tenant() as session:
+            tags = (
+                session.execute(
+                    select(Tag).where(
+                        Tag.tag_key == test_tag_key,
+                        Tag.tag_value.in_(test_tag_values),
+                        Tag.source == test_source,
+                    )
+                )
+                .scalars()
+                .all()
+            )
+
+        assert len(tags) == len(
+            test_tag_values
+        ), f"Expected {len(test_tag_values)} tags, found {len(tags)}"
+
+    def test_concurrent_mixed_tag_operations(
+        self, db_session: Session, tenant_context: None
+    ) -> None:
+        """
+        Test that concurrent single tag and tag list operations on the same
+        tag key/value don't interfere with each other.
+
+        This is a more realistic scenario where different documents might
+        have the same metadata key but different value types (single vs list).
+        """
+        num_documents = 10
+        doc_ids_single = [
+            f"test_doc_single_{uuid4().hex[:8]}" for _ in range(num_documents)
+        ]
+        doc_ids_list = [
+            f"test_doc_list_{uuid4().hex[:8]}" for _ in range(num_documents)
+        ]
+
+        for doc_id in doc_ids_single + doc_ids_list:
+            _create_test_document(db_session, doc_id)
+
+        # Same key but used as both single value and list value
+        test_tag_key = f"mixed_key_{uuid4().hex[:8]}"
+        test_single_value = f"single_value_{uuid4().hex[:8]}"
+        test_list_values = [test_single_value]  # Same value but as list
+        test_source = DocumentSource.FILE
+
+        errors: list[Exception] = []
+
+        def create_single_tag(doc_id: str) -> Tag | None:
+            with get_session_with_current_tenant() as session:
+                return create_or_add_document_tag(
+                    tag_key=test_tag_key,
+                    tag_value=test_single_value,
+                    source=test_source,
+                    document_id=doc_id,
+                    db_session=session,
+                )
+
+        def create_list_tag(doc_id: str) -> list[Tag]:
+            with get_session_with_current_tenant() as session:
+                return create_or_add_document_tag_list(
+                    tag_key=test_tag_key,
+                    tag_values=test_list_values,
+                    source=test_source,
+                    document_id=doc_id,
+                    db_session=session,
+                )
+
+        # Run both types of operations concurrently
+        with ThreadPoolExecutor(max_workers=num_documents * 2) as executor:
+            futures: list[Future[Union[Tag | None] | list[Tag]]] = []
+            for doc_id in doc_ids_single:
+                futures.append(executor.submit(create_single_tag, doc_id))
+            for doc_id in doc_ids_list:
+                futures.append(executor.submit(create_list_tag, doc_id))
+
+            for future in as_completed(futures):
+                try:
+                    future.result()
+                except Exception as e:
+                    errors.append(e)
+
+        # All operations should succeed
+        assert len(errors) == 0, f"Got {len(errors)} errors: {errors}"
+
+        # Should have exactly 2 tags: one with is_list=False, one with is_list=True
+        with get_session_with_current_tenant() as session:
+            tags = (
+                session.execute(
+                    select(Tag).where(
+                        Tag.tag_key == test_tag_key,
+                        Tag.tag_value == test_single_value,
+                        Tag.source == test_source,
+                    )
+                )
+                .scalars()
+                .all()
+            )
+
+        assert (
+            len(tags) == 2
+        ), f"Expected 2 tags (is_list=True and False), found {len(tags)}"
+        is_list_values = {tag.is_list for tag in tags}
+        assert is_list_values == {True, False}

backend/tests/external_dependency_unit/slack_bot/test_slack_bot_federated_search.py

@@ -6,9 +6,11 @@ from unittest.mock import patch
 from uuid import uuid4
 
 # Set environment variables to disable model server for testing
+os.environ["DISABLE_MODEL_SERVER"] = "true"
 os.environ["MODEL_SERVER_HOST"] = "disabled"
 os.environ["MODEL_SERVER_PORT"] = "9000"
 
+from sqlalchemy import inspect
 from sqlalchemy.orm import Session
 from slack_sdk.errors import SlackApiError
 
@@ -760,3 +762,76 @@ def test_multiple_missing_scopes_resilience(
     # Should still return available channels
     assert len(result) == 1, f"Expected 1 channel, got {len(result)}"
     assert result["C1234567890"]["name"] == "general"
+
+
+def test_slack_channel_config_eager_loads_persona(db_session: Session) -> None:
+    """Test that fetch_slack_channel_config_for_channel_or_default eagerly loads persona.
+
+    This prevents lazy loading failures when the session context changes later
+    in the request handling flow (e.g., in handle_regular_answer).
+    """
+    from onyx.db.slack_channel_config import (
+        fetch_slack_channel_config_for_channel_or_default,
+    )
+
+    unique_id = str(uuid4())[:8]
+
+    # Create a persona (using same fields as _create_test_persona_with_slack_config)
+    persona = Persona(
+        name=f"test_eager_load_persona_{unique_id}",
+        description="Test persona for eager loading test",
+        chunks_above=0,
+        chunks_below=0,
+        llm_relevance_filter=True,
+        llm_filter_extraction=True,
+        recency_bias=RecencyBiasSetting.AUTO,
+        system_prompt="You are a helpful assistant.",
+        task_prompt="Answer the user's question.",
+    )
+    db_session.add(persona)
+    db_session.flush()
+
+    # Create a slack bot
+    slack_bot = SlackBot(
+        name=f"Test Bot {unique_id}",
+        bot_token=f"xoxb-test-{unique_id}",
+        app_token=f"xapp-test-{unique_id}",
+        enabled=True,
+    )
+    db_session.add(slack_bot)
+    db_session.flush()
+
+    # Create slack channel config with persona
+    channel_name = f"test-channel-{unique_id}"
+    slack_channel_config = SlackChannelConfig(
+        slack_bot_id=slack_bot.id,
+        persona_id=persona.id,
+        channel_config={"channel_name": channel_name, "disabled": False},
+        enable_auto_filters=False,
+        is_default=False,
+    )
+    db_session.add(slack_channel_config)
+    db_session.commit()
+
+    # Fetch the config using the function under test
+    fetched_config = fetch_slack_channel_config_for_channel_or_default(
+        db_session=db_session,
+        slack_bot_id=slack_bot.id,
+        channel_name=channel_name,
+    )
+
+    assert fetched_config is not None, "Should find the channel config"
+
+    # Check that persona relationship is already loaded (not pending lazy load)
+    insp = inspect(fetched_config)
+    assert insp is not None, "Should be able to inspect the config"
+    assert "persona" not in insp.unloaded, (
+        "Persona should be eagerly loaded, not pending lazy load. "
+        "This is required to prevent fallback to default persona when "
+        "session context changes in handle_regular_answer."
+    )
+
+    # Verify the persona is correct
+    assert fetched_config.persona is not None, "Persona should not be None"
+    assert fetched_config.persona.id == persona.id, "Should load the correct persona"
+    assert fetched_config.persona.name == persona.name

backend/tests/integration/tests/indexing/test_repeated_error_state.py

@@ -11,7 +11,6 @@ from onyx.connectors.mock_connector.connector import MockConnectorCheckpoint
 from onyx.connectors.models import InputType
 from onyx.db.connector_credential_pair import get_connector_credential_pair_from_id
 from onyx.db.engine.sql_engine import get_session_with_current_tenant
-from onyx.db.enums import ConnectorCredentialPairStatus
 from onyx.db.enums import IndexingStatus
 from tests.integration.common_utils.constants import MOCK_CONNECTOR_SERVER_HOST
 from tests.integration.common_utils.constants import MOCK_CONNECTOR_SERVER_PORT
@@ -122,11 +121,13 @@ def test_repeated_error_state_detection_and_recovery(
             )
             assert cc_pair_obj is not None
             if cc_pair_obj.in_repeated_error_state:
-                # Verify the connector is also paused to prevent further indexing attempts
-                assert cc_pair_obj.status == ConnectorCredentialPairStatus.PAUSED, (
-                    f"Expected status to be PAUSED when in repeated error state, "
-                    f"but got {cc_pair_obj.status}"
-                )
+                # Pausing only happens for cloud deployments and the IT don't run with
+                # that auth type :(
+                # if AUTH_TYPE == AuthType.CLOUD:
+                #     assert cc_pair_obj.status == ConnectorCredentialPairStatus.PAUSED, (
+                #         f"Expected status to be PAUSED when in repeated error state, "
+                #         f"but got {cc_pair_obj.status}"
+                #     )
                 break
 
         if time.monotonic() - start_time > 30:

backend/tests/unit/onyx/connectors/sharepoint/test_drive_matching.py

@@ -0,0 +1,200 @@
+from __future__ import annotations
+
+from collections import deque
+from collections.abc import Sequence
+from datetime import datetime
+from types import SimpleNamespace
+from typing import Any
+
+import pytest
+
+from onyx.connectors.sharepoint.connector import SHARED_DOCUMENTS_MAP
+from onyx.connectors.sharepoint.connector import SharepointConnector
+from onyx.connectors.sharepoint.connector import SharepointConnectorCheckpoint
+from onyx.connectors.sharepoint.connector import SiteDescriptor
+
+
+class _FakeQuery:
+    def __init__(self, payload: Sequence[Any]) -> None:
+        self._payload = payload
+
+    def execute_query(self) -> Sequence[Any]:
+        return self._payload
+
+
+class _FakeFolder:
+    def __init__(self, items: Sequence[Any]) -> None:
+        self._items = items
+        self.name = "root"
+
+    def get_by_path(self, _path: str) -> _FakeFolder:
+        return self
+
+    def get_files(
+        self, *, recursive: bool, page_size: int
+    ) -> _FakeQuery:  # noqa: ARG002
+        return _FakeQuery(self._items)
+
+
+class _FakeDrive:
+    def __init__(self, name: str, items: Sequence[Any]) -> None:
+        self.name = name
+        self.root = _FakeFolder(items)
+
+
+class _FakeDrivesCollection:
+    def __init__(self, drives: Sequence[_FakeDrive]) -> None:
+        self._drives = drives
+
+    def get(self) -> _FakeQuery:
+        return _FakeQuery(list(self._drives))
+
+
+class _FakeSite:
+    def __init__(self, drives: Sequence[_FakeDrive]) -> None:
+        self.drives = _FakeDrivesCollection(drives)
+
+
+class _FakeSites:
+    def __init__(self, drives: Sequence[_FakeDrive]) -> None:
+        self._drives = drives
+
+    def get_by_url(self, _url: str) -> _FakeSite:
+        return _FakeSite(self._drives)
+
+
+class _FakeGraphClient:
+    def __init__(self, drives: Sequence[_FakeDrive]) -> None:
+        self.sites = _FakeSites(drives)
+
+
+def _build_connector(drives: Sequence[_FakeDrive]) -> SharepointConnector:
+    connector = SharepointConnector()
+    connector._graph_client = _FakeGraphClient(drives)  # type: ignore[assignment]
+    return connector
+
+
+@pytest.mark.parametrize(
+    ("requested_drive_name", "graph_drive_name"),
+    [
+        ("Shared Documents", "Documents"),
+        ("Freigegebene Dokumente", "Dokumente"),
+        ("Documentos compartidos", "Documentos"),
+    ],
+)
+def test_fetch_driveitems_matches_international_drive_names(
+    requested_drive_name: str, graph_drive_name: str
+) -> None:
+    item = SimpleNamespace(parent_reference=SimpleNamespace(path=None))
+    connector = _build_connector([_FakeDrive(graph_drive_name, [item])])
+    site_descriptor = SiteDescriptor(
+        url="https://example.sharepoint.com/sites/sample",
+        drive_name=requested_drive_name,
+        folder_path=None,
+    )
+
+    results = connector._fetch_driveitems(site_descriptor=site_descriptor)
+
+    assert len(results) == 1
+    drive_item, returned_drive_name = results[0]
+    assert drive_item is item
+    assert returned_drive_name == requested_drive_name
+
+
+@pytest.mark.parametrize(
+    ("requested_drive_name", "graph_drive_name"),
+    [
+        ("Shared Documents", "Documents"),
+        ("Freigegebene Dokumente", "Dokumente"),
+        ("Documentos compartidos", "Documentos"),
+    ],
+)
+def test_get_drive_items_for_drive_name_matches_map(
+    requested_drive_name: str, graph_drive_name: str
+) -> None:
+    item = SimpleNamespace()
+    connector = _build_connector([_FakeDrive(graph_drive_name, [item])])
+    site_descriptor = SiteDescriptor(
+        url="https://example.sharepoint.com/sites/sample",
+        drive_name=requested_drive_name,
+        folder_path=None,
+    )
+
+    results = connector._get_drive_items_for_drive_name(
+        site_descriptor=site_descriptor,
+        drive_name=requested_drive_name,
+    )
+
+    assert len(results) == 1
+    assert results[0] is item
+
+
+def test_load_from_checkpoint_maps_drive_name(monkeypatch: pytest.MonkeyPatch) -> None:
+    connector = SharepointConnector()
+    connector._graph_client = object()  # type: ignore[assignment]
+    connector.include_site_pages = False
+
+    captured_drive_names: list[str] = []
+
+    def fake_get_drive_items(
+        self: SharepointConnector,
+        site_descriptor: SiteDescriptor,
+        drive_name: str,
+        start: datetime | None,
+        end: datetime | None,
+    ) -> list[SimpleNamespace]:
+        assert drive_name == "Documents"
+        return [
+            SimpleNamespace(
+                name="sample.pdf",
+                web_url="https://example.sharepoint.com/sites/sample/sample.pdf",
+            )
+        ]
+
+    def fake_convert(
+        driveitem: SimpleNamespace,
+        drive_name: str,
+        ctx: Any,
+        graph_client: Any,
+        include_permissions: bool,
+    ) -> SimpleNamespace:
+        captured_drive_names.append(drive_name)
+        return SimpleNamespace(sections=["content"])
+
+    monkeypatch.setattr(
+        SharepointConnector,
+        "_get_drive_items_for_drive_name",
+        fake_get_drive_items,
+    )
+    monkeypatch.setattr(
+        "onyx.connectors.sharepoint.connector._convert_driveitem_to_document_with_permissions",
+        fake_convert,
+    )
+
+    checkpoint = SharepointConnectorCheckpoint(has_more=True)
+    checkpoint.cached_site_descriptors = deque()
+    checkpoint.current_site_descriptor = SiteDescriptor(
+        url="https://example.sharepoint.com/sites/sample",
+        drive_name=SHARED_DOCUMENTS_MAP["Documents"],
+        folder_path=None,
+    )
+    checkpoint.cached_drive_names = deque(["Documents"])
+    checkpoint.current_drive_name = None
+    checkpoint.process_site_pages = False
+
+    generator = connector._load_from_checkpoint(
+        start=0,
+        end=0,
+        checkpoint=checkpoint,
+        include_permissions=False,
+    )
+
+    documents: list[Any] = []
+    try:
+        while True:
+            documents.append(next(generator))
+    except StopIteration:
+        pass
+
+    assert len(documents) == 1
+    assert captured_drive_names == [SHARED_DOCUMENTS_MAP["Documents"]]

backend/tests/unit/onyx/llm/test_true_openai_model.py

@@ -128,3 +128,13 @@ class TestIsTrueOpenAIModel:
         model_map = get_model_map()
         if "openai/gpt-3.5-turbo-instruct" in model_map:
             assert is_true_openai_model("openai", "gpt-3.5-turbo-instruct") is True
+
+    def test_azure_openai_model_through_litellm_proxy(self) -> None:
+        """Test that Azure OpenAI models are correctly identified."""
+        assert is_true_openai_model("litellm_proxy", "gpt-4") is True
+        assert is_true_openai_model("litellm_proxy", "gpt-5") is True
+        assert is_true_openai_model("litellm_proxy", "gpt-5.1") is True
+
+        assert is_true_openai_model("litellm_proxy", "azure/gpt-4") is True
+        assert is_true_openai_model("litellm_proxy", "azure/gpt-5") is True
+        assert is_true_openai_model("litellm_proxy", "azure/gpt-5.1") is True

backend/tests/unit/onyx/natural_language_processing/test_search_nlp_models.py

@@ -1,3 +1,4 @@
+import os
 from collections.abc import AsyncGenerator
 from typing import List
 from unittest.mock import AsyncMock
@@ -9,6 +10,12 @@ from httpx import AsyncClient
 from litellm.exceptions import RateLimitError
 
 from onyx.natural_language_processing.search_nlp_models import CloudEmbedding
+from onyx.natural_language_processing.search_nlp_models import (
+    ConnectorClassificationModel,
+)
+from onyx.natural_language_processing.search_nlp_models import (
+    InformationContentClassificationModel,
+)
 from shared_configs.enums import EmbeddingProvider
 from shared_configs.enums import EmbedTextType
 
@@ -81,3 +88,95 @@ async def test_rate_limit_handling() -> None:
                 model_name="fake-model",
                 text_type=EmbedTextType.QUERY,
             )
+
+
+class TestInformationContentClassificationModel:
+    """Test cases for InformationContentClassificationModel with DISABLE_MODEL_SERVER"""
+
+    @patch.dict(os.environ, {"DISABLE_MODEL_SERVER": "true"})
+    def test_predict_with_disable_model_server(self) -> None:
+        """Test that predict returns default classifications when DISABLE_MODEL_SERVER is true"""
+        model = InformationContentClassificationModel()
+        queries = ["What is AI?", "How does Python work?"]
+
+        results = model.predict(queries)
+
+        assert len(results) == 2
+        for result in results:
+            assert result.predicted_label == 1  # 1 indicates informational content
+            assert result.content_boost_factor == 1.0  # Default boost factor
+
+    @patch.dict(os.environ, {"DISABLE_MODEL_SERVER": "false"})
+    @patch("requests.post")
+    def test_predict_with_model_server_enabled(self, mock_post: MagicMock) -> None:
+        """Test that predict makes request when DISABLE_MODEL_SERVER is false"""
+        mock_response = MagicMock()
+        mock_response.json.return_value = [
+            {"predicted_label": 1, "content_boost_factor": 1.0},
+            {"predicted_label": 0, "content_boost_factor": 0.8},
+        ]
+        mock_post.return_value = mock_response
+
+        model = InformationContentClassificationModel()
+        queries = ["test1", "test2"]
+
+        results = model.predict(queries)
+
+        assert len(results) == 2
+        assert results[0].predicted_label == 1
+        assert results[0].content_boost_factor == 1.0
+        assert results[1].predicted_label == 0
+        assert results[1].content_boost_factor == 0.8
+        mock_post.assert_called_once()
+
+
+class TestConnectorClassificationModel:
+    """Test cases for ConnectorClassificationModel with DISABLE_MODEL_SERVER"""
+
+    @patch.dict(os.environ, {"DISABLE_MODEL_SERVER": "true"})
+    def test_predict_with_disable_model_server(self) -> None:
+        """Test that predict returns all connectors when DISABLE_MODEL_SERVER is true"""
+        model = ConnectorClassificationModel()
+        query = "Search for documentation"
+        available_connectors = ["confluence", "slack", "github"]
+
+        results = model.predict(query, available_connectors)
+
+        assert results == available_connectors
+
+    @patch.dict(os.environ, {"DISABLE_MODEL_SERVER": "false"})
+    @patch("requests.post")
+    def test_predict_with_model_server_enabled(self, mock_post: MagicMock) -> None:
+        """Test that predict makes request when DISABLE_MODEL_SERVER is false"""
+        mock_response = MagicMock()
+        mock_response.json.return_value = {"connectors": ["confluence", "github"]}
+        mock_post.return_value = mock_response
+
+        model = ConnectorClassificationModel()
+        query = "Search for documentation"
+        available_connectors = ["confluence", "slack", "github"]
+
+        results = model.predict(query, available_connectors)
+
+        assert results == ["confluence", "github"]
+        mock_post.assert_called_once()
+
+    @patch.dict(os.environ, {"DISABLE_MODEL_SERVER": "1"})
+    @patch("requests.post")
+    def test_predict_with_disable_model_server_numeric(
+        self, mock_post: MagicMock
+    ) -> None:
+        """Test that predict makes request when DISABLE_MODEL_SERVER is 1 (not 'true')"""
+        # "1" should NOT trigger disable (only "true" should)
+        mock_response = MagicMock()
+        mock_response.json.return_value = {"connectors": ["github"]}
+        mock_post.return_value = mock_response
+
+        model = ConnectorClassificationModel()
+        query = "Find issues"
+        available_connectors = ["jira", "github"]
+
+        results = model.predict(query, available_connectors)
+
+        assert results == ["github"]
+        mock_post.assert_called_once()

backend/tests/unit/onyx/utils/test_gpu_utils.py

@@ -0,0 +1,103 @@
+"""
+Test cases for onyx/utils/gpu_utils.py with DISABLE_MODEL_SERVER environment variable
+"""
+
+import os
+from unittest import TestCase
+from unittest.mock import MagicMock
+from unittest.mock import patch
+
+import requests
+
+from onyx.utils.gpu_utils import _get_gpu_status_from_model_server
+
+
+class TestGPUUtils(TestCase):
+    """Test cases for GPU utilities with DISABLE_MODEL_SERVER support"""
+
+    @patch.dict(os.environ, {"DISABLE_MODEL_SERVER": "true"})
+    def test_disable_model_server_true(self) -> None:
+        """Test that GPU status returns False when DISABLE_MODEL_SERVER is true"""
+        result = _get_gpu_status_from_model_server(indexing=False)
+        assert result is False
+
+    @patch.dict(os.environ, {"DISABLE_MODEL_SERVER": "True"})
+    def test_disable_model_server_capital_true(self) -> None:
+        """Test that GPU status returns False when DISABLE_MODEL_SERVER is True (capital)"""
+        # "True" WILL trigger disable because .lower() is called
+        result = _get_gpu_status_from_model_server(indexing=False)
+        assert result is False
+
+    @patch.dict(os.environ, {"DISABLE_MODEL_SERVER": "1"})
+    @patch("requests.get")
+    def test_disable_model_server_one(self, mock_get: MagicMock) -> None:
+        """Test that GPU status makes request when DISABLE_MODEL_SERVER is 1"""
+        # "1" should NOT trigger disable (only "true" should)
+        mock_response = MagicMock()
+        mock_response.json.return_value = {"gpu_available": True}
+        mock_get.return_value = mock_response
+
+        result = _get_gpu_status_from_model_server(indexing=False)
+        assert result is True
+        mock_get.assert_called_once()
+
+    @patch.dict(os.environ, {"DISABLE_MODEL_SERVER": "yes"})
+    @patch("requests.get")
+    def test_disable_model_server_yes(self, mock_get: MagicMock) -> None:
+        """Test that GPU status makes request when DISABLE_MODEL_SERVER is yes"""
+        # "yes" should NOT trigger disable (only "true" should)
+        mock_response = MagicMock()
+        mock_response.json.return_value = {"gpu_available": False}
+        mock_get.return_value = mock_response
+
+        result = _get_gpu_status_from_model_server(indexing=True)
+        assert result is False
+        mock_get.assert_called_once()
+
+    @patch.dict(os.environ, {"DISABLE_MODEL_SERVER": "false"})
+    @patch("requests.get")
+    def test_disable_model_server_false(self, mock_get: MagicMock) -> None:
+        """Test that GPU status makes request when DISABLE_MODEL_SERVER is false"""
+        mock_response = MagicMock()
+        mock_response.json.return_value = {"gpu_available": True}
+        mock_get.return_value = mock_response
+
+        result = _get_gpu_status_from_model_server(indexing=True)
+        assert result is True
+        mock_get.assert_called_once()
+
+    @patch.dict(os.environ, {}, clear=True)
+    @patch("requests.get")
+    def test_disable_model_server_not_set(self, mock_get: MagicMock) -> None:
+        """Test that GPU status makes request when DISABLE_MODEL_SERVER is not set"""
+        mock_response = MagicMock()
+        mock_response.json.return_value = {"gpu_available": False}
+        mock_get.return_value = mock_response
+
+        result = _get_gpu_status_from_model_server(indexing=False)
+        assert result is False
+        mock_get.assert_called_once()
+
+    @patch.dict(os.environ, {"DISABLE_MODEL_SERVER": "true"})
+    def test_disabled_host_fallback(self) -> None:
+        """Test that disabled host is handled correctly via environment variable"""
+        result = _get_gpu_status_from_model_server(indexing=True)
+        assert result is False
+
+    @patch.dict(os.environ, {"DISABLE_MODEL_SERVER": "false"})
+    @patch("requests.get")
+    def test_request_exception_handling(self, mock_get: MagicMock) -> None:
+        """Test that exceptions are properly raised when GPU status request fails"""
+        mock_get.side_effect = requests.RequestException("Connection error")
+
+        with self.assertRaises(requests.RequestException):
+            _get_gpu_status_from_model_server(indexing=False)
+
+    @patch.dict(os.environ, {"DISABLE_MODEL_SERVER": "true"})
+    @patch("requests.get")
+    def test_gpu_status_request_with_disable(self, mock_get: MagicMock) -> None:
+        """Test that no request is made when DISABLE_MODEL_SERVER is true"""
+        result = _get_gpu_status_from_model_server(indexing=True)
+        assert result is False
+        # Verify that no HTTP request was made
+        mock_get.assert_not_called()

deployment/docker_compose/env.prod.template

@@ -63,7 +63,7 @@ DB_READONLY_PASSWORD=password
 #VESPA_LANGUAGE_OVERRIDE=
 
 # Show extra/uncommon connectors
-# See https://docs.onyx.app/admin/connectors/overview for a full list of connectors
+# See https://docs.onyx.app/admins/connectors/overview for a full list of connectors
 SHOW_EXTRA_CONNECTORS=False
 
 # User File Upload Configuration

pyproject.toml

@@ -123,7 +123,7 @@ backend = [
     "mistune==0.8.4",
     "sendgrid==6.11.0",
     "exa_py==1.15.4",
-    "braintrust[openai-agents]==0.2.6",
+    "braintrust[openai-agents]==0.3.9",
     "braintrust-langchain==0.0.4",
     "openai-agents==0.4.2",
     "langfuse==3.10.0",

uv.lock

@@ -533,7 +533,7 @@ wheels = [
 
 [[package]]
 name = "braintrust"
-version = "0.2.6"
+version = "0.3.9"
 source = { registry = "https://pypi.org/simple" }
 dependencies = [
     { name = "chevron" },
@@ -545,10 +545,11 @@ dependencies = [
     { name = "sseclient-py" },
     { name = "tqdm" },
     { name = "typing-extensions" },
+    { name = "wrapt" },
 ]
-sdist = { url = "https://files.pythonhosted.org/packages/e8/66/d8bf0a99a29ec7c6cbf47a1e9416ea2f9dcab7bf6d6539cd6a8e755ec22f/braintrust-0.2.6.tar.gz", hash = "sha256:e9c2ae3cce09a8562fa436bc6ec20039c495a6f82928ac36050b5fbc7b89743a", size = 180921, upload-time = "2025-08-27T18:13:01.804Z" }
+sdist = { url = "https://files.pythonhosted.org/packages/fb/5c/d4086bcf843a49916cefdd0379a138d9024d5fe75e6cea1e672da19cb272/braintrust-0.3.9.tar.gz", hash = "sha256:8c56ccb214bb102291968cdda67d7bf36a4b01fefc6b1998e53c272379eacab3", size = 237224, upload-time = "2025-11-25T22:20:24.368Z" }
 wheels = [
-    { url = "https://files.pythonhosted.org/packages/c2/86/8d4bb3b0ab6dd40d2d428bb0b48d29601ed2a344aa6c618378311a73844a/braintrust-0.2.6-py3-none-any.whl", hash = "sha256:d07f1666881c25a8d65a04cfb84f45a0016531928eb3133172942fae49f9a8f8", size = 210465, upload-time = "2025-08-27T18:13:00.503Z" },
+    { url = "https://files.pythonhosted.org/packages/ed/b6/e21ddf4e815c3fc845c61b5084bd4d35b882700d78ec054e9e9803aa04db/braintrust-0.3.9-py3-none-any.whl", hash = "sha256:2b903671837dea85d74e984403f154b60b77a3dde4c0bffa17121fdcb24f18d0", size = 280622, upload-time = "2025-11-25T22:20:22.553Z" },
 ]
 
 [package.optional-dependencies]
@@ -3656,7 +3657,7 @@ backend = [
     { name = "beautifulsoup4", specifier = "==4.12.3" },
     { name = "boto3", specifier = "==1.39.11" },
     { name = "boto3-stubs", extras = ["s3"], specifier = "==1.39.11" },
-    { name = "braintrust", extras = ["openai-agents"], specifier = "==0.2.6" },
+    { name = "braintrust", extras = ["openai-agents"], specifier = "==0.3.9" },
     { name = "braintrust-langchain", specifier = "==0.0.4" },
     { name = "celery", specifier = "==5.5.1" },
     { name = "chardet", specifier = "==5.2.0" },

web/package-lock.json

@@ -56,15 +56,15 @@
         "lowlight": "^3.3.0",
         "lucide-react": "^0.454.0",
         "mdast-util-find-and-replace": "^3.0.1",
-        "next": "^16.0.1",
+        "next": "^16.0.7",
         "next-themes": "^0.4.4",
         "npm": "^10.8.0",
         "postcss": "^8.5.6",
         "posthog-js": "^1.176.0",
-        "react": "^19.2.0",
+        "react": "^19.2.1",
         "react-datepicker": "^7.6.0",
         "react-day-picker": "^9.11.1",
-        "react-dom": "^19.2.0",
+        "react-dom": "^19.2.1",
         "react-dropzone": "^14.2.3",
         "react-icons": "^4.8.0",
         "react-loader-spinner": "^8.0.0",
@@ -101,8 +101,8 @@
         "@types/js-cookie": "^3.0.6",
         "@types/lodash": "^4.17.20",
         "@types/node": "18.15.11",
-        "@types/react": "19.2.2",
-        "@types/react-dom": "19.2.2",
+        "@types/react": "^19.2.2",
+        "@types/react-dom": "^19.2.2",
         "@types/uuid": "^9.0.8",
         "babel-plugin-react-compiler": "^1.0.0",
         "chromatic": "^11.25.2",
@@ -914,8 +914,6 @@
     },
     "node_modules/@eslint/eslintrc/node_modules/ajv": {
       "version": "6.12.6",
-      "resolved": "https://registry.npmjs.org/ajv/-/ajv-6.12.6.tgz",
-      "integrity": "sha512-j3fVLgvTo527anyYyJOGTYJbG+vnnQYvE0m5mmkc1TK+nxAppkCLMIL0aZ4dblVCNoGShhm+kzE4ZUykBoMg4g==",
       "dev": true,
       "license": "MIT",
       "dependencies": {
@@ -942,8 +940,6 @@
     },
     "node_modules/@eslint/eslintrc/node_modules/json-schema-traverse": {
       "version": "0.4.1",
-      "resolved": "https://registry.npmjs.org/json-schema-traverse/-/json-schema-traverse-0.4.1.tgz",
-      "integrity": "sha512-xbbCH5dCYU5T8LcEhhuh7HJ88HXuW3qsI3Y0zOZFKfZEHcpWiHU/Jxzk629Brsab/mMiHQti9wMP+845RPe3Vg==",
       "dev": true,
       "license": "MIT"
     },
@@ -1285,9 +1281,7 @@
       }
     },
     "node_modules/@istanbuljs/load-nyc-config/node_modules/js-yaml": {
-      "version": "3.14.2",
-      "resolved": "https://registry.npmjs.org/js-yaml/-/js-yaml-3.14.2.tgz",
-      "integrity": "sha512-PMSmkqxr106Xa156c2M265Z+FTrPl+oxd/rgOQy2tijQeK5TxQ43psO1ZCwhVOSdnn+RzkzlRz/eY4BgJBYVpg==",
+      "version": "3.14.1",
       "dev": true,
       "license": "MIT",
       "dependencies": {
@@ -1684,7 +1678,9 @@
       }
     },
     "node_modules/@next/env": {
-      "version": "16.0.1",
+      "version": "16.0.7",
+      "resolved": "https://registry.npmjs.org/@next/env/-/env-16.0.7.tgz",
+      "integrity": "sha512-gpaNgUh5nftFKRkRQGnVi5dpcYSKGcZZkQffZ172OrG/XkrnS7UBTQ648YY+8ME92cC4IojpI2LqTC8sTDhAaw==",
       "license": "MIT"
     },
     "node_modules/@next/eslint-plugin-next": {
@@ -1722,7 +1718,9 @@
       }
     },
     "node_modules/@next/swc-darwin-arm64": {
-      "version": "16.0.1",
+      "version": "16.0.7",
+      "resolved": "https://registry.npmjs.org/@next/swc-darwin-arm64/-/swc-darwin-arm64-16.0.7.tgz",
+      "integrity": "sha512-LlDtCYOEj/rfSnEn/Idi+j1QKHxY9BJFmxx7108A6D8K0SB+bNgfYQATPk/4LqOl4C0Wo3LACg2ie6s7xqMpJg==",
       "cpu": [
         "arm64"
       ],
@@ -1735,111 +1733,6 @@
         "node": ">= 10"
       }
     },
-    "node_modules/@next/swc-darwin-x64": {
-      "version": "16.0.1",
-      "resolved": "https://registry.npmjs.org/@next/swc-darwin-x64/-/swc-darwin-x64-16.0.1.tgz",
-      "integrity": "sha512-kETZBocRux3xITiZtOtVoVvXyQLB7VBxN7L6EPqgI5paZiUlnsgYv4q8diTNYeHmF9EiehydOBo20lTttCbHAg==",
-      "cpu": [
-        "x64"
-      ],
-      "optional": true,
-      "os": [
-        "darwin"
-      ],
-      "engines": {
-        "node": ">= 10"
-      }
-    },
-    "node_modules/@next/swc-linux-arm64-gnu": {
-      "version": "16.0.1",
-      "resolved": "https://registry.npmjs.org/@next/swc-linux-arm64-gnu/-/swc-linux-arm64-gnu-16.0.1.tgz",
-      "integrity": "sha512-hWg3BtsxQuSKhfe0LunJoqxjO4NEpBmKkE+P2Sroos7yB//OOX3jD5ISP2wv8QdUwtRehMdwYz6VB50mY6hqAg==",
-      "cpu": [
-        "arm64"
-      ],
-      "optional": true,
-      "os": [
-        "linux"
-      ],
-      "engines": {
-        "node": ">= 10"
-      }
-    },
-    "node_modules/@next/swc-linux-arm64-musl": {
-      "version": "16.0.1",
-      "resolved": "https://registry.npmjs.org/@next/swc-linux-arm64-musl/-/swc-linux-arm64-musl-16.0.1.tgz",
-      "integrity": "sha512-UPnOvYg+fjAhP3b1iQStcYPWeBFRLrugEyK/lDKGk7kLNua8t5/DvDbAEFotfV1YfcOY6bru76qN9qnjLoyHCQ==",
-      "cpu": [
-        "arm64"
-      ],
-      "optional": true,
-      "os": [
-        "linux"
-      ],
-      "engines": {
-        "node": ">= 10"
-      }
-    },
-    "node_modules/@next/swc-linux-x64-gnu": {
-      "version": "16.0.1",
-      "resolved": "https://registry.npmjs.org/@next/swc-linux-x64-gnu/-/swc-linux-x64-gnu-16.0.1.tgz",
-      "integrity": "sha512-Et81SdWkcRqAJziIgFtsFyJizHoWne4fzJkvjd6V4wEkWTB4MX6J0uByUb0peiJQ4WeAt6GGmMszE5KrXK6WKg==",
-      "cpu": [
-        "x64"
-      ],
-      "optional": true,
-      "os": [
-        "linux"
-      ],
-      "engines": {
-        "node": ">= 10"
-      }
-    },
-    "node_modules/@next/swc-linux-x64-musl": {
-      "version": "16.0.1",
-      "resolved": "https://registry.npmjs.org/@next/swc-linux-x64-musl/-/swc-linux-x64-musl-16.0.1.tgz",
-      "integrity": "sha512-qBbgYEBRrC1egcG03FZaVfVxrJm8wBl7vr8UFKplnxNRprctdP26xEv9nJ07Ggq4y1adwa0nz2mz83CELY7N6Q==",
-      "cpu": [
-        "x64"
-      ],
-      "optional": true,
-      "os": [
-        "linux"
-      ],
-      "engines": {
-        "node": ">= 10"
-      }
-    },
-    "node_modules/@next/swc-win32-arm64-msvc": {
-      "version": "16.0.1",
-      "resolved": "https://registry.npmjs.org/@next/swc-win32-arm64-msvc/-/swc-win32-arm64-msvc-16.0.1.tgz",
-      "integrity": "sha512-cPuBjYP6I699/RdbHJonb3BiRNEDm5CKEBuJ6SD8k3oLam2fDRMKAvmrli4QMDgT2ixyRJ0+DTkiODbIQhRkeQ==",
-      "cpu": [
-        "arm64"
-      ],
-      "optional": true,
-      "os": [
-        "win32"
-      ],
-      "engines": {
-        "node": ">= 10"
-      }
-    },
-    "node_modules/@next/swc-win32-x64-msvc": {
-      "version": "16.0.1",
-      "resolved": "https://registry.npmjs.org/@next/swc-win32-x64-msvc/-/swc-win32-x64-msvc-16.0.1.tgz",
-      "integrity": "sha512-XeEUJsE4JYtfrXe/LaJn3z1pD19fK0Q6Er8Qoufi+HqvdO4LEPyCxLUt4rxA+4RfYo6S9gMlmzCMU2F+AatFqQ==",
-      "cpu": [
-        "x64"
-      ],
-      "optional": true,
-      "os": [
-        "win32"
-      ],
-      "engines": {
-        "node": ">= 10"
-      }
-    },
     "node_modules/@nodelib/fs.scandir": {
       "version": "2.1.5",
       "license": "MIT",
@@ -5092,6 +4985,8 @@
     },
     "node_modules/@types/react": {
       "version": "19.2.2",
+      "resolved": "https://registry.npmjs.org/@types/react/-/react-19.2.2.tgz",
+      "integrity": "sha512-6mDvHUFSjyT2B2yeNx2nUgMxh9LtOWvkhIU3uePn2I2oyNymUAX1NIsdgviM4CH+JSrp2D2hsMvJOkxY+0wNRA==",
       "license": "MIT",
       "dependencies": {
         "csstype": "^3.0.2"
@@ -5099,6 +4994,8 @@
     },
     "node_modules/@types/react-dom": {
       "version": "19.2.2",
+      "resolved": "https://registry.npmjs.org/@types/react-dom/-/react-dom-19.2.2.tgz",
+      "integrity": "sha512-9KQPoO6mZCi7jcIStSnlOWn2nEF3mNmyr3rIAsGnAbQKYbRLyqmeSc39EVgtxXVia+LMT8j3knZLAZAh+xLmrw==",
       "devOptional": true,
       "license": "MIT",
       "peerDependencies": {
@@ -5608,8 +5505,6 @@
     },
     "node_modules/ajv": {
       "version": "8.17.1",
-      "resolved": "https://registry.npmjs.org/ajv/-/ajv-8.17.1.tgz",
-      "integrity": "sha512-B/gBuNg5SiMTrPkC+A2+cW0RszwxYmn6VYxB/inlBStS5nx6xHIt/ehKRhIMhqusl7a8LjQoZnjCs5vhwxOQ1g==",
       "license": "MIT",
       "dependencies": {
         "fast-deep-equal": "^3.1.3",
@@ -5718,8 +5613,6 @@
     },
     "node_modules/anymatch/node_modules/picomatch": {
       "version": "2.3.1",
-      "resolved": "https://registry.npmjs.org/picomatch/-/picomatch-2.3.1.tgz",
-      "integrity": "sha512-JU3teHTNjmE2VCGFzuY8EXzCDVwEqB2a8fsIvwaStHhAWJEeVd1o1QD80CU6+ZdEXXSLbSsuLwJjkCBWqRQUVA==",
       "license": "MIT",
       "engines": {
         "node": ">=8.6"
@@ -5937,8 +5830,6 @@
     },
     "node_modules/autoprefixer": {
       "version": "10.4.22",
-      "resolved": "https://registry.npmjs.org/autoprefixer/-/autoprefixer-10.4.22.tgz",
-      "integrity": "sha512-ARe0v/t9gO28Bznv6GgqARmVqcWOV3mfgUPn9becPHMiD3o9BwlRgaeccZnwTpZ7Zwqrm+c1sUSsMxIzQzc8Xg==",
       "funding": [
         {
           "type": "opencollective",
@@ -6410,8 +6301,6 @@
     },
     "node_modules/caniuse-lite": {
       "version": "1.0.30001754",
-      "resolved": "https://registry.npmjs.org/caniuse-lite/-/caniuse-lite-1.0.30001754.tgz",
-      "integrity": "sha512-x6OeBXueoAceOmotzx3PO4Zpt4rzpeIFsSr6AAePTZxSkXiYDUmpypEl7e2+8NCd9bD7bXjqyef8CJYPC1jfxg==",
       "funding": [
         {
           "type": "opencollective",
@@ -8135,8 +8024,6 @@
     },
     "node_modules/eslint/node_modules/ajv": {
       "version": "6.12.6",
-      "resolved": "https://registry.npmjs.org/ajv/-/ajv-6.12.6.tgz",
-      "integrity": "sha512-j3fVLgvTo527anyYyJOGTYJbG+vnnQYvE0m5mmkc1TK+nxAppkCLMIL0aZ4dblVCNoGShhm+kzE4ZUykBoMg4g==",
       "dev": true,
       "license": "MIT",
       "dependencies": {
@@ -8152,8 +8039,6 @@
     },
     "node_modules/eslint/node_modules/json-schema-traverse": {
       "version": "0.4.1",
-      "resolved": "https://registry.npmjs.org/json-schema-traverse/-/json-schema-traverse-0.4.1.tgz",
-      "integrity": "sha512-xbbCH5dCYU5T8LcEhhuh7HJ88HXuW3qsI3Y0zOZFKfZEHcpWiHU/Jxzk629Brsab/mMiHQti9wMP+845RPe3Vg==",
       "dev": true,
       "license": "MIT"
     },
@@ -8522,8 +8407,6 @@
     },
     "node_modules/fork-ts-checker-webpack-plugin/node_modules/ajv": {
       "version": "6.12.6",
-      "resolved": "https://registry.npmjs.org/ajv/-/ajv-6.12.6.tgz",
-      "integrity": "sha512-j3fVLgvTo527anyYyJOGTYJbG+vnnQYvE0m5mmkc1TK+nxAppkCLMIL0aZ4dblVCNoGShhm+kzE4ZUykBoMg4g==",
       "dev": true,
       "license": "MIT",
       "dependencies": {
@@ -8547,8 +8430,6 @@
     },
     "node_modules/fork-ts-checker-webpack-plugin/node_modules/json-schema-traverse": {
       "version": "0.4.1",
-      "resolved": "https://registry.npmjs.org/json-schema-traverse/-/json-schema-traverse-0.4.1.tgz",
-      "integrity": "sha512-xbbCH5dCYU5T8LcEhhuh7HJ88HXuW3qsI3Y0zOZFKfZEHcpWiHU/Jxzk629Brsab/mMiHQti9wMP+845RPe3Vg==",
       "dev": true,
       "license": "MIT"
     },
@@ -8621,8 +8502,6 @@
     },
     "node_modules/fraction.js": {
       "version": "5.3.4",
-      "resolved": "https://registry.npmjs.org/fraction.js/-/fraction.js-5.3.4.tgz",
-      "integrity": "sha512-1X1NTtiJphryn/uLQz3whtY6jK3fTqoE3ohKs0tT+Ujr1W59oopxmoEh7Lu5p6vBaPbgoM0bzveAW4Qi5RyWDQ==",
       "license": "MIT",
       "engines": {
         "node": "*"
@@ -10698,8 +10577,6 @@
     },
     "node_modules/jest-util/node_modules/picomatch": {
       "version": "2.3.1",
-      "resolved": "https://registry.npmjs.org/picomatch/-/picomatch-2.3.1.tgz",
-      "integrity": "sha512-JU3teHTNjmE2VCGFzuY8EXzCDVwEqB2a8fsIvwaStHhAWJEeVd1o1QD80CU6+ZdEXXSLbSsuLwJjkCBWqRQUVA==",
       "dev": true,
       "license": "MIT",
       "engines": {
@@ -10790,9 +10667,7 @@
       "license": "MIT"
     },
     "node_modules/js-yaml": {
-      "version": "4.1.1",
-      "resolved": "https://registry.npmjs.org/js-yaml/-/js-yaml-4.1.1.tgz",
-      "integrity": "sha512-qQKT4zQxXl8lLwBtHMWwaTcGfFOZviOJet3Oy/xmGk2gZH677CJM9EvtfdSkgWcATZhj/55JZ0rmy3myCT5lsA==",
+      "version": "4.1.0",
       "dev": true,
       "license": "MIT",
       "dependencies": {
@@ -10875,8 +10750,6 @@
     },
     "node_modules/json-schema-traverse": {
       "version": "1.0.0",
-      "resolved": "https://registry.npmjs.org/json-schema-traverse/-/json-schema-traverse-1.0.0.tgz",
-      "integrity": "sha512-NM8/P9n3XjXhIZn1lLhkFaACTOURQXjWhV4BA/RnOv8xvgqtqpAX9IO4mRQxSx1Rlo4tqzeqb0sOlruaOy3dug==",
       "license": "MIT"
     },
     "node_modules/json-stable-stringify-without-jsonify": {
@@ -12016,8 +11889,6 @@
     },
     "node_modules/micromatch/node_modules/picomatch": {
       "version": "2.3.1",
-      "resolved": "https://registry.npmjs.org/picomatch/-/picomatch-2.3.1.tgz",
-      "integrity": "sha512-JU3teHTNjmE2VCGFzuY8EXzCDVwEqB2a8fsIvwaStHhAWJEeVd1o1QD80CU6+ZdEXXSLbSsuLwJjkCBWqRQUVA==",
       "license": "MIT",
       "engines": {
         "node": ">=8.6"
@@ -12142,10 +12013,12 @@
       "license": "MIT"
     },
     "node_modules/next": {
-      "version": "16.0.1",
+      "version": "16.0.7",
+      "resolved": "https://registry.npmjs.org/next/-/next-16.0.7.tgz",
+      "integrity": "sha512-3mBRJyPxT4LOxAJI6IsXeFtKfiJUbjCLgvXO02fV8Wy/lIhPvP94Fe7dGhUgHXcQy4sSuYwQNcOLhIfOm0rL0A==",
       "license": "MIT",
       "dependencies": {
-        "@next/env": "16.0.1",
+        "@next/env": "16.0.7",
         "@swc/helpers": "0.5.15",
         "caniuse-lite": "^1.0.30001579",
         "postcss": "8.4.31",
@@ -12158,14 +12031,14 @@
         "node": ">=20.9.0"
       },
       "optionalDependencies": {
-        "@next/swc-darwin-arm64": "16.0.1",
-        "@next/swc-darwin-x64": "16.0.1",
-        "@next/swc-linux-arm64-gnu": "16.0.1",
-        "@next/swc-linux-arm64-musl": "16.0.1",
-        "@next/swc-linux-x64-gnu": "16.0.1",
-        "@next/swc-linux-x64-musl": "16.0.1",
-        "@next/swc-win32-arm64-msvc": "16.0.1",
-        "@next/swc-win32-x64-msvc": "16.0.1",
+        "@next/swc-darwin-arm64": "16.0.7",
+        "@next/swc-darwin-x64": "16.0.7",
+        "@next/swc-linux-arm64-gnu": "16.0.7",
+        "@next/swc-linux-arm64-musl": "16.0.7",
+        "@next/swc-linux-x64-gnu": "16.0.7",
+        "@next/swc-linux-x64-musl": "16.0.7",
+        "@next/swc-win32-arm64-msvc": "16.0.7",
+        "@next/swc-win32-x64-msvc": "16.0.7",
         "sharp": "^0.34.4"
       },
       "peerDependencies": {
@@ -15259,8 +15132,6 @@
     },
     "node_modules/picomatch": {
       "version": "4.0.3",
-      "resolved": "https://registry.npmjs.org/picomatch/-/picomatch-4.0.3.tgz",
-      "integrity": "sha512-5gTmgEY/sqK6gFXLIsQNH19lWb4ebPDLA4SdLP7dsWkIXHWlG66oPuVvXSGFPppYZz8ZDZq0dYYrbHfBCVUb1Q==",
       "license": "MIT",
       "engines": {
         "node": ">=12"
@@ -15872,7 +15743,9 @@
       }
     },
     "node_modules/react": {
-      "version": "19.2.0",
+      "version": "19.2.1",
+      "resolved": "https://registry.npmjs.org/react/-/react-19.2.1.tgz",
+      "integrity": "sha512-DGrYcCWK7tvYMnWh79yrPHt+vdx9tY+1gPZa7nJQtO/p8bLTDaHp4dzwEhQB7pZ4Xe3ok4XKuEPrVuc+wlpkmw==",
       "license": "MIT",
       "engines": {
         "node": ">=0.10.0"
@@ -15932,13 +15805,15 @@
       }
     },
     "node_modules/react-dom": {
-      "version": "19.2.0",
+      "version": "19.2.1",
+      "resolved": "https://registry.npmjs.org/react-dom/-/react-dom-19.2.1.tgz",
+      "integrity": "sha512-ibrK8llX2a4eOskq1mXKu/TGZj9qzomO+sNfO98M6d9zIPOEhlBkMkBUBLd1vgS0gQsLDBzA+8jJBVXDnfHmJg==",
       "license": "MIT",
       "dependencies": {
         "scheduler": "^0.27.0"
       },
       "peerDependencies": {
-        "react": "^19.2.0"
+        "react": "^19.2.1"
       }
     },
     "node_modules/react-dropzone": {
@@ -16139,8 +16014,6 @@
     },
     "node_modules/readdirp/node_modules/picomatch": {
       "version": "2.3.1",
-      "resolved": "https://registry.npmjs.org/picomatch/-/picomatch-2.3.1.tgz",
-      "integrity": "sha512-JU3teHTNjmE2VCGFzuY8EXzCDVwEqB2a8fsIvwaStHhAWJEeVd1o1QD80CU6+ZdEXXSLbSsuLwJjkCBWqRQUVA==",
       "license": "MIT",
       "engines": {
         "node": ">=8.6"
@@ -18249,8 +18122,6 @@
     },
     "node_modules/uri-js": {
       "version": "4.4.1",
-      "resolved": "https://registry.npmjs.org/uri-js/-/uri-js-4.4.1.tgz",
-      "integrity": "sha512-7rKUyy33Q1yc98pQ1DAmLtwX109F7TIfWlW1Ydo8Wl1ii1SeHieeh0HHfPeL2fMXK6z0s8ecKs9frCuLJvndBg==",
       "dev": true,
       "license": "BSD-2-Clause",
       "dependencies": {
@@ -18987,8 +18858,6 @@
     },
     "node_modules/zustand": {
       "version": "5.0.8",
-      "resolved": "https://registry.npmjs.org/zustand/-/zustand-5.0.8.tgz",
-      "integrity": "sha512-gyPKpIaxY9XcO2vSMrLbiER7QMAMGOQZVRdJ6Zi782jkbzZygq5GI9nG8g+sMgitRtndwaBSl7uiqC49o1SSiw==",
       "license": "MIT",
       "engines": {
         "node": ">=12.20.0"

web/package.json

@@ -71,15 +71,15 @@
     "lowlight": "^3.3.0",
     "lucide-react": "^0.454.0",
     "mdast-util-find-and-replace": "^3.0.1",
-    "next": "^16.0.1",
+    "next": "^16.0.7",
     "next-themes": "^0.4.4",
     "npm": "^10.8.0",
     "postcss": "^8.5.6",
     "posthog-js": "^1.176.0",
-    "react": "^19.2.0",
+    "react": "^19.2.1",
     "react-datepicker": "^7.6.0",
     "react-day-picker": "^9.11.1",
-    "react-dom": "^19.2.0",
+    "react-dom": "^19.2.1",
     "react-dropzone": "^14.2.3",
     "react-icons": "^4.8.0",
     "react-loader-spinner": "^8.0.0",
@@ -116,8 +116,8 @@
     "@types/js-cookie": "^3.0.6",
     "@types/lodash": "^4.17.20",
     "@types/node": "18.15.11",
-    "@types/react": "19.2.2",
-    "@types/react-dom": "19.2.2",
+    "@types/react": "^19.2.2",
+    "@types/react-dom": "^19.2.2",
     "@types/uuid": "^9.0.8",
     "babel-plugin-react-compiler": "^1.0.0",
     "chromatic": "^11.25.2",
@@ -136,7 +136,7 @@
   },
   "overrides": {
     "react-is": "^19.0.0-rc-69d4b800-20241021",
-    "@types/react": "19.2.2",
-    "@types/react-dom": "19.2.2"
+    "@types/react": "^19.2.2",
+    "@types/react-dom": "^19.2.2"
   }
 }

web/src/app/admin/actions/ActionEditor.tsx

@@ -20,6 +20,7 @@ import Link from "next/link";
 import Separator from "@/refresh-components/Separator";
 import Checkbox from "@/refresh-components/inputs/Checkbox";
 import { useAuthType } from "@/lib/hooks";
+import { AuthType, DOCS_ADMINS_PATH } from "@/lib/constants";
 import { InfoIcon } from "@/components/icons/icons";
 import {
   Table,
@@ -169,7 +170,7 @@ function ActionForm({
         {/* Documentation Link */}
         <div className="rounded-md border border-border bg-background-50 p-4">
           <Link
-            href="https://docs.onyx.app/admin/actions/overview"
+            href={`${DOCS_ADMINS_PATH}/actions/overview`}
             className="flex items-center gap-2 group"
             target="_blank"
             rel="noopener noreferrer"

web/src/app/admin/bots/SlackTokensForm.tsx

@@ -7,6 +7,7 @@ import { createSlackBot, updateSlackBot } from "./new/lib";
 import Button from "@/refresh-components/buttons/Button";
 import Separator from "@/refresh-components/Separator";
 import { useEffect } from "react";
+import { DOCS_ADMINS_PATH } from "@/lib/constants";
 
 export const SlackTokensForm = ({
   isUpdate,
@@ -102,7 +103,7 @@ export const SlackTokensForm = ({
               Please refer to our{" "}
               <a
                 className="text-blue-500 hover:underline"
-                href="https://docs.onyx.app/admin/getting_started/slack_bot_setup"
+                href={`${DOCS_ADMINS_PATH}/getting_started/slack_bot_setup`}
                 target="_blank"
                 rel="noopener noreferrer"
               >

web/src/app/admin/bots/page.tsx

@@ -9,6 +9,7 @@ import { SlackBotTable } from "./SlackBotTable";
 import { useSlackBots } from "./[bot-id]/hooks";
 import { ValidSources } from "@/lib/types";
 import CreateButton from "@/refresh-components/buttons/CreateButton";
+import { DOCS_ADMINS_PATH } from "@/lib/constants";
 
 const Main = () => {
   const {
@@ -60,7 +61,7 @@ const Main = () => {
         Follow the{" "}
         <a
           className="text-blue-500 hover:underline"
-          href="https://docs.onyx.app/admin/getting_started/slack_bot_setup"
+          href={`${DOCS_ADMINS_PATH}/getting_started/slack_bot_setup`}
           target="_blank"
           rel="noopener noreferrer"
         >

web/src/app/admin/connectors/[connector]/pages/gdrive/Credential.tsx

@@ -5,7 +5,10 @@ import * as Yup from "yup";
 import { useRouter } from "next/navigation";
 import { adminDeleteCredential } from "@/lib/credential";
 import { setupGoogleDriveOAuth } from "@/lib/googleDrive";
-import { GOOGLE_DRIVE_AUTH_IS_ADMIN_COOKIE_NAME } from "@/lib/constants";
+import {
+  DOCS_ADMINS_PATH,
+  GOOGLE_DRIVE_AUTH_IS_ADMIN_COOKIE_NAME,
+} from "@/lib/constants";
 import Cookies from "js-cookie";
 import { TextFormField, SectionHeader } from "@/components/Field";
 import { Form, Formik } from "formik";
@@ -300,7 +303,7 @@ export const DriveJsonUploadSection = ({
         <a
           className="text-primary hover:text-primary/80 flex items-center gap-1 text-sm"
           target="_blank"
-          href="https://docs.onyx.app/admin/connectors/official/google_drive/overview"
+          href={`${DOCS_ADMINS_PATH}/connectors/official/google_drive/overview`}
           rel="noreferrer"
         >
           <FiLink className="h-3 w-3" />

web/src/app/admin/connectors/[connector]/pages/gmail/Credential.tsx

@@ -6,7 +6,10 @@ import * as Yup from "yup";
 import { useRouter } from "next/navigation";
 import { adminDeleteCredential } from "@/lib/credential";
 import { setupGmailOAuth } from "@/lib/gmail";
-import { GMAIL_AUTH_IS_ADMIN_COOKIE_NAME } from "@/lib/constants";
+import {
+  DOCS_ADMINS_PATH,
+  GMAIL_AUTH_IS_ADMIN_COOKIE_NAME,
+} from "@/lib/constants";
 import Cookies from "js-cookie";
 import { TextFormField, SectionHeader } from "@/components/Field";
 import { Form, Formik } from "formik";
@@ -298,7 +301,7 @@ export const GmailJsonUploadSection = ({
         <a
           className="text-primary hover:text-primary/80 flex items-center gap-1 text-sm"
           target="_blank"
-          href="https://docs.onyx.app/admin/connectors/official/gmail/overview"
+          href={`${DOCS_ADMINS_PATH}/connectors/official/gmail/overview`}
           rel="noreferrer"
         >
           <FiLink className="h-3 w-3" />

web/src/app/chat/hooks/useChatController.ts

@@ -1025,6 +1025,8 @@ export function useChatController({
 
   // fetch # of allowed document tokens for the selected Persona
   useEffect(() => {
+    if (!liveAssistant?.id) return; // avoid calling with undefined persona id
+
     async function fetchMaxTokens() {
       const response = await fetch(
         `/api/chat/max-selected-document-tokens?persona_id=${liveAssistant?.id}`

web/src/app/chat/message/messageComponents/AIMessage.tsx

@@ -101,7 +101,7 @@ export default function AIMessage({
       // Toggle logic
       if (currentFeedback === clickedFeedback) {
         // Clicking same button - remove feedback
-        await handleFeedbackChange(nodeId, null);
+        await handleFeedbackChange(messageId, null);
       }
 
       // Clicking like (will automatically clear dislike if it was active).
@@ -113,12 +113,12 @@ export default function AIMessage({
           // Open modal for positive feedback
           setFeedbackModalProps({
             feedbackType: "like",
-            messageId: nodeId,
+            messageId,
           });
           modal.toggle(true);
         } else {
           // No modal needed - just submit like (this replaces any existing feedback)
-          await handleFeedbackChange(nodeId, "like");
+          await handleFeedbackChange(messageId, "like");
         }
       }
 

web/src/app/ee/admin/whitelabeling/WhitelabelingForm.tsx

@@ -18,6 +18,7 @@ import { ImageUpload } from "./ImageUpload";
 import { AdvancedOptionsToggle } from "@/components/AdvancedOptionsToggle";
 import Link from "next/link";
 import Separator from "@/refresh-components/Separator";
+import { DOCS_ADMINS_PATH } from "@/lib/constants";
 
 export function WhitelabelingForm() {
   const router = useRouter();
@@ -201,9 +202,7 @@ export function WhitelabelingForm() {
                 <Text>
                   Read{" "}
                   <Link
-                    href={
-                      "https://docs.onyx.app/admin/advanced_configs/white_labeling"
-                    }
+                    href={`${DOCS_ADMINS_PATH}/advanced_configs/white_labeling`}
                     className="text-link cursor-pointer"
                   >
                     the docs

web/src/components/FederatedConnectorSelector.tsx

@@ -3,24 +3,14 @@ import {
   FederatedConnectorDetail,
   FederatedConnectorConfig,
   federatedSourceToRegularSource,
-  ValidSources,
 } from "@/lib/types";
 import { SourceIcon } from "@/components/SourceIcon";
 import SvgX from "@/icons/x";
-import SvgSettings from "@/icons/settings";
 import { Label } from "@/components/ui/label";
 import { ErrorMessage } from "formik";
 import Text from "@/refresh-components/texts/Text";
 import IconButton from "@/refresh-components/buttons/IconButton";
-import { Input } from "@/components/ui/input";
 import InputTypeIn from "@/refresh-components/inputs/InputTypeIn";
-import {
-  Dialog,
-  DialogContent,
-  DialogHeader,
-  DialogTitle,
-} from "@/components/ui/dialog";
-import Button from "@/refresh-components/buttons/Button";
 
 interface FederatedConnectorSelectorProps {
   name: string;
@@ -33,194 +23,6 @@ interface FederatedConnectorSelectorProps {
   showError?: boolean;
 }
 
-interface EntityConfigDialogProps {
-  connectorId: number;
-  connectorName: string;
-  connectorSource: ValidSources | null;
-  currentEntities: Record<string, any>;
-  onSave: (entities: Record<string, any>) => void;
-  onClose: () => void;
-  isOpen: boolean;
-}
-
-const EntityConfigDialog = ({
-  connectorId,
-  connectorName,
-  connectorSource,
-  currentEntities,
-  onSave,
-  onClose,
-  isOpen,
-}: EntityConfigDialogProps) => {
-  const [entities, setEntities] =
-    useState<Record<string, any>>(currentEntities);
-  const [entitySchema, setEntitySchema] = useState<Record<string, any> | null>(
-    null
-  );
-  const [isLoading, setIsLoading] = useState(false);
-  const [error, setError] = useState<string | null>(null);
-
-  useEffect(() => {
-    if (isOpen) {
-      setEntities(currentEntities || {});
-    }
-  }, [currentEntities, isOpen]);
-
-  useEffect(() => {
-    if (isOpen && connectorId) {
-      const fetchEntitySchema = async () => {
-        setIsLoading(true);
-        setError(null);
-        try {
-          const response = await fetch(
-            `/api/federated/${connectorId}/entities`
-          );
-          if (!response.ok) {
-            throw new Error(
-              `Failed to fetch entity schema: ${response.statusText}`
-            );
-          }
-          const data = await response.json();
-          setEntitySchema(data.entities);
-        } catch (err) {
-          setError(
-            err instanceof Error ? err.message : "Failed to load entity schema"
-          );
-        } finally {
-          setIsLoading(false);
-        }
-      };
-      fetchEntitySchema();
-    }
-  }, [isOpen, connectorId]);
-
-  const handleSave = () => {
-    onSave(entities);
-    onClose();
-  };
-
-  const handleEntityChange = (key: string, value: any) => {
-    setEntities((prev) => ({
-      ...prev,
-      [key]: value,
-    }));
-  };
-
-  if (!connectorSource) {
-    return null;
-  }
-
-  return (
-    <Dialog open={isOpen} onOpenChange={onClose}>
-      <DialogContent className="max-w-md">
-        <DialogHeader>
-          <DialogTitle className="flex items-center gap-2">
-            <SourceIcon
-              sourceType={federatedSourceToRegularSource(connectorSource)}
-              iconSize={20}
-            />
-            Configure {connectorName}
-          </DialogTitle>
-        </DialogHeader>
-
-        <div className="space-y-4">
-          {isLoading && (
-            <div className="text-center py-4">
-              <div className="animate-spin h-6 w-6 border-2 border-blue-500 border-t-transparent rounded-full mx-auto mb-2"></div>
-              <p className="text-sm text-muted-foreground">
-                Loading configuration...
-              </p>
-            </div>
-          )}
-
-          {error && (
-            <div className="text-red-500 text-sm p-3 bg-red-50 rounded-md">
-              {error}
-            </div>
-          )}
-
-          {entitySchema && !isLoading && (
-            <div className="space-y-3">
-              <p className="text-sm text-muted-foreground">
-                Configure which entities to include from this connector:
-              </p>
-
-              {Object.entries(entitySchema).map(
-                ([key, field]: [string, any]) => (
-                  <div key={key} className="space-y-2">
-                    <Label className="text-sm font-medium">
-                      {field.description || key}
-                      {field.required && (
-                        <span className="text-red-500 ml-1">*</span>
-                      )}
-                    </Label>
-
-                    {field.type === "list" ? (
-                      <div className="space-y-2">
-                        <Input
-                          type="text"
-                          placeholder={
-                            field.example || `Enter ${key} (comma-separated)`
-                          }
-                          value={
-                            Array.isArray(entities[key])
-                              ? entities[key].join(", ")
-                              : ""
-                          }
-                          onChange={(e) => {
-                            const value = e.target.value;
-                            const list = value
-                              ? value
-                                  .split(",")
-                                  .map((item) => item.trim())
-                                  .filter(Boolean)
-                              : [];
-                            handleEntityChange(key, list);
-                          }}
-                        />
-                        <p className="text-xs text-muted-foreground">
-                          {field.description && field.description !== key
-                            ? field.description
-                            : `Enter ${key} separated by commas`}
-                        </p>
-                      </div>
-                    ) : (
-                      <div className="space-y-2">
-                        <Input
-                          type="text"
-                          placeholder={field.example || `Enter ${key}`}
-                          value={entities[key] || ""}
-                          onChange={(e) =>
-                            handleEntityChange(key, e.target.value)
-                          }
-                        />
-                        {field.description && field.description !== key && (
-                          <p className="text-xs text-muted-foreground">
-                            {field.description}
-                          </p>
-                        )}
-                      </div>
-                    )}
-                  </div>
-                )
-              )}
-            </div>
-          )}
-
-          <div className="flex justify-end gap-2 pt-4">
-            <Button secondary onClick={onClose}>
-              Cancel
-            </Button>
-            <Button onClick={handleSave} disabled={isLoading}>
-              Save Configuration
-            </Button>
-          </div>
-        </div>
-      </DialogContent>
-    </Dialog>
-  );
-};
-
 export const FederatedConnectorSelector = ({
   name,
   label,
@@ -233,19 +35,6 @@ export const FederatedConnectorSelector = ({
 }: FederatedConnectorSelectorProps) => {
   const [open, setOpen] = useState(false);
   const [searchQuery, setSearchQuery] = useState("");
-  const [configDialogState, setConfigDialogState] = useState<{
-    isOpen: boolean;
-    connectorId: number | null;
-    connectorName: string;
-    connectorSource: ValidSources | null;
-    currentEntities: Record<string, any>;
-  }>({
-    isOpen: false,
-    connectorId: null,
-    connectorName: "",
-    connectorSource: null,
-    currentEntities: {},
-  });
   const dropdownRef = useRef<HTMLDivElement>(null);
   const inputRef = useRef<HTMLInputElement>(null);
 
@@ -307,36 +96,6 @@ export const FederatedConnectorSelector = ({
     );
   };
 
-  const openConfigDialog = (connectorId: number) => {
-    const connector = federatedConnectors.find((c) => c.id === connectorId);
-    const config = selectedConfigs.find(
-      (c) => c.federated_connector_id === connectorId
-    );
-
-    if (connector) {
-      setConfigDialogState({
-        isOpen: true,
-        connectorId,
-        connectorName: connector.name,
-        connectorSource: connector.source,
-        currentEntities: config?.entities || {},
-      });
-    }
-  };
-
-  const saveEntityConfig = (entities: Record<string, any>) => {
-    const updatedConfigs = selectedConfigs.map((config) => {
-      if (config.federated_connector_id === configDialogState.connectorId) {
-        return {
-          ...config,
-          entities,
-        };
-      }
-      return config;
-    });
-    onChange(updatedConfigs);
-  };
-
   useEffect(() => {
     const handleClickOutside = (event: MouseEvent) => {
       if (
@@ -472,14 +231,6 @@ export const FederatedConnectorSelector = ({
                     )}
                   </div>
                   <div className="flex items-center ml-2 gap-1">
-                    <IconButton
-                      internal
-                      type="button"
-                      tooltip="Configure entities"
-                      aria-label="Configure entities"
-                      onClick={() => openConfigDialog(connector.id)}
-                      icon={SvgSettings}
-                    />
                     <IconButton
                       internal
                       type="button"
@@ -500,18 +251,6 @@ export const FederatedConnectorSelector = ({
         </div>
       )}
 
-      <EntityConfigDialog
-        connectorId={configDialogState.connectorId!}
-        connectorName={configDialogState.connectorName}
-        connectorSource={configDialogState.connectorSource}
-        currentEntities={configDialogState.currentEntities}
-        onSave={saveEntityConfig}
-        onClose={() =>
-          setConfigDialogState((prev) => ({ ...prev, isOpen: false }))
-        }
-        isOpen={configDialogState.isOpen}
-      />
-
       {showError && (
         <ErrorMessage
           name={name}

web/src/components/admin/assistants/FormSections.tsx

@@ -5,6 +5,7 @@ import { FastField, useFormikContext } from "formik";
 import { TextFormField } from "@/components/Field";
 import { FiChevronRight, FiChevronDown } from "react-icons/fi";
 import Checkbox from "@/refresh-components/inputs/Checkbox";
+import { DOCS_ADMINS_PATH } from "@/lib/constants";
 
 const MAX_DESCRIPTION_LENGTH = 600;
 import { useState, useEffect } from "react";
@@ -82,7 +83,7 @@ export const TaskPromptField = memo(function TaskPromptField() {
             form.setFieldValue("task_prompt", e.target.value);
           }}
           explanationText="Learn about prompting in our docs!"
-          explanationLink="https://docs.onyx.app/admin/agents/overview"
+          explanationLink={`${DOCS_ADMINS_PATH}/agents/overview`}
           className="[&_textarea]:placeholder:text-text-muted/50"
         />
       )}

web/src/components/embedding/interfaces.tsx

@@ -12,6 +12,7 @@ import {
   VoyageIconSVG,
 } from "@/components/icons/icons";
 import { SwitchoverType } from "@/app/admin/embeddings/interfaces";
+import { DOCS_ADMINS_PATH } from "@/lib/constants";
 
 export enum EmbeddingProvider {
   OPENAI = "openai",
@@ -180,7 +181,7 @@ export const AVAILABLE_CLOUD_PROVIDERS: CloudEmbeddingProvider[] = [
     provider_type: EmbeddingProvider.COHERE,
     website: "https://cohere.ai",
     icon: CohereIcon,
-    docsLink: "https://docs.onyx.app/admin/advanced_configs/search_configs",
+    docsLink: `${DOCS_ADMINS_PATH}/advanced_configs/search_configs`,
     description:
       "AI company specializing in NLP models for various text-based tasks",
     apiLink: "https://dashboard.cohere.ai/api-keys",
@@ -222,7 +223,7 @@ export const AVAILABLE_CLOUD_PROVIDERS: CloudEmbeddingProvider[] = [
     icon: OpenAIISVG,
     description: "AI industry leader known for ChatGPT and DALL-E",
     apiLink: "https://platform.openai.com/api-keys",
-    docsLink: "https://docs.onyx.app/admin/advanced_configs/search_configs",
+    docsLink: `${DOCS_ADMINS_PATH}/advanced_configs/search_configs`,
     costslink: "https://openai.com/pricing",
     embedding_models: [
       {
@@ -260,7 +261,7 @@ export const AVAILABLE_CLOUD_PROVIDERS: CloudEmbeddingProvider[] = [
     provider_type: EmbeddingProvider.GOOGLE,
     website: "https://ai.google",
     icon: GoogleIcon,
-    docsLink: "https://docs.onyx.app/admin/advanced_configs/search_configs",
+    docsLink: `${DOCS_ADMINS_PATH}/advanced_configs/search_configs`,
     description:
       "Offers a wide range of AI services including language and vision models",
     apiLink: "https://console.cloud.google.com/apis/credentials",
@@ -299,7 +300,7 @@ export const AVAILABLE_CLOUD_PROVIDERS: CloudEmbeddingProvider[] = [
     website: "https://www.voyageai.com",
     icon: VoyageIconSVG,
     description: "Advanced NLP research startup born from Stanford AI Labs",
-    docsLink: "https://docs.onyx.app/admin/advanced_configs/search_configs",
+    docsLink: `${DOCS_ADMINS_PATH}/advanced_configs/search_configs`,
     apiLink: "https://www.voyageai.com/dashboard",
     costslink: "https://www.voyageai.com/pricing",
     embedding_models: [

web/src/components/errorPages/ErrorPage.tsx

@@ -1,6 +1,7 @@
 import ErrorPageLayout from "@/components/errorPages/ErrorPageLayout";
 import Text from "@/refresh-components/texts/Text";
 import SvgAlertCircle from "@/icons/alert-circle";
+import { DOCS_BASE_URL } from "@/lib/constants";
 
 export default function Error() {
   return (
@@ -19,7 +20,7 @@ export default function Error() {
         If you're an admin, please review our{" "}
         <a
           className="text-action-link-05"
-          href="https://docs.onyx.app/?utm_source=app&utm_medium=error_page&utm_campaign=config_error"
+          href={`${DOCS_BASE_URL}?utm_source=app&utm_medium=error_page&utm_campaign=config_error`}
           target="_blank"
           rel="noopener noreferrer"
         >

web/src/lib/connectors/connectors.tsx

@@ -2,6 +2,7 @@ import * as Yup from "yup";
 import { ConfigurableSources, ValidInputTypes, ValidSources } from "../types";
 import { AccessTypeGroupSelectorFormType } from "@/components/admin/connectors/AccessTypeGroupSelector";
 import { Credential } from "@/lib/connectors/credentials"; // Import Credential type
+import { DOCS_ADMINS_PATH } from "@/lib/constants";
 
 export function isLoadState(connector_name: string): boolean {
   // TODO: centralize connector metadata like this somewhere instead of hardcoding it here
@@ -791,7 +792,7 @@ export const connectorConfigs: Record<
                   "\n    }" +
                   "\n  }" +
                   "\n}" +
-                  "\n\n[See our docs](https://docs.onyx.app/admin/connectors/official/salesforce) for more details.",
+                  `\n\n[See our docs](${DOCS_ADMINS_PATH}/connectors/official/salesforce) for more details.`,
               },
             ],
           },
@@ -813,6 +814,7 @@ export const connectorConfigs: Record<
         description: `• If no sites are specified, all sites in your organization will be indexed (Sites.Read.All permission required).
 • Specifying 'https://onyxai.sharepoint.com/sites/support' for example only indexes this site.
 • Specifying 'https://onyxai.sharepoint.com/sites/support/subfolder' for example only indexes this folder.
+• Specifying sites currently works for SharePoint instances using English, Spanish, or German. Contact the Onyx team if you need another language supported.
 `,
       },
     ],

web/src/lib/constants.ts

@@ -9,6 +9,11 @@ export type AuthType =
 export const HOST_URL = process.env.WEB_DOMAIN || "http://127.0.0.1:3000";
 
 export const INTERNAL_URL = process.env.INTERNAL_URL || "http://127.0.0.1:8080";
+
+// Documentation URLs
+export const DOCS_BASE_URL = "https://docs.onyx.app";
+export const DOCS_ADMINS_PATH = `${DOCS_BASE_URL}/admins`;
+
 export const MCP_INTERNAL_URL =
   process.env.MCP_INTERNAL_URL || "http://127.0.0.1:8090";
 
@@ -93,19 +98,8 @@ export const NEXT_PUBLIC_INCLUDE_ERROR_POPUP_SUPPORT_LINK =
 export const NEXT_PUBLIC_STRIPE_PUBLISHABLE_KEY =
   process.env.NEXT_PUBLIC_STRIPE_PUBLISHABLE_KEY;
 
-// Add support for custom URL protocols in markdown links
-export const ALLOWED_URL_PROTOCOLS = [
-  "http:",
-  "https:",
-  "mailto:",
-  "tel:",
-  "slack:",
-  "vscode:",
-  "file:",
-  "sms:",
-  "spotify:",
-  "zoommtg:",
-];
+// Restrict markdown links to safe protocols
+export const ALLOWED_URL_PROTOCOLS = ["http:", "https:", "mailto:"] as const;
 
 export const MAX_CHARACTERS_PERSONA_DESCRIPTION = 5000000;
 export const MAX_STARTER_MESSAGES = 4;

web/src/lib/sources.ts

@@ -52,6 +52,10 @@ import {
 import { ValidSources } from "./types";
 import { SourceCategory, SourceMetadata } from "./search/interfaces";
 import { Persona } from "@/app/admin/assistants/interfaces";
+import SvgGlobe from "@/icons/globe";
+import React from "react";
+import SvgFileText from "@/icons/file-text";
+import { DOCS_ADMINS_PATH } from "./constants";
 
 interface PartialSourceMetadata {
   icon: React.FC<{ size?: number; className?: string }>;
@@ -76,7 +80,7 @@ const slackMetadata = {
   displayName: "Slack",
   category: SourceCategory.Messaging,
   isPopular: true,
-  docs: "https://docs.onyx.app/admin/connectors/official/slack",
+  docs: `${DOCS_ADMINS_PATH}/connectors/official/slack`,
   oauthSupported: true,
   federated: true,
   federatedTooltip:
@@ -91,7 +95,7 @@ export const SOURCE_METADATA_MAP: SourceMap = {
     icon: ConfluenceIcon,
     displayName: "Confluence",
     category: SourceCategory.Wiki,
-    docs: "https://docs.onyx.app/admin/connectors/official/confluence",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/confluence`,
     oauthSupported: true,
     isPopular: true,
   },
@@ -99,80 +103,80 @@ export const SOURCE_METADATA_MAP: SourceMap = {
     icon: SharepointIcon,
     displayName: "Sharepoint",
     category: SourceCategory.Wiki,
-    docs: "https://docs.onyx.app/admin/connectors/official/sharepoint",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/sharepoint`,
     isPopular: true,
   },
   notion: {
     icon: NotionIcon,
     displayName: "Notion",
     category: SourceCategory.Wiki,
-    docs: "https://docs.onyx.app/admin/connectors/official/notion",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/notion`,
   },
   bookstack: {
     icon: BookstackIcon,
     displayName: "BookStack",
     category: SourceCategory.Wiki,
-    docs: "https://docs.onyx.app/admin/connectors/official/bookstack",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/bookstack`,
   },
   document360: {
     icon: Document360Icon,
     displayName: "Document360",
     category: SourceCategory.Wiki,
-    docs: "https://docs.onyx.app/admin/connectors/official/document360",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/document360`,
   },
   discourse: {
     icon: DiscourseIcon,
     displayName: "Discourse",
     category: SourceCategory.Wiki,
-    docs: "https://docs.onyx.app/admin/connectors/official/discourse",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/discourse`,
   },
   gitbook: {
     icon: GitbookIcon,
     displayName: "GitBook",
     category: SourceCategory.Wiki,
-    docs: "https://docs.onyx.app/admin/connectors/official/gitbook",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/gitbook`,
   },
   slab: {
     icon: SlabIcon,
     displayName: "Slab",
     category: SourceCategory.Wiki,
-    docs: "https://docs.onyx.app/admin/connectors/official/slab",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/slab`,
   },
   outline: {
     icon: OutlineIcon,
     displayName: "Outline",
     category: SourceCategory.Wiki,
-    docs: "https://docs.onyx.app/admin/connectors/official/outline",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/outline`,
   },
   google_sites: {
     icon: GoogleSitesIcon,
     displayName: "Google Sites",
     category: SourceCategory.Wiki,
-    docs: "https://docs.onyx.app/admin/connectors/official/google_sites",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/google_sites`,
   },
   guru: {
     icon: GuruIcon,
     displayName: "Guru",
     category: SourceCategory.Wiki,
-    docs: "https://docs.onyx.app/admin/connectors/official/guru",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/guru`,
   },
   mediawiki: {
     icon: MediaWikiIcon,
     displayName: "MediaWiki",
     category: SourceCategory.Wiki,
-    docs: "https://docs.onyx.app/admin/connectors/official/mediawiki",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/mediawiki`,
   },
   axero: {
     icon: AxeroIcon,
     displayName: "Axero",
     category: SourceCategory.Wiki,
-    docs: "https://docs.onyx.app/admin/connectors/official/axero",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/axero`,
   },
   wikipedia: {
     icon: WikipediaIcon,
     displayName: "Wikipedia",
     category: SourceCategory.Wiki,
-    docs: "https://docs.onyx.app/admin/connectors/official/wikipedia",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/wikipedia`,
   },
 
   // Cloud Storage
@@ -180,7 +184,7 @@ export const SOURCE_METADATA_MAP: SourceMap = {
     icon: GoogleDriveIcon,
     displayName: "Google Drive",
     category: SourceCategory.Storage,
-    docs: "https://docs.onyx.app/admin/connectors/official/google_drive/overview",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/google_drive/overview`,
     oauthSupported: true,
     isPopular: true,
   },
@@ -188,37 +192,37 @@ export const SOURCE_METADATA_MAP: SourceMap = {
     icon: DropboxIcon,
     displayName: "Dropbox",
     category: SourceCategory.Storage,
-    docs: "https://docs.onyx.app/admin/connectors/official/dropbox",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/dropbox`,
   },
   s3: {
     icon: S3Icon,
     displayName: "S3",
     category: SourceCategory.Storage,
-    docs: "https://docs.onyx.app/admin/connectors/official/s3",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/s3`,
   },
   google_cloud_storage: {
     icon: GoogleStorageIcon,
     displayName: "Google Storage",
     category: SourceCategory.Storage,
-    docs: "https://docs.onyx.app/admin/connectors/official/google_storage",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/google_storage`,
   },
   egnyte: {
     icon: EgnyteIcon,
     displayName: "Egnyte",
     category: SourceCategory.Storage,
-    docs: "https://docs.onyx.app/admin/connectors/official/egnyte",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/egnyte`,
   },
   oci_storage: {
     icon: OCIStorageIcon,
     displayName: "Oracle Storage",
     category: SourceCategory.Storage,
-    docs: "https://docs.onyx.app/admin/connectors/official/oci_storage",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/oci_storage`,
   },
   r2: {
     icon: R2Icon,
     displayName: "R2",
     category: SourceCategory.Storage,
-    docs: "https://docs.onyx.app/admin/connectors/official/r2",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/r2`,
   },
 
   // Ticketing & Task Management
@@ -226,51 +230,51 @@ export const SOURCE_METADATA_MAP: SourceMap = {
     icon: JiraIcon,
     displayName: "Jira",
     category: SourceCategory.TicketingAndTaskManagement,
-    docs: "https://docs.onyx.app/admin/connectors/official/jira",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/jira`,
     isPopular: true,
   },
   zendesk: {
     icon: ZendeskIcon,
     displayName: "Zendesk",
     category: SourceCategory.TicketingAndTaskManagement,
-    docs: "https://docs.onyx.app/admin/connectors/official/zendesk",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/zendesk`,
     isPopular: true,
   },
   airtable: {
     icon: AirtableIcon,
     displayName: "Airtable",
     category: SourceCategory.TicketingAndTaskManagement,
-    docs: "https://docs.onyx.app/admin/connectors/official/airtable",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/airtable`,
   },
   linear: {
     icon: LinearIcon,
     displayName: "Linear",
     category: SourceCategory.TicketingAndTaskManagement,
-    docs: "https://docs.onyx.app/admin/connectors/official/linear",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/linear`,
   },
   freshdesk: {
     icon: FreshdeskIcon,
     displayName: "Freshdesk",
     category: SourceCategory.TicketingAndTaskManagement,
-    docs: "https://docs.onyx.app/admin/connectors/official/freshdesk",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/freshdesk`,
   },
   asana: {
     icon: AsanaIcon,
     displayName: "Asana",
     category: SourceCategory.TicketingAndTaskManagement,
-    docs: "https://docs.onyx.app/admin/connectors/official/asana",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/asana`,
   },
   clickup: {
     icon: ClickupIcon,
     displayName: "Clickup",
     category: SourceCategory.TicketingAndTaskManagement,
-    docs: "https://docs.onyx.app/admin/connectors/official/clickup",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/clickup`,
   },
   productboard: {
     icon: ProductboardIcon,
     displayName: "Productboard",
     category: SourceCategory.TicketingAndTaskManagement,
-    docs: "https://docs.onyx.app/admin/connectors/official/productboard",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/productboard`,
   },
   testrail: {
     icon: TestRailIcon,
@@ -285,13 +289,13 @@ export const SOURCE_METADATA_MAP: SourceMap = {
     icon: TeamsIcon,
     displayName: "Teams",
     category: SourceCategory.Messaging,
-    docs: "https://docs.onyx.app/admin/connectors/official/teams",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/teams`,
   },
   gmail: {
     icon: GmailIcon,
     displayName: "Gmail",
     category: SourceCategory.Messaging,
-    docs: "https://docs.onyx.app/admin/connectors/official/gmail/overview",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/gmail/overview`,
   },
   imap: {
     icon: EmailIcon,
@@ -302,7 +306,7 @@ export const SOURCE_METADATA_MAP: SourceMap = {
     icon: ColorDiscordIcon,
     displayName: "Discord",
     category: SourceCategory.Messaging,
-    docs: "https://docs.onyx.app/admin/connectors/official/discord",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/discord`,
   },
   xenforo: {
     icon: XenforoIcon,
@@ -313,7 +317,7 @@ export const SOURCE_METADATA_MAP: SourceMap = {
     icon: ZulipIcon,
     displayName: "Zulip",
     category: SourceCategory.Messaging,
-    docs: "https://docs.onyx.app/admin/connectors/official/zulip",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/zulip`,
   },
 
   // Sales
@@ -321,34 +325,34 @@ export const SOURCE_METADATA_MAP: SourceMap = {
     icon: SalesforceIcon,
     displayName: "Salesforce",
     category: SourceCategory.Sales,
-    docs: "https://docs.onyx.app/admin/connectors/official/salesforce",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/salesforce`,
     isPopular: true,
   },
   hubspot: {
     icon: HubSpotIcon,
     displayName: "HubSpot",
     category: SourceCategory.Sales,
-    docs: "https://docs.onyx.app/admin/connectors/official/hubspot",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/hubspot`,
     isPopular: true,
   },
   gong: {
     icon: GongIcon,
     displayName: "Gong",
     category: SourceCategory.Sales,
-    docs: "https://docs.onyx.app/admin/connectors/official/gong",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/gong`,
     isPopular: true,
   },
   fireflies: {
     icon: FirefliesIcon,
     displayName: "Fireflies",
     category: SourceCategory.Sales,
-    docs: "https://docs.onyx.app/admin/connectors/official/fireflies",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/fireflies`,
   },
   highspot: {
     icon: HighspotIcon,
     displayName: "Highspot",
     category: SourceCategory.Sales,
-    docs: "https://docs.onyx.app/admin/connectors/official/highspot",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/highspot`,
   },
   loopio: {
     icon: LoopioIcon,
@@ -361,20 +365,20 @@ export const SOURCE_METADATA_MAP: SourceMap = {
     icon: GithubIcon,
     displayName: "Github",
     category: SourceCategory.CodeRepository,
-    docs: "https://docs.onyx.app/admin/connectors/official/github",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/github`,
     isPopular: true,
   },
   gitlab: {
     icon: GitlabIcon,
     displayName: "Gitlab",
     category: SourceCategory.CodeRepository,
-    docs: "https://docs.onyx.app/admin/connectors/official/gitlab",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/gitlab`,
   },
   bitbucket: {
     icon: BitbucketIcon,
     displayName: "Bitbucket",
     category: SourceCategory.CodeRepository,
-    docs: "https://docs.onyx.app/connectors/bitbucket",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/bitbucket`,
   },
 
   // Others
@@ -382,14 +386,14 @@ export const SOURCE_METADATA_MAP: SourceMap = {
     icon: GlobeIcon2,
     displayName: "Web",
     category: SourceCategory.Other,
-    docs: "https://docs.onyx.app/admin/connectors/official/web",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/web`,
     isPopular: true,
   },
   file: {
     icon: FileIcon2,
     displayName: "File",
     category: SourceCategory.Other,
-    docs: "https://docs.onyx.app/admin/connectors/official/file",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/file`,
     isPopular: true,
   },
   user_file: {
@@ -397,7 +401,7 @@ export const SOURCE_METADATA_MAP: SourceMap = {
     icon: FileIcon2,
     displayName: "File",
     category: SourceCategory.Other,
-    docs: "https://docs.onyx.app/admin/connectors/official/file",
+    docs: `${DOCS_ADMINS_PATH}/connectors/official/file`,
     isPopular: false, // Needs to be false to hide from the Add Connector page
   },
 

web/src/lib/utils.ts

@@ -11,28 +11,33 @@ export const truncateString = (str: string, maxLength: number) => {
 };
 
 /**
- * Custom URL transformer function for ReactMarkdown
- * Allows specific protocols to be used in markdown links
- * We use this with the urlTransform prop in ReactMarkdown
+ * Custom URL transformer function for ReactMarkdown.
+ * Only allows a small, safe set of protocols and strips everything else.
+ * Returning null removes the href attribute entirely.
  */
-export function transformLinkUri(href: string) {
-  if (!href) return href;
+export function transformLinkUri(href: string): string | null {
+  if (!href) return null;
+
+  const trimmedHref = href.trim();
+  if (!trimmedHref) return null;
 
-  const url = href.trim();
   try {
-    const parsedUrl = new URL(url);
-    if (
-      ALLOWED_URL_PROTOCOLS.some((protocol) =>
-        parsedUrl.protocol.startsWith(protocol)
-      )
-    ) {
-      return url;
+    const parsedUrl = new URL(trimmedHref);
+    const protocol = parsedUrl.protocol.toLowerCase();
+
+    if (ALLOWED_URL_PROTOCOLS.some((allowed) => allowed === protocol)) {
+      return trimmedHref;
     }
+
+    return null;
   } catch {
-    // If it's not a valid URL with protocol, return the original href
-    return href;
+    // Allow relative URLs, but drop anything that looks like a protocol-prefixed link
+    if (/^[a-zA-Z][a-zA-Z\d+.-]*:\S*/.test(trimmedHref)) {
+      return null;
+    }
+
+    return trimmedHref;
   }
-  return href;
 }
 
 export function isSubset(parent: string[], child: string[]): boolean {

web/src/refresh-components/onboarding/constants.tsx

@@ -13,6 +13,7 @@ import SvgOpenai from "@/icons/openai";
 import SvgOpenrouter from "@/icons/openrouter";
 import { LLMProviderName } from "@/app/admin/configuration/llm/interfaces";
 import InlineExternalLink from "../InlineExternalLink";
+import { DOCS_ADMINS_PATH, DOCS_BASE_URL } from "@/lib/constants";
 
 type StepConfig = {
   index: number;
@@ -79,7 +80,7 @@ export const FINAL_SETUP_CONFIG: FinalStepItemProps[] = [
     description: "Set up models to create images in your chats.",
     icon: SvgImage,
     buttonText: "Image Generation",
-    buttonHref: "https://docs.onyx.app/overview/core_features/image_generation",
+    buttonHref: `${DOCS_BASE_URL}/overview/core_features/image_generation`,
   },
   {
     title: "Invite your team",
@@ -216,7 +217,9 @@ export const MODAL_CONTENT_MAP: Record<string, any> = {
       BEDROCK_AUTH_METHOD: (
         <>
           {"See "}
-          <InlineExternalLink href="https://docs.onyx.app/admin/ai_models/bedrock#authentication-methods">
+          <InlineExternalLink
+            href={`${DOCS_ADMINS_PATH}/ai_models/bedrock#authentication-methods`}
+          >
             documentation
           </InlineExternalLink>
           {" for more instructions."}