Make better

.
2026-03-07 00:25:48 +00:00 · 2026-03-06 11:22:38 -08:00 · 2026-03-06 11:22:38 -08:00 · 2026-03-06 11:22:38 -08:00 · 2026-03-06 11:22:38 -08:00
55 changed files with 1863 additions and 2049 deletions
--- a/.cursor/skills/onyx-cli/SKILL.md
+++ b/.cursor/skills/onyx-cli/SKILL.md
@@ -106,34 +106,13 @@ onyx-cli ask --json "What authentication methods do we support?"

 Outputs JSON-encoded parsed stream events (one object per line). Key event objects include message deltas, stop, errors, search-start, and citation payloads.

-Each line is a JSON object with this envelope:
-
-```json
-{"type": "<event_type>", "event": { ... }}
-```
-
 | Event Type | Description |
 |------------|-------------|
 | `message_delta` | Content token — concatenate all `content` fields for the full answer |
 | `stop` | Stream complete |
 | `error` | Error with `error` message field |
 | `search_tool_start` | Onyx started searching documents |
-| `citation_info` | Source citation — see shape below |
-
-`citation_info` event shape:
-
-```json
-{
-  "type": "citation_info",
-  "event": {
-    "citation_number": 1,
-    "document_id": "abc123def456",
-    "placement": {"turn_index": 0, "tab_index": 0, "sub_turn_index": null}
-  }
-}
-```
-
-`placement` is metadata about where in the conversation the citation appeared and can be ignored for most use cases.
+| `citation_info` | Source citation with `citation_number` and `document_id` |

 ### Specify an agent

@@ -150,10 +129,6 @@ Uses a specific Onyx agent/persona instead of the default.
 | `--agent-id` | int | Agent ID to use (overrides default) |
 | `--json` | bool | Output raw NDJSON events instead of plain text |

-## Statelessness
-
-Each `onyx-cli ask` call creates an independent chat session. There is no built-in way to chain context across multiple `ask` invocations — every call starts fresh. If you need multi-turn conversation with memory, use the interactive TUI (`onyx-cli` or `onyx-cli chat`) instead.
-
 ## When to Use

 Use `onyx-cli ask` when:
--- a/.github/workflows/pr-desktop-build.yml
+++ b/.github/workflows/pr-desktop-build.yml
@@ -57,7 +57,7 @@ jobs:
          cache-dependency-path: ./desktop/package-lock.json

      - name: Setup Rust
-        uses: dtolnay/rust-toolchain@efa25f7f19611383d5b0ccf2d1c8914531636bf9
+        uses: dtolnay/rust-toolchain@4be9e76fd7c4901c61fb841f559994984270fce7
        with:
          toolchain: stable
          targets: ${{ matrix.target }}
--- a/.github/workflows/release-cli.yml
+++ b/.github/workflows/release-cli.yml
@@ -1,39 +0,0 @@
-name: Release CLI
-
-on:
-  push:
-    tags:
-      - "cli/v*.*.*"
-
-jobs:
-  pypi:
-    runs-on: ubuntu-latest
-    environment:
-      name: release-cli
-    permissions:
-      id-token: write
-    timeout-minutes: 10
-    strategy:
-      matrix:
-        os-arch:
-          - { goos: "linux", goarch: "amd64" }
-          - { goos: "linux", goarch: "arm64" }
-          - { goos: "windows", goarch: "amd64" }
-          - { goos: "windows", goarch: "arm64" }
-          - { goos: "darwin", goarch: "amd64" }
-          - { goos: "darwin", goarch: "arm64" }
-    steps:
-      - uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # ratchet:actions/checkout@v6
-        with:
-          persist-credentials: false
-      - uses: astral-sh/setup-uv@61cb8a9741eeb8a550a1b8544337180c0fc8476b # ratchet:astral-sh/setup-uv@v7
-        with:
-          enable-cache: false
-          version: "0.9.9"
-      - run: |
-          GOOS="${{ matrix.os-arch.goos }}" \
-          GOARCH="${{ matrix.os-arch.goarch }}" \
-          uv build --wheel
-        working-directory: cli
-      - run: uv publish
-        working-directory: cli
--- a/.github/workflows/release-devtools.yml
+++ b/.github/workflows/release-devtools.yml
@@ -22,10 +22,12 @@ jobs:
          - { goos: "windows", goarch: "arm64" }
          - { goos: "darwin", goarch: "amd64" }
          - { goos: "darwin", goarch: "arm64" }
+          - { goos: "", goarch: "" }
    steps:
      - uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # ratchet:actions/checkout@v6
        with:
          persist-credentials: false
+          fetch-depth: 0
      - uses: astral-sh/setup-uv@61cb8a9741eeb8a550a1b8544337180c0fc8476b # ratchet:astral-sh/setup-uv@v7
        with:
          enable-cache: false
--- a/backend/onyx/db/llm.py
+++ b/backend/onyx/db/llm.py
@@ -270,10 +270,34 @@ def upsert_llm_provider(
        mc.name for mc in llm_provider_upsert_request.model_configurations
    }

+    default_model = fetch_default_llm_model(db_session)
+
+    # Build a lookup of requested visibility by model name
+    requested_visibility = {
+        mc.name: mc.is_visible
+        for mc in llm_provider_upsert_request.model_configurations
+    }
+
    # Delete removed models
    removed_ids = [
        mc.id for name, mc in existing_by_name.items() if name not in models_to_exist
    ]
+
+    # Prevent removing and hiding the default model
+    if default_model:
+        for name, mc in existing_by_name.items():
+            if mc.id == default_model.id:
+                if name not in models_to_exist:
+                    raise ValueError(
+                        f"Cannot remove the default model '{name}'. "
+                        "Please change the default model before removing."
+                    )
+                if not requested_visibility.get(name, True):
+                    raise ValueError(
+                        f"Cannot hide the default model '{name}'. "
+                        "Please change the default model before hiding."
+                    )
+
    if removed_ids:
        db_session.query(ModelConfiguration).filter(
            ModelConfiguration.id.in_(removed_ids)
@@ -538,7 +562,6 @@ def fetch_default_model(
        .options(selectinload(ModelConfiguration.llm_provider))
        .join(LLMModelFlow)
        .where(
-            ModelConfiguration.is_visible == True,  # noqa: E712
            LLMModelFlow.llm_model_flow_type == flow_type,
            LLMModelFlow.is_default == True,  # noqa: E712
        )
--- a/backend/onyx/document_index/FILTER_SEMANTICS.md
+++ b/backend/onyx/document_index/FILTER_SEMANTICS.md
@@ -1,103 +0,0 @@
-# Vector DB Filter Semantics
-
-How `IndexFilters` fields combine into the final query filter. Applies to both Vespa and OpenSearch.
-
-## Filter categories
-
-| Category | Fields | Join logic |
-|---|---|---|
-| **Visibility** | `hidden` | Always applied (unless `include_hidden`) |
-| **Tenant** | `tenant_id` | AND (multi-tenant only) |
-| **ACL** | `access_control_list` | OR within, AND with rest |
-| **Narrowing** | `source_type`, `tags`, `time_cutoff` | Each OR within, AND with rest |
-| **Knowledge scope** | `document_set`, `user_file_ids`, `attached_document_ids`, `hierarchy_node_ids` | OR within group, AND with rest |
-| **Additive scope** | `project_id`, `persona_id` | OR'd into knowledge scope **only when** a knowledge scope filter already exists |
-
-## How filters combine
-
-All categories are AND'd together. Within the knowledge scope category, individual filters are OR'd.
-
-```
-NOT hidden
-AND tenant = T                          -- if multi-tenant
-AND (acl contains A1 OR acl contains A2)
-AND (source_type = S1 OR ...)           -- if set
-AND (tag = T1 OR ...)                   -- if set
-AND <knowledge scope>                   -- see below
-AND time >= cutoff                      -- if set
-```
-
-## Knowledge scope rules
-
-The knowledge scope filter controls **what knowledge an assistant can access**.
-
-### No explicit knowledge attached
-
-When `document_set`, `user_file_ids`, `attached_document_ids`, and `hierarchy_node_ids` are all empty/None:
-
- **No knowledge scope filter is applied.** The assistant can see everything (subject to ACL).
- `project_id` and `persona_id` are ignored — they never restrict on their own.
-
-### One explicit knowledge type
-
-```
-- Only document sets
-AND (document_sets contains "Engineering" OR document_sets contains "Legal")
-
-- Only user files
-AND (document_id = "uuid-1" OR document_id = "uuid-2")
-```
-
-### Multiple explicit knowledge types (OR'd)
-
-```
-- Document sets + user files
-AND (
-    document_sets contains "Engineering"
-    OR document_id = "uuid-1"
-)
-```
-
-### Explicit knowledge + overflowing user files
-
-When an explicit knowledge restriction is in effect **and** `project_id` or `persona_id` is set (user files overflowed the LLM context window), the additive scopes widen the filter:
-
-```
-- Document sets + persona user files overflowed
-AND (
-    document_sets contains "Engineering"
-    OR personas contains 42
-)
-
-- User files + project files overflowed
-AND (
-    document_id = "uuid-1"
-    OR user_project contains 7
-)
-```
-
-### Only project_id or persona_id (no explicit knowledge)
-
-No knowledge scope filter. The assistant searches everything.
-
-```
-- Just ACL, no restriction
-NOT hidden
-AND (acl contains ...)
-```
-
-## Field reference
-
-| Filter field | Vespa field | Vespa type | Purpose |
-|---|---|---|---|
-| `document_set` | `document_sets` | `weightedset<string>` | Connector doc sets attached to assistant |
-| `user_file_ids` | `document_id` | `string` | User files uploaded to assistant |
-| `attached_document_ids` | `document_id` | `string` | Documents explicitly attached (OpenSearch only) |
-| `hierarchy_node_ids` | `ancestor_hierarchy_node_ids` | `array<int>` | Folder/space nodes (OpenSearch only) |
-| `project_id` | `user_project` | `array<int>` | Project tag for overflowing user files |
-| `persona_id` | `personas` | `array<int>` | Persona tag for overflowing user files |
-| `access_control_list` | `access_control_list` | `weightedset<string>` | ACL entries for the requesting user |
-| `source_type` | `source_type` | `string` | Connector source type (e.g. `web`, `jira`) |
-| `tags` | `metadata_list` | `array<string>` | Document metadata tags |
-| `time_cutoff` | `doc_updated_at` | `long` | Minimum document update timestamp |
-| `tenant_id` | `tenant_id` | `string` | Tenant isolation (multi-tenant) |
--- a/backend/onyx/document_index/opensearch/search.py
+++ b/backend/onyx/document_index/opensearch/search.py
@@ -698,6 +698,41 @@ class DocumentQuery:
            """
            return {"terms": {ANCESTOR_HIERARCHY_NODE_IDS_FIELD_NAME: node_ids}}

+        def _get_assistant_knowledge_filter(
+            attached_doc_ids: list[str] | None,
+            node_ids: list[int] | None,
+            file_ids: list[UUID] | None,
+            document_sets: list[str] | None,
+        ) -> dict[str, Any]:
+            """Combined filter for assistant knowledge.
+
+            When an assistant has attached knowledge, search should be scoped to:
+            - Documents explicitly attached (by document ID), OR
+            - Documents under attached hierarchy nodes (by ancestor node IDs), OR
+            - User-uploaded files attached to the assistant, OR
+            - Documents in the assistant's document sets (if any)
+            """
+            knowledge_filter: dict[str, Any] = {
+                "bool": {"should": [], "minimum_should_match": 1}
+            }
+            if attached_doc_ids:
+                knowledge_filter["bool"]["should"].append(
+                    _get_attached_document_id_filter(attached_doc_ids)
+                )
+            if node_ids:
+                knowledge_filter["bool"]["should"].append(
+                    _get_hierarchy_node_filter(node_ids)
+                )
+            if file_ids:
+                knowledge_filter["bool"]["should"].append(
+                    _get_user_file_id_filter(file_ids)
+                )
+            if document_sets:
+                knowledge_filter["bool"]["should"].append(
+                    _get_document_set_filter(document_sets)
+                )
+            return knowledge_filter
+
        filter_clauses: list[dict[str, Any]] = []

        if not include_hidden:
@@ -723,53 +758,41 @@ class DocumentQuery:
            # document's metadata list.
            filter_clauses.append(_get_tag_filter(tags))

-        # Knowledge scope: explicit knowledge attachments restrict what
-        # an assistant can see.  When none are set the assistant
-        # searches everything.
-        #
-        # project_id / persona_id are additive: they make overflowing
-        # user files findable but must NOT trigger the restriction on
-        # their own (an agent with no explicit knowledge should search
-        # everything).
-        has_knowledge_scope = (
+        # Check if this is an assistant knowledge search (has any assistant-scoped knowledge)
+        has_assistant_knowledge = (
            attached_document_ids
            or hierarchy_node_ids
            or user_file_ids
            or document_sets
        )

-        if has_knowledge_scope:
-            knowledge_filter: dict[str, Any] = {
-                "bool": {"should": [], "minimum_should_match": 1}
-            }
-            if attached_document_ids:
-                knowledge_filter["bool"]["should"].append(
-                    _get_attached_document_id_filter(attached_document_ids)
+        if has_assistant_knowledge:
+            # If assistant has attached knowledge, scope search to that knowledge.
+            # Document sets are included in the OR filter so directly attached
+            # docs are always findable even if not in the document sets.
+            filter_clauses.append(
+                _get_assistant_knowledge_filter(
+                    attached_document_ids,
+                    hierarchy_node_ids,
+                    user_file_ids,
+                    document_sets,
                )
-            if hierarchy_node_ids:
-                knowledge_filter["bool"]["should"].append(
-                    _get_hierarchy_node_filter(hierarchy_node_ids)
-                )
-            if user_file_ids:
-                knowledge_filter["bool"]["should"].append(
-                    _get_user_file_id_filter(user_file_ids)
-                )
-            if document_sets:
-                knowledge_filter["bool"]["should"].append(
-                    _get_document_set_filter(document_sets)
-                )
-            # Additive: widen scope to also cover overflowing user
-            # files, but only when an explicit restriction is already
-            # in effect.
-            if project_id is not None:
-                knowledge_filter["bool"]["should"].append(
-                    _get_user_project_filter(project_id)
-                )
-            if persona_id is not None:
-                knowledge_filter["bool"]["should"].append(
-                    _get_persona_filter(persona_id)
-                )
-            filter_clauses.append(knowledge_filter)
+            )
+        elif user_file_ids:
+            # Fallback for non-assistant user file searches (e.g., project searches)
+            # If at least one user file ID is provided, the caller will only
+            # retrieve documents where the document ID is in this input list of
+            # file IDs.
+            filter_clauses.append(_get_user_file_id_filter(user_file_ids))
+
+        if project_id is not None:
+            # If a project ID is provided, the caller will only retrieve
+            # documents where the project ID provided here is present in the
+            # document's user projects list.
+            filter_clauses.append(_get_user_project_filter(project_id))
+
+        if persona_id is not None:
+            filter_clauses.append(_get_persona_filter(persona_id))

        if time_cutoff is not None:
            # If a time cutoff is provided, the caller will only retrieve
--- a/backend/onyx/document_index/vespa/shared_utils/vespa_request_builders.py
+++ b/backend/onyx/document_index/vespa/shared_utils/vespa_request_builders.py
@@ -23,8 +23,11 @@ from shared_configs.configs import MULTI_TENANT
 logger = setup_logger()


-def build_tenant_id_filter(tenant_id: str) -> str:
-    return f'({TENANT_ID} contains "{tenant_id}")'
+def build_tenant_id_filter(tenant_id: str, include_trailing_and: bool = False) -> str:
+    filter_str = f'({TENANT_ID} contains "{tenant_id}")'
+    if include_trailing_and:
+        filter_str += " and "
+    return filter_str


 def build_vespa_filters(
@@ -34,22 +37,30 @@ def build_vespa_filters(
    remove_trailing_and: bool = False,  # Set to True when using as a complete Vespa query
 ) -> str:
    def _build_or_filters(key: str, vals: list[str] | None) -> str:
-        """For string-based 'contains' filters, e.g. WSET fields or array<string> fields.
-        Returns a bare clause like '(key contains "v1" or key contains "v2")' or ""."""
+        """For string-based 'contains' filters, e.g. WSET fields or array<string> fields."""
        if not key or not vals:
            return ""
        eq_elems = [f'{key} contains "{val}"' for val in vals if val]
        if not eq_elems:
            return ""
-        return f"({' or '.join(eq_elems)})"
+        or_clause = " or ".join(eq_elems)
+        return f"({or_clause}) and "

    def _build_int_or_filters(key: str, vals: list[int] | None) -> str:
-        """For an integer field filter.
-        Returns a bare clause or ""."""
+        """
+        For an integer field filter.
+        If vals is not None, we want *only* docs whose key matches one of vals.
+        """
+        # If `vals` is None => skip the filter entirely
        if vals is None or not vals:
            return ""
+
+        # Otherwise build the OR filter
        eq_elems = [f"{key} = {val}" for val in vals]
-        return f"({' or '.join(eq_elems)})"
+        or_clause = " or ".join(eq_elems)
+        result = f"({or_clause}) and "
+
+        return result

    def _build_kg_filter(
        kg_entities: list[str] | None,
@@ -62,12 +73,16 @@ def build_vespa_filters(
        combined_filter_parts = []

        def _build_kge(entity: str) -> str:
+            # TYPE-SUBTYPE::ID -> "TYPE-SUBTYPE::ID"
+            # TYPE-SUBTYPE::*  -> ({prefix: true}"TYPE-SUBTYPE")
+            # TYPE::*          -> ({prefix: true}"TYPE")
            GENERAL = "::*"
            if entity.endswith(GENERAL):
                return f'({{prefix: true}}"{entity.split(GENERAL, 1)[0]}")'
            else:
                return f'"{entity}"'

+        # OR the entities (give new design)
        if kg_entities:
            filter_parts = []
            for kg_entity in kg_entities:
@@ -89,7 +104,8 @@ def build_vespa_filters(

        # TODO: remove kg terms entirely from prompts and codebase

-        return f"({' and '.join(combined_filter_parts)})"
+        # AND the combined filter parts
+        return f"({' and '.join(combined_filter_parts)}) and "

    def _build_kg_source_filters(
        kg_sources: list[str] | None,
@@ -98,14 +114,16 @@ def build_vespa_filters(
            return ""

        source_phrases = [f'{DOCUMENT_ID} contains "{source}"' for source in kg_sources]
-        return f"({' or '.join(source_phrases)})"
+
+        return f"({' or '.join(source_phrases)}) and "

    def _build_kg_chunk_id_zero_only_filter(
        kg_chunk_id_zero_only: bool,
    ) -> str:
        if not kg_chunk_id_zero_only:
            return ""
-        return "(chunk_id = 0)"
+
+        return "(chunk_id = 0 ) and "

    def _build_time_filter(
        cutoff: datetime | None,
@@ -117,8 +135,8 @@ def build_vespa_filters(
        cutoff_secs = int(cutoff.timestamp())

        if include_untimed:
-            return f"!({DOC_UPDATED_AT} < {cutoff_secs})"
-        return f"({DOC_UPDATED_AT} >= {cutoff_secs})"
+            return f"!({DOC_UPDATED_AT} < {cutoff_secs}) and "
+        return f"({DOC_UPDATED_AT} >= {cutoff_secs}) and "

    def _build_user_project_filter(
        project_id: int | None,
@@ -129,7 +147,8 @@ def build_vespa_filters(
            pid = int(project_id)
        except Exception:
            return ""
-        return f'({USER_PROJECT} contains "{pid}")'
+        # Vespa YQL 'contains' expects a string literal; quote the integer
+        return f'({USER_PROJECT} contains "{pid}") and '

    def _build_persona_filter(
        persona_id: int | None,
@@ -141,94 +160,73 @@ def build_vespa_filters(
        except Exception:
            logger.warning(f"Invalid persona ID: {persona_id}")
            return ""
-        return f'({PERSONAS} contains "{pid}")'
+        return f'({PERSONAS} contains "{pid}") and '

-    def _append(parts: list[str], clause: str) -> None:
-        if clause:
-            parts.append(clause)
-
-    # Collect all top-level filter clauses, then join with " and " at the end.
-    filter_parts: list[str] = []
-
-    if not include_hidden:
-        filter_parts.append(f"!({HIDDEN}=true)")
+    # Start building the filter string
+    filter_str = f"!({HIDDEN}=true) and " if not include_hidden else ""

    # TODO: add error condition if MULTI_TENANT and no tenant_id filter is set
+    # If running in multi-tenant mode
    if filters.tenant_id and MULTI_TENANT:
-        filter_parts.append(build_tenant_id_filter(filters.tenant_id))
+        filter_str += build_tenant_id_filter(
+            filters.tenant_id, include_trailing_and=True
+        )

    # ACL filters
    if filters.access_control_list is not None:
-        _append(
-            filter_parts,
-            _build_or_filters(ACCESS_CONTROL_LIST, filters.access_control_list),
+        filter_str += _build_or_filters(
+            ACCESS_CONTROL_LIST, filters.access_control_list
        )

    # Source type filters
    source_strs = (
        [s.value for s in filters.source_type] if filters.source_type else None
    )
-    _append(filter_parts, _build_or_filters(SOURCE_TYPE, source_strs))
+    filter_str += _build_or_filters(SOURCE_TYPE, source_strs)

    # Tag filters
    tag_attributes = None
    if filters.tags:
+        # build e.g. "tag_key|tag_value"
        tag_attributes = [
            f"{tag.tag_key}{INDEX_SEPARATOR}{tag.tag_value}" for tag in filters.tags
        ]
-    _append(filter_parts, _build_or_filters(METADATA_LIST, tag_attributes))
+    filter_str += _build_or_filters(METADATA_LIST, tag_attributes)

-    # Knowledge scope: explicit knowledge attachments (document_sets,
-    # user_file_ids) restrict what an assistant can see.  When none are
-    # set, the assistant can see everything.
-    #
-    # project_id / persona_id are additive: they make overflowing user
-    # files findable in Vespa but must NOT trigger the restriction on
-    # their own (an agent with no explicit knowledge should search
-    # everything).
-    knowledge_scope_parts: list[str] = []
-
-    _append(
-        knowledge_scope_parts, _build_or_filters(DOCUMENT_SETS, filters.document_set)
-    )
+    # Document sets
+    filter_str += _build_or_filters(DOCUMENT_SETS, filters.document_set)

+    # Convert UUIDs to strings for user_file_ids
    user_file_ids_str = (
        [str(uuid) for uuid in filters.user_file_ids] if filters.user_file_ids else None
    )
-    _append(knowledge_scope_parts, _build_or_filters(DOCUMENT_ID, user_file_ids_str))
+    filter_str += _build_or_filters(DOCUMENT_ID, user_file_ids_str)

-    # Only include project/persona scopes when an explicit knowledge
-    # restriction is already in effect — they widen the scope to also
-    # cover overflowing user files but never restrict on their own.
-    if knowledge_scope_parts:
-        _append(knowledge_scope_parts, _build_user_project_filter(filters.project_id))
-        _append(knowledge_scope_parts, _build_persona_filter(filters.persona_id))
+    # User project filter (array<int> attribute membership)
+    filter_str += _build_user_project_filter(filters.project_id)

-    if len(knowledge_scope_parts) > 1:
-        filter_parts.append("(" + " or ".join(knowledge_scope_parts) + ")")
-    elif len(knowledge_scope_parts) == 1:
-        filter_parts.append(knowledge_scope_parts[0])
+    # Persona filter (array<int> attribute membership)
+    filter_str += _build_persona_filter(filters.persona_id)

    # Time filter
-    _append(filter_parts, _build_time_filter(filters.time_cutoff))
+    filter_str += _build_time_filter(filters.time_cutoff)

    # # Knowledge Graph Filters
-    # _append(filter_parts, _build_kg_filter(
+    # filter_str += _build_kg_filter(
    #     kg_entities=filters.kg_entities,
    #     kg_relationships=filters.kg_relationships,
    #     kg_terms=filters.kg_terms,
-    # ))
+    # )

-    # _append(filter_parts, _build_kg_source_filters(filters.kg_sources))
+    # filter_str += _build_kg_source_filters(filters.kg_sources)

-    # _append(filter_parts, _build_kg_chunk_id_zero_only_filter(
+    # filter_str += _build_kg_chunk_id_zero_only_filter(
    #     filters.kg_chunk_id_zero_only or False
-    # ))
+    # )

-    filter_str = " and ".join(filter_parts)
-
-    if filter_str and not remove_trailing_and:
-        filter_str += " and "
+    # Trim trailing " and "
+    if remove_trailing_and and filter_str.endswith(" and "):
+        filter_str = filter_str[:-5]

    return filter_str

--- a/backend/onyx/llm/model_metadata_enrichments.json
+++ b/backend/onyx/llm/model_metadata_enrichments.json
@@ -1512,10 +1512,6 @@
    "display_name": "Claude Opus 4.5",
    "model_vendor": "anthropic"
  },
-  "claude-opus-4-6": {
-    "display_name": "Claude Opus 4.6",
-    "model_vendor": "anthropic"
-  },
  "claude-opus-4-5-20251101": {
    "display_name": "Claude Opus 4.5",
    "model_vendor": "anthropic",
@@ -1530,10 +1526,6 @@
    "display_name": "Claude Sonnet 4.5",
    "model_vendor": "anthropic"
  },
-  "claude-sonnet-4-6": {
-    "display_name": "Claude Sonnet 4.6",
-    "model_vendor": "anthropic"
-  },
  "claude-sonnet-4-5-20250929": {
    "display_name": "Claude Sonnet 4.5",
    "model_vendor": "anthropic",
--- a/backend/onyx/llm/well_known_providers/constants.py
+++ b/backend/onyx/llm/well_known_providers/constants.py
@@ -1,8 +1,37 @@
 from onyx.llm.constants import LlmProviderNames

 OPENAI_PROVIDER_NAME = "openai"
+# Curated list of OpenAI models to show by default in the UI
+OPENAI_VISIBLE_MODEL_NAMES = {
+    "gpt-5",
+    "gpt-5-mini",
+    "o1",
+    "o3-mini",
+    "gpt-4o",
+    "gpt-4o-mini",
+}

 BEDROCK_PROVIDER_NAME = "bedrock"
+BEDROCK_DEFAULT_MODEL = "anthropic.claude-3-5-sonnet-20241022-v2:0"
+
+
+def _fallback_bedrock_regions() -> list[str]:
+    # Fall back to a conservative set of well-known Bedrock regions if boto3 data isn't available.
+    return [
+        "us-east-1",
+        "us-east-2",
+        "us-gov-east-1",
+        "us-gov-west-1",
+        "us-west-2",
+        "ap-northeast-1",
+        "ap-south-1",
+        "ap-southeast-1",
+        "ap-southeast-2",
+        "ap-east-1",
+        "ca-central-1",
+        "eu-central-1",
+        "eu-west-2",
+    ]


 OLLAMA_PROVIDER_NAME = "ollama_chat"
@@ -22,6 +51,13 @@ OPENROUTER_PROVIDER_NAME = "openrouter"

 ANTHROPIC_PROVIDER_NAME = "anthropic"

+# Curated list of Anthropic models to show by default in the UI
+ANTHROPIC_VISIBLE_MODEL_NAMES = {
+    "claude-opus-4-5",
+    "claude-sonnet-4-5",
+    "claude-haiku-4-5",
+}
+
 AZURE_PROVIDER_NAME = "azure"


@@ -29,6 +65,13 @@ VERTEXAI_PROVIDER_NAME = "vertex_ai"
 VERTEX_CREDENTIALS_FILE_KWARG = "vertex_credentials"
 VERTEX_CREDENTIALS_FILE_KWARG_ENV_VAR_FORMAT = "CREDENTIALS_FILE"
 VERTEX_LOCATION_KWARG = "vertex_location"
+VERTEXAI_DEFAULT_MODEL = "gemini-2.5-flash"
+# Curated list of Vertex AI models to show by default in the UI
+VERTEXAI_VISIBLE_MODEL_NAMES = {
+    "gemini-2.5-flash",
+    "gemini-2.5-flash-lite",
+    "gemini-2.5-pro",
+}

 AWS_REGION_NAME_KWARG = "aws_region_name"
 AWS_REGION_NAME_KWARG_ENV_VAR_FORMAT = "AWS_REGION_NAME"
--- a/backend/onyx/llm/well_known_providers/recommended-models.json
+++ b/backend/onyx/llm/well_known_providers/recommended-models.json
@@ -16,10 +16,6 @@
          "name": "claude-opus-4-6",
          "display_name": "Claude Opus 4.6"
        },
-        {
-          "name": "claude-sonnet-4-6",
-          "display_name": "Claude Sonnet 4.6"
-        },
        {
          "name": "claude-opus-4-5",
          "display_name": "Claude Opus 4.5"
--- a/backend/onyx/server/features/build/sandbox/kubernetes/docker/templates/outputs/web/package-lock.json
+++ b/backend/onyx/server/features/build/sandbox/kubernetes/docker/templates/outputs/web/package-lock.json
@@ -6758,12 +6758,12 @@
      }
    },
    "node_modules/express-rate-limit": {
-      "version": "8.3.0",
-      "resolved": "https://registry.npmjs.org/express-rate-limit/-/express-rate-limit-8.3.0.tgz",
-      "integrity": "sha512-KJzBawY6fB9FiZGdE/0aftepZ91YlaGIrV8vgblRM3J8X+dHx/aiowJWwkx6LIGyuqGiANsjSwwrbb8mifOJ4Q==",
+      "version": "8.2.1",
+      "resolved": "https://registry.npmjs.org/express-rate-limit/-/express-rate-limit-8.2.1.tgz",
+      "integrity": "sha512-PCZEIEIxqwhzw4KF0n7QF4QqruVTcF73O5kFKUnGOyjbCCgizBBiFaYpd/fnBLUMPw/BWw9OsiN7GgrNYr7j6g==",
      "license": "MIT",
      "dependencies": {
-        "ip-address": "10.1.0"
+        "ip-address": "10.0.1"
      },
      "engines": {
        "node": ">= 16"
@@ -7556,9 +7556,9 @@
      }
    },
    "node_modules/ip-address": {
-      "version": "10.1.0",
-      "resolved": "https://registry.npmjs.org/ip-address/-/ip-address-10.1.0.tgz",
-      "integrity": "sha512-XXADHxXmvT9+CRxhXg56LJovE+bmWnEWB78LB83VZTprKTmaC5QfruXocxzTZ2Kl0DNwKuBdlIhjL8LeY8Sf8Q==",
+      "version": "10.0.1",
+      "resolved": "https://registry.npmjs.org/ip-address/-/ip-address-10.0.1.tgz",
+      "integrity": "sha512-NWv9YLW4PoW2B7xtzaS3NCot75m6nK7Icdv0o3lfMceJVRfSoQwqD4wEH5rLwoKJwUiZ/rfpiVBhnaF0FK4HoA==",
      "license": "MIT",
      "engines": {
        "node": ">= 12"
--- a/backend/onyx/server/manage/llm/api.py
+++ b/backend/onyx/server/manage/llm/api.py
@@ -65,6 +65,7 @@ from onyx.server.manage.llm.models import LLMProviderUpsertRequest
 from onyx.server.manage.llm.models import LLMProviderView
 from onyx.server.manage.llm.models import LMStudioFinalModelResponse
 from onyx.server.manage.llm.models import LMStudioModelsRequest
+from onyx.server.manage.llm.models import ModelConfigurationUpsertRequest
 from onyx.server.manage.llm.models import OllamaFinalModelResponse
 from onyx.server.manage.llm.models import OllamaModelDetails
 from onyx.server.manage.llm.models import OllamaModelsRequest
@@ -445,16 +446,17 @@ def put_llm_provider(
        not existing_provider or not existing_provider.is_auto_mode
    )

-    # Before the upsert, check if this provider currently owns the global
-    # CHAT default. The upsert may cascade-delete model_configurations
-    # (and their flow mappings), so we need to remember this beforehand.
-    was_default_provider = False
-    if existing_provider and transitioning_to_auto_mode:
-        current_default = fetch_default_llm_model(db_session)
-        was_default_provider = (
-            current_default is not None
-            and current_default.llm_provider_id == existing_provider.id
-        )
+    # When transitioning to auto mode, preserve existing model configurations
+    # so the upsert doesn't try to delete them (which would trip the default
+    # model protection guard). sync_auto_mode_models will handle the model
+    # lifecycle afterward — adding new models, hiding removed ones, and
+    # updating the default. This is safe even if sync fails: the provider
+    # keeps its old models and default rather than losing them.
+    if transitioning_to_auto_mode and existing_provider:
+        llm_provider_upsert_request.model_configurations = [
+            ModelConfigurationUpsertRequest.from_model(mc)
+            for mc in existing_provider.model_configurations
+        ]

    try:
        result = upsert_llm_provider(
@@ -468,7 +470,6 @@ def put_llm_provider(

            config = fetch_llm_recommendations_from_github()
            if config and llm_provider_upsert_request.provider in config.providers:
-                # Refetch the provider to get the updated model
                updated_provider = fetch_existing_llm_provider_by_id(
                    id=result.id, db_session=db_session
                )
@@ -478,20 +479,6 @@ def put_llm_provider(
                        updated_provider,
                        config,
                    )
-
-                    # If this provider was the default before the transition,
-                    # restore the default using the recommended model.
-                    if was_default_provider:
-                        recommended = config.get_default_model(
-                            llm_provider_upsert_request.provider
-                        )
-                        if recommended:
-                            update_default_provider(
-                                provider_id=updated_provider.id,
-                                model_name=recommended.name,
-                                db_session=db_session,
-                            )
-
                    # Refresh result with synced models
                    result = LLMProviderView.from_model(updated_provider)

--- a/backend/tests/external_dependency_unit/llm/test_llm_provider_auto_mode.py
+++ b/backend/tests/external_dependency_unit/llm/test_llm_provider_auto_mode.py
@@ -1152,3 +1152,179 @@ class TestAutoModeTransitionsAndResync:
        finally:
            db_session.rollback()
            _cleanup_provider(db_session, provider_name)
+
+    def test_sync_updates_default_when_recommended_default_changes(
+        self,
+        db_session: Session,
+        provider_name: str,
+    ) -> None:
+        """When the provider owns the CHAT default and a sync arrives with a
+        different recommended default model (both models still in config),
+        the global default should be updated to the new recommendation.
+
+        Steps:
+        1. Create auto-mode provider with config v1: default=gpt-4o.
+        2. Set gpt-4o as the global CHAT default.
+        3. Re-sync with config v2: default=gpt-4o-mini (gpt-4o still present).
+        4. Verify the CHAT default switched to gpt-4o-mini and both models
+           remain visible.
+        """
+        config_v1 = _create_mock_llm_recommendations(
+            provider=LlmProviderNames.OPENAI,
+            default_model_name="gpt-4o",
+            additional_models=["gpt-4o-mini"],
+        )
+        config_v2 = _create_mock_llm_recommendations(
+            provider=LlmProviderNames.OPENAI,
+            default_model_name="gpt-4o-mini",
+            additional_models=["gpt-4o"],
+        )
+
+        try:
+            with patch(
+                "onyx.server.manage.llm.api.fetch_llm_recommendations_from_github",
+                return_value=config_v1,
+            ):
+                put_llm_provider(
+                    llm_provider_upsert_request=LLMProviderUpsertRequest(
+                        name=provider_name,
+                        provider=LlmProviderNames.OPENAI,
+                        api_key="sk-test-key-00000000000000000000000000000000000",
+                        api_key_changed=True,
+                        is_auto_mode=True,
+                        model_configurations=[],
+                    ),
+                    is_creation=True,
+                    _=_create_mock_admin(),
+                    db_session=db_session,
+                )
+
+            # Set gpt-4o as the global CHAT default
+            db_session.expire_all()
+            provider = fetch_existing_llm_provider(
+                name=provider_name, db_session=db_session
+            )
+            assert provider is not None
+            update_default_provider(provider.id, "gpt-4o", db_session)
+
+            default_before = fetch_default_llm_model(db_session)
+            assert default_before is not None
+            assert default_before.name == "gpt-4o"
+
+            # Re-sync with config v2 (recommended default changed)
+            db_session.expire_all()
+            provider = fetch_existing_llm_provider(
+                name=provider_name, db_session=db_session
+            )
+            assert provider is not None
+
+            changes = sync_auto_mode_models(
+                db_session=db_session,
+                provider=provider,
+                llm_recommendations=config_v2,
+            )
+            assert changes > 0, "Sync should report changes when default switches"
+
+            # Both models should remain visible
+            db_session.expire_all()
+            provider = fetch_existing_llm_provider(
+                name=provider_name, db_session=db_session
+            )
+            assert provider is not None
+            visibility = {
+                mc.name: mc.is_visible for mc in provider.model_configurations
+            }
+            assert visibility["gpt-4o"] is True
+            assert visibility["gpt-4o-mini"] is True
+
+            # The CHAT default should now be gpt-4o-mini
+            default_after = fetch_default_llm_model(db_session)
+            assert default_after is not None
+            assert (
+                default_after.name == "gpt-4o-mini"
+            ), f"Default should be updated to 'gpt-4o-mini', got '{default_after.name}'"
+
+        finally:
+            db_session.rollback()
+            _cleanup_provider(db_session, provider_name)
+
+    def test_sync_idempotent_when_default_already_matches(
+        self,
+        db_session: Session,
+        provider_name: str,
+    ) -> None:
+        """When the provider owns the CHAT default and it already matches the
+        recommended default, re-syncing should report zero changes.
+
+        This is a regression test for the bug where changes was unconditionally
+        incremented even when the default was already correct.
+        """
+        config = _create_mock_llm_recommendations(
+            provider=LlmProviderNames.OPENAI,
+            default_model_name="gpt-4o",
+            additional_models=["gpt-4o-mini"],
+        )
+
+        try:
+            with patch(
+                "onyx.server.manage.llm.api.fetch_llm_recommendations_from_github",
+                return_value=config,
+            ):
+                put_llm_provider(
+                    llm_provider_upsert_request=LLMProviderUpsertRequest(
+                        name=provider_name,
+                        provider=LlmProviderNames.OPENAI,
+                        api_key="sk-test-key-00000000000000000000000000000000000",
+                        api_key_changed=True,
+                        is_auto_mode=True,
+                        model_configurations=[],
+                    ),
+                    is_creation=True,
+                    _=_create_mock_admin(),
+                    db_session=db_session,
+                )
+
+            # Set gpt-4o (the recommended default) as global CHAT default
+            db_session.expire_all()
+            provider = fetch_existing_llm_provider(
+                name=provider_name, db_session=db_session
+            )
+            assert provider is not None
+            update_default_provider(provider.id, "gpt-4o", db_session)
+
+            # First sync to stabilize state
+            db_session.expire_all()
+            provider = fetch_existing_llm_provider(
+                name=provider_name, db_session=db_session
+            )
+            assert provider is not None
+            sync_auto_mode_models(
+                db_session=db_session,
+                provider=provider,
+                llm_recommendations=config,
+            )
+
+            # Second sync — default already matches, should be a no-op
+            db_session.expire_all()
+            provider = fetch_existing_llm_provider(
+                name=provider_name, db_session=db_session
+            )
+            assert provider is not None
+            changes = sync_auto_mode_models(
+                db_session=db_session,
+                provider=provider,
+                llm_recommendations=config,
+            )
+            assert changes == 0, (
+                f"Expected 0 changes when default already matches recommended, "
+                f"got {changes}"
+            )
+
+            # Default should still be gpt-4o
+            default_model = fetch_default_llm_model(db_session)
+            assert default_model is not None
+            assert default_model.name == "gpt-4o"
+
+        finally:
+            db_session.rollback()
+            _cleanup_provider(db_session, provider_name)
--- a/backend/tests/external_dependency_unit/llm/test_llm_provider_default_model_protection.py
+++ b/backend/tests/external_dependency_unit/llm/test_llm_provider_default_model_protection.py
@@ -0,0 +1,220 @@
+"""
+This should act as the main point of reference for testing that default model
+logic is consisten.
+
+ -
+"""
+
+from collections.abc import Generator
+from uuid import uuid4
+
+import pytest
+from sqlalchemy.orm import Session
+
+from onyx.db.llm import fetch_existing_llm_provider
+from onyx.db.llm import remove_llm_provider
+from onyx.db.llm import update_default_provider
+from onyx.db.llm import update_default_vision_provider
+from onyx.db.llm import upsert_llm_provider
+from onyx.llm.constants import LlmProviderNames
+from onyx.server.manage.llm.models import LLMProviderUpsertRequest
+from onyx.server.manage.llm.models import LLMProviderView
+from onyx.server.manage.llm.models import ModelConfigurationUpsertRequest
+
+
+def _create_test_provider(
+    db_session: Session,
+    name: str,
+    models: list[ModelConfigurationUpsertRequest] | None = None,
+) -> LLMProviderView:
+    """Helper to create a test LLM provider with multiple models."""
+    if models is None:
+        models = [
+            ModelConfigurationUpsertRequest(
+                name="gpt-4o", is_visible=True, supports_image_input=True
+            ),
+            ModelConfigurationUpsertRequest(
+                name="gpt-4o-mini", is_visible=True, supports_image_input=False
+            ),
+        ]
+    return upsert_llm_provider(
+        LLMProviderUpsertRequest(
+            name=name,
+            provider=LlmProviderNames.OPENAI,
+            api_key="sk-test-key-00000000000000000000000000000000000",
+            api_key_changed=True,
+            model_configurations=models,
+        ),
+        db_session=db_session,
+    )
+
+
+def _cleanup_provider(db_session: Session, name: str) -> None:
+    """Helper to clean up a test provider by name."""
+    provider = fetch_existing_llm_provider(name=name, db_session=db_session)
+    if provider:
+        remove_llm_provider(db_session, provider.id)
+
+
+@pytest.fixture
+def provider_name(db_session: Session) -> Generator[str, None, None]:
+    """Generate a unique provider name for each test, with automatic cleanup."""
+    name = f"test-provider-{uuid4().hex[:8]}"
+    yield name
+    db_session.rollback()
+    _cleanup_provider(db_session, name)
+
+
+class TestDefaultModelProtection:
+    """Tests that the default model cannot be removed or hidden."""
+
+    def test_cannot_remove_default_text_model(
+        self,
+        db_session: Session,
+        provider_name: str,
+    ) -> None:
+        """Removing the default text model from a provider should raise ValueError."""
+        provider = _create_test_provider(db_session, provider_name)
+        update_default_provider(provider.id, "gpt-4o", db_session)
+
+        # Try to update the provider without the default model
+        with pytest.raises(ValueError, match="Cannot remove the default model"):
+            upsert_llm_provider(
+                LLMProviderUpsertRequest(
+                    id=provider.id,
+                    name=provider_name,
+                    provider=LlmProviderNames.OPENAI,
+                    api_key="sk-test-key-00000000000000000000000000000000000",
+                    api_key_changed=True,
+                    model_configurations=[
+                        ModelConfigurationUpsertRequest(
+                            name="gpt-4o-mini", is_visible=True
+                        ),
+                    ],
+                ),
+                db_session=db_session,
+            )
+
+    def test_cannot_hide_default_text_model(
+        self,
+        db_session: Session,
+        provider_name: str,
+    ) -> None:
+        """Setting is_visible=False on the default text model should raise ValueError."""
+        provider = _create_test_provider(db_session, provider_name)
+        update_default_provider(provider.id, "gpt-4o", db_session)
+
+        # Try to hide the default model
+        with pytest.raises(ValueError, match="Cannot hide the default model"):
+            upsert_llm_provider(
+                LLMProviderUpsertRequest(
+                    id=provider.id,
+                    name=provider_name,
+                    provider=LlmProviderNames.OPENAI,
+                    api_key="sk-test-key-00000000000000000000000000000000000",
+                    api_key_changed=True,
+                    model_configurations=[
+                        ModelConfigurationUpsertRequest(
+                            name="gpt-4o", is_visible=False
+                        ),
+                        ModelConfigurationUpsertRequest(
+                            name="gpt-4o-mini", is_visible=True
+                        ),
+                    ],
+                ),
+                db_session=db_session,
+            )
+
+    def test_cannot_remove_default_vision_model(
+        self,
+        db_session: Session,
+        provider_name: str,
+    ) -> None:
+        """Removing the default vision model from a provider should raise ValueError."""
+        provider = _create_test_provider(db_session, provider_name)
+        # Set gpt-4o as both the text and vision default
+        update_default_provider(provider.id, "gpt-4o", db_session)
+        update_default_vision_provider(provider.id, "gpt-4o", db_session)
+
+        # Try to remove the default vision model
+        with pytest.raises(ValueError, match="Cannot remove the default model"):
+            upsert_llm_provider(
+                LLMProviderUpsertRequest(
+                    id=provider.id,
+                    name=provider_name,
+                    provider=LlmProviderNames.OPENAI,
+                    api_key="sk-test-key-00000000000000000000000000000000000",
+                    api_key_changed=True,
+                    model_configurations=[
+                        ModelConfigurationUpsertRequest(
+                            name="gpt-4o-mini", is_visible=True
+                        ),
+                    ],
+                ),
+                db_session=db_session,
+            )
+
+    def test_can_remove_non_default_model(
+        self,
+        db_session: Session,
+        provider_name: str,
+    ) -> None:
+        """Removing a non-default model should succeed."""
+        provider = _create_test_provider(db_session, provider_name)
+        update_default_provider(provider.id, "gpt-4o", db_session)
+
+        # Remove gpt-4o-mini (not default) — should succeed
+        updated = upsert_llm_provider(
+            LLMProviderUpsertRequest(
+                id=provider.id,
+                name=provider_name,
+                provider=LlmProviderNames.OPENAI,
+                api_key="sk-test-key-00000000000000000000000000000000000",
+                api_key_changed=True,
+                model_configurations=[
+                    ModelConfigurationUpsertRequest(
+                        name="gpt-4o", is_visible=True, supports_image_input=True
+                    ),
+                ],
+            ),
+            db_session=db_session,
+        )
+
+        model_names = {mc.name for mc in updated.model_configurations}
+        assert "gpt-4o" in model_names
+        assert "gpt-4o-mini" not in model_names
+
+    def test_can_hide_non_default_model(
+        self,
+        db_session: Session,
+        provider_name: str,
+    ) -> None:
+        """Hiding a non-default model should succeed."""
+        provider = _create_test_provider(db_session, provider_name)
+        update_default_provider(provider.id, "gpt-4o", db_session)
+
+        # Hide gpt-4o-mini (not default) — should succeed
+        updated = upsert_llm_provider(
+            LLMProviderUpsertRequest(
+                id=provider.id,
+                name=provider_name,
+                provider=LlmProviderNames.OPENAI,
+                api_key="sk-test-key-00000000000000000000000000000000000",
+                api_key_changed=True,
+                model_configurations=[
+                    ModelConfigurationUpsertRequest(
+                        name="gpt-4o", is_visible=True, supports_image_input=True
+                    ),
+                    ModelConfigurationUpsertRequest(
+                        name="gpt-4o-mini", is_visible=False
+                    ),
+                ],
+            ),
+            db_session=db_session,
+        )
+
+        model_visibility = {
+            mc.name: mc.is_visible for mc in updated.model_configurations
+        }
+        assert model_visibility["gpt-4o"] is True
+        assert model_visibility["gpt-4o-mini"] is False
--- a/backend/tests/unit/onyx/utils/test_vespa_query.py
+++ b/backend/tests/unit/onyx/utils/test_vespa_query.py
@@ -20,6 +20,8 @@ from onyx.document_index.vespa_constants import TENANT_ID
 from onyx.document_index.vespa_constants import USER_PROJECT
 from shared_configs.configs import MULTI_TENANT

+# Import the function under test
+

 class TestBuildVespaFilters:
    def test_empty_filters(self) -> None:
@@ -177,27 +179,11 @@ class TestBuildVespaFilters:
        assert f"!({HIDDEN}=true) and " == result

    def test_user_project_filter(self) -> None:
-        """Test user project filtering.
-
-        project_id alone does NOT trigger a knowledge scope restriction
-        (an agent with no explicit knowledge should search everything).
-        It only participates when explicit knowledge filters are present.
-        """
-        # project_id alone → no restriction
+        """Test user project filtering (replacement for user folder IDs)."""
+        # Single project id
        filters = IndexFilters(access_control_list=[], project_id=789)
        result = build_vespa_filters(filters)
-        assert f"!({HIDDEN}=true) and " == result
-
-        # project_id with user_file_ids → both OR'd
-        id1 = UUID("00000000-0000-0000-0000-000000000123")
-        filters = IndexFilters(
-            access_control_list=[], project_id=789, user_file_ids=[id1]
-        )
-        result = build_vespa_filters(filters)
-        assert (
-            f'!({HIDDEN}=true) and (({DOCUMENT_ID} contains "{str(id1)}") or ({USER_PROJECT} contains "789")) and '
-            == result
-        )
+        assert f'!({HIDDEN}=true) and ({USER_PROJECT} contains "789") and ' == result

        # No project id
        filters = IndexFilters(access_control_list=[], project_id=None)
@@ -231,11 +217,7 @@ class TestBuildVespaFilters:
        )

    def test_combined_filters(self) -> None:
-        """Test combining multiple filter types.
-
-        Knowledge-scope filters (document_set, user_file_ids, project_id,
-        persona_id) are OR'd together, while all other filters are AND'd.
-        """
+        """Test combining multiple filter types."""
        id1 = UUID("00000000-0000-0000-0000-000000000123")
        filters = IndexFilters(
            access_control_list=["user1", "group1"],
@@ -249,6 +231,7 @@ class TestBuildVespaFilters:

        result = build_vespa_filters(filters)

+        # Build expected result piece by piece for readability
        expected = f"!({HIDDEN}=true) and "
        expected += (
            '(access_control_list contains "user1" or '
@@ -256,13 +239,9 @@ class TestBuildVespaFilters:
        )
        expected += f'({SOURCE_TYPE} contains "web") and '
        expected += f'({METADATA_LIST} contains "color{INDEX_SEPARATOR}red") and '
-        # Knowledge scope filters are OR'd together
-        expected += (
-            f'(({DOCUMENT_SETS} contains "set1")'
-            f' or ({DOCUMENT_ID} contains "{str(id1)}")'
-            f' or ({USER_PROJECT} contains "789")'
-            f") and "
-        )
+        expected += f'({DOCUMENT_SETS} contains "set1") and '
+        expected += f'({DOCUMENT_ID} contains "{str(id1)}") and '
+        expected += f'({USER_PROJECT} contains "789") and '
        cutoff_secs = int(datetime(2023, 1, 1, tzinfo=timezone.utc).timestamp())
        expected += f"!({DOC_UPDATED_AT} < {cutoff_secs}) and "

@@ -272,32 +251,6 @@ class TestBuildVespaFilters:
        result_no_trailing = build_vespa_filters(filters, remove_trailing_and=True)
        assert expected[:-5] == result_no_trailing  # Remove trailing " and "

-    def test_knowledge_scope_single_filter_not_wrapped(self) -> None:
-        """When only one knowledge-scope filter is present it should not
-        be wrapped in an extra OR group."""
-        filters = IndexFilters(access_control_list=[], document_set=["set1"])
-        result = build_vespa_filters(filters)
-        assert f'!({HIDDEN}=true) and ({DOCUMENT_SETS} contains "set1") and ' == result
-
-    def test_knowledge_scope_document_set_and_user_files_ored(self) -> None:
-        """Document set filter and user file IDs must be OR'd so that
-        connector documents (in the set) and user files (with specific
-        IDs) can both be found."""
-        id1 = UUID("00000000-0000-0000-0000-000000000123")
-        filters = IndexFilters(
-            access_control_list=[],
-            document_set=["engineering"],
-            user_file_ids=[id1],
-        )
-        result = build_vespa_filters(filters)
-        expected = (
-            f"!({HIDDEN}=true) and "
-            f'(({DOCUMENT_SETS} contains "engineering")'
-            f' or ({DOCUMENT_ID} contains "{str(id1)}")'
-            f") and "
-        )
-        assert expected == result
-
    def test_empty_or_none_values(self) -> None:
        """Test with empty or None values in filter lists."""
        # Empty strings in document set
--- a/cli/README.md
+++ b/cli/README.md
@@ -1,8 +1,5 @@
 # Onyx CLI

-[![Release CLI](https://github.com/onyx-dot-app/onyx/actions/workflows/release-cli.yml/badge.svg)](https://github.com/onyx-dot-app/onyx/actions/workflows/release-cli.yml)
-[![PyPI](https://img.shields.io/pypi/v/onyx-cli.svg)](https://pypi.org/project/onyx-cli/)
-
 A terminal interface for chatting with your [Onyx](https://github.com/onyx-dot-app/onyx) agent. Built with Go using [Bubble Tea](https://github.com/charmbracelet/bubbletea) for the TUI framework.

 ## Installation
@@ -31,7 +28,7 @@ Environment variables override config file values:

 | Variable | Required | Description |
 |----------|----------|-------------|
-| `ONYX_SERVER_URL` | No | Server base URL (default: `https://cloud.onyx.app`) |
+| `ONYX_SERVER_URL` | No | Server base URL (default: `http://localhost:3000`) |
 | `ONYX_API_KEY` | Yes | API key for authentication |
 | `ONYX_PERSONA_ID` | No | Default agent/persona ID |

@@ -71,17 +68,17 @@ onyx-cli agents --json
 | `ask` | Ask a one-shot question (non-interactive) |
 | `agents` | List available agents |
 | `configure` | Configure server URL and API key |
-| `validate-config` | Validate configuration and test connection |

 ## Slash Commands (in TUI)

 | Command | Description |
 |---------|-------------|
 | `/help` | Show help message |
-| `/clear` | Clear chat and start a new session |
+| `/new` | Start a new chat session |
 | `/agent` | List and switch agents |
 | `/attach <path>` | Attach a file to next message |
 | `/sessions` | List recent chat sessions |
+| `/clear` | Clear the chat display |
 | `/configure` | Re-run connection setup |
 | `/connectors` | Open connectors in browser |
 | `/settings` | Open settings in browser |
@@ -119,43 +116,3 @@ go build -o onyx-cli .
 # Lint
 staticcheck ./...
 ```
-
-## Publishing to PyPI
-
-The CLI is distributed as a Python package via [PyPI](https://pypi.org/project/onyx-cli/). The build system uses [hatchling](https://hatch.pypa.io/) with [manygo](https://github.com/nicholasgasior/manygo) to cross-compile Go binaries into platform-specific wheels.
-
-### CI release (recommended)
-
-Tag a release and push — the `release-cli.yml` workflow builds wheels for all platforms and publishes to PyPI automatically:
-
-```shell
-tag --prefix cli
-```
-
-To do this manually:
-
-```shell
-git tag cli/v0.1.0
-git push origin cli/v0.1.0
-```
-
-The workflow builds wheels for: linux/amd64, linux/arm64, darwin/amd64, darwin/arm64, windows/amd64, windows/arm64.
-
-### Manual release
-
-Build a wheel locally with `uv`. Set `GOOS` and `GOARCH` to cross-compile for other platforms (Go handles this natively — no cross-compiler needed):
-
-```shell
-# Build for current platform
-uv build --wheel
-
-# Cross-compile for a different platform
-GOOS=linux GOARCH=amd64 uv build --wheel
-
-# Upload to PyPI
-uv publish
-```
-
-### Versioning
-
-Versions are derived from git tags with the `cli/` prefix (e.g. `cli/v0.1.0`). The tag is parsed by `internal/_version.py` and injected into the Go binary via `-ldflags` at build time.
--- a/cli/hatch_build.py
+++ b/cli/hatch_build.py
@@ -1,43 +0,0 @@
-from __future__ import annotations
-
-import os
-import subprocess
-from typing import Any
-
-import manygo
-from hatchling.builders.hooks.plugin.interface import BuildHookInterface
-
-
-class CustomBuildHook(BuildHookInterface):
-    """Build hook to compile the Go binary and include it in the wheel."""
-
-    def initialize(self, version: Any, build_data: Any) -> None:  # noqa: ARG002
-        """Build the Go binary before packaging."""
-        build_data["pure_python"] = False
-
-        # Set platform tag for cross-compilation
-        goos = os.getenv("GOOS")
-        goarch = os.getenv("GOARCH")
-        if manygo.is_goos(goos) and manygo.is_goarch(goarch):
-            build_data["tag"] = "py3-none-" + manygo.get_platform_tag(
-                goos=goos,
-                goarch=goarch,
-            )
-
-        # Get config and environment
-        binary_name = self.config["binary_name"]
-        tag_prefix = self.config.get("tag_prefix", binary_name)
-        tag = os.getenv("GITHUB_REF_NAME", "dev").removeprefix(f"{tag_prefix}/")
-        commit = os.getenv("GITHUB_SHA", "none")
-
-        # Build the Go binary if it doesn't exist
-        # Build the Go binary (always rebuild to ensure correct version injection)
-        if not os.path.exists(binary_name):
-            print(f"Building Go binary '{binary_name}'...")
-            pkg = "github.com/onyx-dot-app/onyx/cli/cmd"
-            ldflags = f"-X {pkg}.version={tag}" f" -X {pkg}.commit={commit}" " -s -w"
-            subprocess.check_call(  # noqa: S603
-                ["go", "build", f"-ldflags={ldflags}", "-o", binary_name],
-            )
-
-        build_data["shared_scripts"] = {binary_name: binary_name}
--- a/cli/internal/_version.py
+++ b/cli/internal/_version.py
@@ -1,11 +0,0 @@
-from __future__ import annotations
-
-import os
-import re
-
-# Must match tag_prefix in pyproject.toml [tool.hatch.build.targets.wheel.hooks.custom]
-TAG_PREFIX = "cli"
-
-_tag = os.environ.get("GITHUB_REF_NAME", "v0.0.0-dev").removeprefix(f"{TAG_PREFIX}/")
-_match = re.search(r"v?(\d+\.\d+\.\d+)", _tag)
-__version__ = _match.group(1) if _match else "0.0.0"
--- a/cli/pyproject.toml
+++ b/cli/pyproject.toml
@@ -1,39 +0,0 @@
-[build-system]
-requires = ["hatchling", "go-bin~=1.24.11", "manygo"]
-build-backend = "hatchling.build"
-
-[project]
-name = "onyx-cli"
-readme = "README.md"
-description = "Terminal interface for chatting with your Onyx agent"
-authors = [{ name = "Onyx AI", email = "founders@onyx.app" }]
-requires-python = ">=3.9"
-keywords = [
-  "onyx", "cli", "chat", "ai", "enterprise-search",
-]
-classifiers = [
-    "Programming Language :: Go",
-    "License :: OSI Approved :: MIT License",
-    "Operating System :: POSIX :: Linux",
-    "Operating System :: MacOS",
-    "Operating System :: Microsoft :: Windows",
-]
-dynamic = ["version"]
-
-[project.urls]
-Repository = "https://github.com/onyx-dot-app/onyx"
-
-[tool.hatch.build]
-include = ["go.mod", "go.sum", "main.go", "**/*.go", "**/*.py", "README.md"]
-
-[tool.hatch.version]
-source = "code"
-path = "internal/_version.py"
-
-[tool.hatch.build.targets.wheel.hooks.custom]
-path = "hatch_build.py"
-binary_name = "onyx-cli"
-tag_prefix = "cli"
-
-[tool.uv]
-managed = false
--- a/desktop/package-lock.json
+++ b/desktop/package-lock.json
@@ -8,16 +8,16 @@
      "name": "onyx-desktop",
      "version": "0.0.0-dev",
      "dependencies": {
-        "@tauri-apps/api": "^2.10.1"
+        "@tauri-apps/api": "^2.0.0"
      },
      "devDependencies": {
-        "@tauri-apps/cli": "^2.10.1"
+        "@tauri-apps/cli": "^2.0.0"
      }
    },
    "node_modules/@tauri-apps/api": {
-      "version": "2.10.1",
-      "resolved": "https://registry.npmjs.org/@tauri-apps/api/-/api-2.10.1.tgz",
-      "integrity": "sha512-hKL/jWf293UDSUN09rR69hrToyIXBb8CjGaWC7gfinvnQrBVvnLr08FeFi38gxtugAVyVcTa5/FD/Xnkb1siBw==",
+      "version": "2.9.1",
+      "resolved": "https://registry.npmjs.org/@tauri-apps/api/-/api-2.9.1.tgz",
+      "integrity": "sha512-IGlhP6EivjXHepbBic618GOmiWe4URJiIeZFlB7x3czM0yDHHYviH1Xvoiv4FefdkQtn6v7TuwWCRfOGdnVUGw==",
      "license": "Apache-2.0 OR MIT",
      "funding": {
        "type": "opencollective",
@@ -25,9 +25,9 @@
      }
    },
    "node_modules/@tauri-apps/cli": {
-      "version": "2.10.1",
-      "resolved": "https://registry.npmjs.org/@tauri-apps/cli/-/cli-2.10.1.tgz",
-      "integrity": "sha512-jQNGF/5quwORdZSSLtTluyKQ+o6SMa/AUICfhf4egCGFdMHqWssApVgYSbg+jmrZoc8e1DscNvjTnXtlHLS11g==",
+      "version": "2.9.6",
+      "resolved": "https://registry.npmjs.org/@tauri-apps/cli/-/cli-2.9.6.tgz",
+      "integrity": "sha512-3xDdXL5omQ3sPfBfdC8fCtDKcnyV7OqyzQgfyT5P3+zY6lcPqIYKQBvUasNvppi21RSdfhy44ttvJmftb0PCDw==",
      "dev": true,
      "license": "Apache-2.0 OR MIT",
      "bin": {
@@ -41,23 +41,23 @@
        "url": "https://opencollective.com/tauri"
      },
      "optionalDependencies": {
-        "@tauri-apps/cli-darwin-arm64": "2.10.1",
-        "@tauri-apps/cli-darwin-x64": "2.10.1",
-        "@tauri-apps/cli-linux-arm-gnueabihf": "2.10.1",
-        "@tauri-apps/cli-linux-arm64-gnu": "2.10.1",
-        "@tauri-apps/cli-linux-arm64-musl": "2.10.1",
-        "@tauri-apps/cli-linux-riscv64-gnu": "2.10.1",
-        "@tauri-apps/cli-linux-x64-gnu": "2.10.1",
-        "@tauri-apps/cli-linux-x64-musl": "2.10.1",
-        "@tauri-apps/cli-win32-arm64-msvc": "2.10.1",
-        "@tauri-apps/cli-win32-ia32-msvc": "2.10.1",
-        "@tauri-apps/cli-win32-x64-msvc": "2.10.1"
+        "@tauri-apps/cli-darwin-arm64": "2.9.6",
+        "@tauri-apps/cli-darwin-x64": "2.9.6",
+        "@tauri-apps/cli-linux-arm-gnueabihf": "2.9.6",
+        "@tauri-apps/cli-linux-arm64-gnu": "2.9.6",
+        "@tauri-apps/cli-linux-arm64-musl": "2.9.6",
+        "@tauri-apps/cli-linux-riscv64-gnu": "2.9.6",
+        "@tauri-apps/cli-linux-x64-gnu": "2.9.6",
+        "@tauri-apps/cli-linux-x64-musl": "2.9.6",
+        "@tauri-apps/cli-win32-arm64-msvc": "2.9.6",
+        "@tauri-apps/cli-win32-ia32-msvc": "2.9.6",
+        "@tauri-apps/cli-win32-x64-msvc": "2.9.6"
      }
    },
    "node_modules/@tauri-apps/cli-darwin-arm64": {
-      "version": "2.10.1",
-      "resolved": "https://registry.npmjs.org/@tauri-apps/cli-darwin-arm64/-/cli-darwin-arm64-2.10.1.tgz",
-      "integrity": "sha512-Z2OjCXiZ+fbYZy7PmP3WRnOpM9+Fy+oonKDEmUE6MwN4IGaYqgceTjwHucc/kEEYZos5GICve35f7ZiizgqEnQ==",
+      "version": "2.9.6",
+      "resolved": "https://registry.npmjs.org/@tauri-apps/cli-darwin-arm64/-/cli-darwin-arm64-2.9.6.tgz",
+      "integrity": "sha512-gf5no6N9FCk1qMrti4lfwP77JHP5haASZgVbBgpZG7BUepB3fhiLCXGUK8LvuOjP36HivXewjg72LTnPDScnQQ==",
      "cpu": [
        "arm64"
      ],
@@ -72,9 +72,9 @@
      }
    },
    "node_modules/@tauri-apps/cli-darwin-x64": {
-      "version": "2.10.1",
-      "resolved": "https://registry.npmjs.org/@tauri-apps/cli-darwin-x64/-/cli-darwin-x64-2.10.1.tgz",
-      "integrity": "sha512-V/irQVvjPMGOTQqNj55PnQPVuH4VJP8vZCN7ajnj+ZS8Kom1tEM2hR3qbbIRoS3dBKs5mbG8yg1WC+97dq17Pw==",
+      "version": "2.9.6",
+      "resolved": "https://registry.npmjs.org/@tauri-apps/cli-darwin-x64/-/cli-darwin-x64-2.9.6.tgz",
+      "integrity": "sha512-oWh74WmqbERwwrwcueJyY6HYhgCksUc6NT7WKeXyrlY/FPmNgdyQAgcLuTSkhRFuQ6zh4Np1HZpOqCTpeZBDcw==",
      "cpu": [
        "x64"
      ],
@@ -89,9 +89,9 @@
      }
    },
    "node_modules/@tauri-apps/cli-linux-arm-gnueabihf": {
-      "version": "2.10.1",
-      "resolved": "https://registry.npmjs.org/@tauri-apps/cli-linux-arm-gnueabihf/-/cli-linux-arm-gnueabihf-2.10.1.tgz",
-      "integrity": "sha512-Hyzwsb4VnCWKGfTw+wSt15Z2pLw2f0JdFBfq2vHBOBhvg7oi6uhKiF87hmbXOBXUZaGkyRDkCHsdzJcIfoJC2w==",
+      "version": "2.9.6",
+      "resolved": "https://registry.npmjs.org/@tauri-apps/cli-linux-arm-gnueabihf/-/cli-linux-arm-gnueabihf-2.9.6.tgz",
+      "integrity": "sha512-/zde3bFroFsNXOHN204DC2qUxAcAanUjVXXSdEGmhwMUZeAQalNj5cz2Qli2elsRjKN/hVbZOJj0gQ5zaYUjSg==",
      "cpu": [
        "arm"
      ],
@@ -106,9 +106,9 @@
      }
    },
    "node_modules/@tauri-apps/cli-linux-arm64-gnu": {
-      "version": "2.10.1",
-      "resolved": "https://registry.npmjs.org/@tauri-apps/cli-linux-arm64-gnu/-/cli-linux-arm64-gnu-2.10.1.tgz",
-      "integrity": "sha512-OyOYs2t5GkBIvyWjA1+h4CZxTcdz1OZPCWAPz5DYEfB0cnWHERTnQ/SLayQzncrT0kwRoSfSz9KxenkyJoTelA==",
+      "version": "2.9.6",
+      "resolved": "https://registry.npmjs.org/@tauri-apps/cli-linux-arm64-gnu/-/cli-linux-arm64-gnu-2.9.6.tgz",
+      "integrity": "sha512-pvbljdhp9VOo4RnID5ywSxgBs7qiylTPlK56cTk7InR3kYSTJKYMqv/4Q/4rGo/mG8cVppesKIeBMH42fw6wjg==",
      "cpu": [
        "arm64"
      ],
@@ -123,9 +123,9 @@
      }
    },
    "node_modules/@tauri-apps/cli-linux-arm64-musl": {
-      "version": "2.10.1",
-      "resolved": "https://registry.npmjs.org/@tauri-apps/cli-linux-arm64-musl/-/cli-linux-arm64-musl-2.10.1.tgz",
-      "integrity": "sha512-MIj78PDDGjkg3NqGptDOGgfXks7SYJwhiMh8SBoZS+vfdz7yP5jN18bNaLnDhsVIPARcAhE1TlsZe/8Yxo2zqg==",
+      "version": "2.9.6",
+      "resolved": "https://registry.npmjs.org/@tauri-apps/cli-linux-arm64-musl/-/cli-linux-arm64-musl-2.9.6.tgz",
+      "integrity": "sha512-02TKUndpodXBCR0oP//6dZWGYcc22Upf2eP27NvC6z0DIqvkBBFziQUcvi2n6SrwTRL0yGgQjkm9K5NIn8s6jw==",
      "cpu": [
        "arm64"
      ],
@@ -140,9 +140,9 @@
      }
    },
    "node_modules/@tauri-apps/cli-linux-riscv64-gnu": {
-      "version": "2.10.1",
-      "resolved": "https://registry.npmjs.org/@tauri-apps/cli-linux-riscv64-gnu/-/cli-linux-riscv64-gnu-2.10.1.tgz",
-      "integrity": "sha512-X0lvOVUg8PCVaoEtEAnpxmnkwlE1gcMDTqfhbefICKDnOTJ5Est3qL0SrWxizDackIOKBcvtpejrSiVpuJI1kw==",
+      "version": "2.9.6",
+      "resolved": "https://registry.npmjs.org/@tauri-apps/cli-linux-riscv64-gnu/-/cli-linux-riscv64-gnu-2.9.6.tgz",
+      "integrity": "sha512-fmp1hnulbqzl1GkXl4aTX9fV+ubHw2LqlLH1PE3BxZ11EQk+l/TmiEongjnxF0ie4kV8DQfDNJ1KGiIdWe1GvQ==",
      "cpu": [
        "riscv64"
      ],
@@ -157,9 +157,9 @@
      }
    },
    "node_modules/@tauri-apps/cli-linux-x64-gnu": {
-      "version": "2.10.1",
-      "resolved": "https://registry.npmjs.org/@tauri-apps/cli-linux-x64-gnu/-/cli-linux-x64-gnu-2.10.1.tgz",
-      "integrity": "sha512-2/12bEzsJS9fAKybxgicCDFxYD1WEI9kO+tlDwX5znWG2GwMBaiWcmhGlZ8fi+DMe9CXlcVarMTYc0L3REIRxw==",
+      "version": "2.9.6",
+      "resolved": "https://registry.npmjs.org/@tauri-apps/cli-linux-x64-gnu/-/cli-linux-x64-gnu-2.9.6.tgz",
+      "integrity": "sha512-vY0le8ad2KaV1PJr+jCd8fUF9VOjwwQP/uBuTJvhvKTloEwxYA/kAjKK9OpIslGA9m/zcnSo74czI6bBrm2sYA==",
      "cpu": [
        "x64"
      ],
@@ -174,9 +174,9 @@
      }
    },
    "node_modules/@tauri-apps/cli-linux-x64-musl": {
-      "version": "2.10.1",
-      "resolved": "https://registry.npmjs.org/@tauri-apps/cli-linux-x64-musl/-/cli-linux-x64-musl-2.10.1.tgz",
-      "integrity": "sha512-Y8J0ZzswPz50UcGOFuXGEMrxbjwKSPgXftx5qnkuMs2rmwQB5ssvLb6tn54wDSYxe7S6vlLob9vt0VKuNOaCIQ==",
+      "version": "2.9.6",
+      "resolved": "https://registry.npmjs.org/@tauri-apps/cli-linux-x64-musl/-/cli-linux-x64-musl-2.9.6.tgz",
+      "integrity": "sha512-TOEuB8YCFZTWVDzsO2yW0+zGcoMiPPwcUgdnW1ODnmgfwccpnihDRoks+ABT1e3fHb1ol8QQWsHSCovb3o2ENQ==",
      "cpu": [
        "x64"
      ],
@@ -191,9 +191,9 @@
      }
    },
    "node_modules/@tauri-apps/cli-win32-arm64-msvc": {
-      "version": "2.10.1",
-      "resolved": "https://registry.npmjs.org/@tauri-apps/cli-win32-arm64-msvc/-/cli-win32-arm64-msvc-2.10.1.tgz",
-      "integrity": "sha512-iSt5B86jHYAPJa/IlYw++SXtFPGnWtFJriHn7X0NFBVunF6zu9+/zOn8OgqIWSl8RgzhLGXQEEtGBdR4wzpVgg==",
+      "version": "2.9.6",
+      "resolved": "https://registry.npmjs.org/@tauri-apps/cli-win32-arm64-msvc/-/cli-win32-arm64-msvc-2.9.6.tgz",
+      "integrity": "sha512-ujmDGMRc4qRLAnj8nNG26Rlz9klJ0I0jmZs2BPpmNNf0gM/rcVHhqbEkAaHPTBVIrtUdf7bGvQAD2pyIiUrBHQ==",
      "cpu": [
        "arm64"
      ],
@@ -208,9 +208,9 @@
      }
    },
    "node_modules/@tauri-apps/cli-win32-ia32-msvc": {
-      "version": "2.10.1",
-      "resolved": "https://registry.npmjs.org/@tauri-apps/cli-win32-ia32-msvc/-/cli-win32-ia32-msvc-2.10.1.tgz",
-      "integrity": "sha512-gXyxgEzsFegmnWywYU5pEBURkcFN/Oo45EAwvZrHMh+zUSEAvO5E8TXsgPADYm31d1u7OQU3O3HsYfVBf2moHw==",
+      "version": "2.9.6",
+      "resolved": "https://registry.npmjs.org/@tauri-apps/cli-win32-ia32-msvc/-/cli-win32-ia32-msvc-2.9.6.tgz",
+      "integrity": "sha512-S4pT0yAJgFX8QRCyKA1iKjZ9Q/oPjCZf66A/VlG5Yw54Nnr88J1uBpmenINbXxzyhduWrIXBaUbEY1K80ZbpMg==",
      "cpu": [
        "ia32"
      ],
@@ -225,9 +225,9 @@
      }
    },
    "node_modules/@tauri-apps/cli-win32-x64-msvc": {
-      "version": "2.10.1",
-      "resolved": "https://registry.npmjs.org/@tauri-apps/cli-win32-x64-msvc/-/cli-win32-x64-msvc-2.10.1.tgz",
-      "integrity": "sha512-6Cn7YpPFwzChy0ERz6djKEmUehWrYlM+xTaNzGPgZocw3BD7OfwfWHKVWxXzdjEW2KfKkHddfdxK1XXTYqBRLg==",
+      "version": "2.9.6",
+      "resolved": "https://registry.npmjs.org/@tauri-apps/cli-win32-x64-msvc/-/cli-win32-x64-msvc-2.9.6.tgz",
+      "integrity": "sha512-ldWuWSSkWbKOPjQMJoYVj9wLHcOniv7diyI5UAJ4XsBdtaFB0pKHQsqw/ItUma0VXGC7vB4E9fZjivmxur60aw==",
      "cpu": [
        "x64"
      ],
--- a/desktop/package.json
+++ b/desktop/package.json
@@ -9,10 +9,10 @@
    "build:dmg": "tauri build --target universal-apple-darwin",
    "build:linux": "tauri build --bundles deb,rpm"
  },
-  "dependencies": {
-    "@tauri-apps/api": "^2.10.1"
-  },
  "devDependencies": {
-    "@tauri-apps/cli": "^2.10.1"
+    "@tauri-apps/cli": "^2.0.0"
+  },
+  "dependencies": {
+    "@tauri-apps/api": "^2.0.0"
  }
 }
--- a/desktop/src-tauri/Cargo.lock
+++ b/desktop/src-tauri/Cargo.lock
--- a/desktop/src-tauri/Cargo.toml
+++ b/desktop/src-tauri/Cargo.toml
@@ -6,18 +6,18 @@ authors = ["you"]
 edition = "2021"

 [build-dependencies]
-tauri-build = { version = "2.5", features = [] }
+tauri-build = { version = "2.0", features = [] }

 [dependencies]
-tauri = { version = "2.10", features = ["macos-private-api", "tray-icon", "image-png"] }
-tauri-plugin-shell = "2.3.5"
-tauri-plugin-window-state = "2.4.1"
+tauri = { version = "2.0", features = ["macos-private-api", "tray-icon", "image-png"] }
+tauri-plugin-shell = "2.0"
+tauri-plugin-window-state = "2.0"
 serde = { version = "1.0", features = ["derive"] }
 serde_json = "1.0"
 uuid = { version = "1.0", features = ["v4"] }
 directories = "5.0"
 tokio = { version = "1", features = ["time"] }
-window-vibrancy = "0.7.1"
+window-vibrancy = "0.5"
 url = "2.5"

 [features]
--- a/tools/ods/hatch_build.py
+++ b/tools/ods/hatch_build.py
@@ -26,16 +26,20 @@ class CustomBuildHook(BuildHookInterface):

        # Get config and environment
        binary_name = self.config["binary_name"]
-        tag_prefix = self.config.get("tag_prefix", binary_name)
-        tag = os.getenv("GITHUB_REF_NAME", "dev").removeprefix(f"{tag_prefix}/")
+        tag = os.getenv("GITHUB_REF_NAME", "dev").removeprefix(f"{binary_name}/")
        commit = os.getenv("GITHUB_SHA", "none")

        # Build the Go binary if it doesn't exist
        if not os.path.exists(binary_name):
            print(f"Building Go binary '{binary_name}'...")
-            ldflags = f"-X main.version={tag} -X main.commit={commit} -s -w"
            subprocess.check_call(  # noqa: S603
-                ["go", "build", f"-ldflags={ldflags}", "-o", binary_name],
+                [
+                    "go",
+                    "build",
+                    f"-ldflags=-X main.version={tag} -X main.commit={commit} -s -w",
+                    "-o",
+                    binary_name,
+                ],
            )

        build_data["shared_scripts"] = {binary_name: binary_name}
--- a/tools/ods/internal/_version.py
+++ b/tools/ods/internal/_version.py
@@ -3,9 +3,6 @@ from __future__ import annotations
 import os
 import re

-# Must match tag_prefix in pyproject.toml [tool.hatch.build.targets.wheel.hooks.custom]
-TAG_PREFIX: str = "ods"
-
-_tag = os.environ.get("GITHUB_REF_NAME", "v0.0.0-dev").removeprefix(f"{TAG_PREFIX}/")
+_tag = os.environ.get("GITHUB_REF_NAME", "v0.0.0-dev").removeprefix("ods/")
 _match = re.search(r"v?(\d+\.\d+\.\d+)", _tag)
 __version__ = _match.group(1) if _match else "0.0.0"
--- a/tools/ods/pyproject.toml
+++ b/tools/ods/pyproject.toml
@@ -14,9 +14,7 @@ keywords = [
 classifiers = [
    "Programming Language :: Go",
    "License :: OSI Approved :: MIT License",
-    "Operating System :: POSIX :: Linux",
-    "Operating System :: MacOS",
-    "Operating System :: Microsoft :: Windows",
+    "Operating System :: OS Independent",
 ]
 dynamic = ["version"]
 dependencies = [
@@ -29,7 +27,7 @@ dependencies = [
 Repository = "https://github.com/onyx-dot-app/onyx"

 [tool.hatch.build]
-include = ["go.mod", "go.sum", "main.go", "**/*.go", "**/*.py", "README.md"]
+include = ["go.mod", "go.sum", "main.go", "**/*.go", "**/*.py"]

 [tool.hatch.version]
 source = "code"
@@ -38,7 +36,6 @@ path = "internal/_version.py"
 [tool.hatch.build.targets.wheel.hooks.custom]
 path = "hatch_build.py"
 binary_name = "ods"
-tag_prefix = "ods"

 [tool.uv]
 managed = false
--- a/web/package-lock.json
+++ b/web/package-lock.json
--- a/web/package.json
+++ b/web/package.json
@@ -132,7 +132,7 @@
    "eslint-plugin-unused-imports": "^4.1.4",
    "identity-obj-proxy": "^3.0.0",
    "jest": "^29.7.0",
-    "jest-environment-jsdom": "^30.2.0",
+    "jest-environment-jsdom": "^29.7.0",
    "prettier": "3.1.0",
    "stats.js": "^0.17.0",
    "tailwindcss": "^3.4.17",
--- a/web/src/app/admin/api-key/OnyxApiKeyForm.tsx
+++ b/web/src/app/admin/api-key/OnyxApiKeyForm.tsx
@@ -6,7 +6,7 @@ import { Button } from "@opal/components";
 import { Disabled } from "@opal/core";
 import Text from "@/refresh-components/texts/Text";
 import InputTypeIn from "@/refresh-components/inputs/InputTypeIn";
-import InputSelect from "@/refresh-components/inputs/InputSelect";
+import InputComboBox from "@/refresh-components/inputs/InputComboBox";
 import { FormikField } from "@/refresh-components/form/FormikField";
 import { FormField } from "@/refresh-components/form/FormField";
 import { USER_ROLE_LABELS, UserRole } from "@/lib/types";
@@ -107,25 +107,26 @@ export default function OnyxApiKeyForm({
                    <FormField name="role" state={state} className="w-full">
                      <FormField.Label>Role:</FormField.Label>
                      <FormField.Control>
-                        <InputSelect
+                        <InputComboBox
                          value={field.value}
                          onValueChange={(value) => helper.setValue(value)}
-                        >
-                          <InputSelect.Trigger placeholder="Select a role" />
-                          <InputSelect.Content>
-                            <InputSelect.Item
-                              value={UserRole.LIMITED.toString()}
-                            >
-                              {USER_ROLE_LABELS[UserRole.LIMITED]}
-                            </InputSelect.Item>
-                            <InputSelect.Item value={UserRole.BASIC.toString()}>
-                              {USER_ROLE_LABELS[UserRole.BASIC]}
-                            </InputSelect.Item>
-                            <InputSelect.Item value={UserRole.ADMIN.toString()}>
-                              {USER_ROLE_LABELS[UserRole.ADMIN]}
-                            </InputSelect.Item>
-                          </InputSelect.Content>
-                        </InputSelect>
+                          options={[
+                            {
+                              label: USER_ROLE_LABELS[UserRole.LIMITED],
+                              value: UserRole.LIMITED.toString(),
+                            },
+                            {
+                              label: USER_ROLE_LABELS[UserRole.BASIC],
+                              value: UserRole.BASIC.toString(),
+                            },
+                            {
+                              label: USER_ROLE_LABELS[UserRole.ADMIN],
+                              value: UserRole.ADMIN.toString(),
+                            },
+                          ]}
+                          placeholder="Select a role"
+                          strict
+                        />
                      </FormField.Control>
                      <FormField.Description>
                        Select the role for this API key. Limited has access to
--- a/web/src/app/app/message/FileDisplay.tsx
+++ b/web/src/app/app/message/FileDisplay.tsx
@@ -7,13 +7,15 @@ import { InMessageImage } from "@/app/app/components/files/images/InMessageImage
 import CsvContent from "@/components/tools/CSVContent";
 import TextViewModal from "@/sections/modals/TextViewModal";
 import { MinimalOnyxDocument } from "@/lib/search/interfaces";
+import { cn } from "@/lib/utils";
 import ExpandableContentWrapper from "@/components/tools/ExpandableContentWrapper";

 interface FileDisplayProps {
  files: FileDescriptor[];
+  alignBubble?: boolean;
 }

-export default function FileDisplay({ files }: FileDisplayProps) {
+export default function FileDisplay({ files, alignBubble }: FileDisplayProps) {
  const [close, setClose] = useState(true);
  const [previewingFile, setPreviewingFile] = useState<FileDescriptor | null>(
    null
@@ -41,47 +43,59 @@ export default function FileDisplay({ files }: FileDisplayProps) {
      )}

      {textFiles.length > 0 && (
-        <div id="onyx-file" className="flex flex-col items-end gap-2 py-2">
-          {textFiles.map((file) => (
-            <Attachment
-              key={file.id}
-              fileName={file.name || file.id}
-              open={() => setPreviewingFile(file)}
-            />
-          ))}
+        <div
+          id="onyx-file"
+          className={cn("m-2 auto", alignBubble && "ml-auto")}
+        >
+          <div className="flex flex-col items-end gap-2">
+            {textFiles.map((file) => (
+              <Attachment
+                key={file.id}
+                fileName={file.name || file.id}
+                open={() => setPreviewingFile(file)}
+              />
+            ))}
+          </div>
        </div>
      )}

      {imageFiles.length > 0 && (
-        <div id="onyx-image" className="flex flex-col items-end gap-2 py-2">
-          {imageFiles.map((file) => (
-            <InMessageImage key={file.id} fileId={file.id} />
-          ))}
+        <div
+          id="onyx-image"
+          className={cn("m-2 auto", alignBubble && "ml-auto")}
+        >
+          <div className="flex flex-col items-end gap-2">
+            {imageFiles.map((file) => (
+              <InMessageImage key={file.id} fileId={file.id} />
+            ))}
+          </div>
        </div>
      )}

      {csvFiles.length > 0 && (
-        <div className="flex flex-col items-end gap-2 py-2">
-          {csvFiles.map((file) => {
-            return (
-              <div key={file.id} className="w-fit">
-                {close ? (
-                  <>
-                    <ExpandableContentWrapper
-                      fileDescriptor={file}
-                      close={() => setClose(false)}
-                      ContentComponent={CsvContent}
+        <div className={cn("m-2 auto", alignBubble && "ml-auto")}>
+          <div className="flex flex-col items-end gap-2">
+            {csvFiles.map((file) => {
+              return (
+                <div key={file.id} className="w-fit">
+                  {close ? (
+                    <>
+                      <ExpandableContentWrapper
+                        fileDescriptor={file}
+                        close={() => setClose(false)}
+                        ContentComponent={CsvContent}
+                      />
+                    </>
+                  ) : (
+                    <Attachment
+                      open={() => setClose(true)}
+                      fileName={file.name || file.id}
                    />
-                  </>
-                ) : (
-                  <Attachment
-                    open={() => setClose(true)}
-                    fileName={file.name || file.id}
-                  />
-                )}
-              </div>
-            );
-          })}
+                  )}
+                </div>
+              );
+            })}
+          </div>
        </div>
      )}
    </>
--- a/web/src/app/app/message/HumanMessage.tsx
+++ b/web/src/app/app/message/HumanMessage.tsx
@@ -195,7 +195,7 @@ const HumanMessage = React.memo(function HumanMessage({
      id="onyx-human-message"
      className="group flex flex-col justify-end w-full relative"
    >
-      <FileDisplay files={files || []} />
+      <FileDisplay alignBubble files={files || []} />
      {isEditing ? (
        <MessageEditing
          content={content}
--- a/web/src/app/auth/login/EmailPasswordForm.test.tsx
+++ b/web/src/app/auth/login/EmailPasswordForm.test.tsx
@@ -22,6 +22,9 @@ describe("Email/Password Login Workflow", () => {
  beforeEach(() => {
    jest.clearAllMocks();
    fetchSpy = jest.spyOn(global, "fetch");
+    // Mock window.location.href for redirect testing
+    delete (window as any).location;
+    window.location = { href: "" } as any;
  });

  afterEach(() => {
@@ -50,9 +53,9 @@ describe("Email/Password Login Workflow", () => {
    const loginButton = screen.getByRole("button", { name: /sign in/i });
    await user.click(loginButton);

-    // Verify success message is shown after login
+    // After successful login, user should be redirected to /chat
    await waitFor(() => {
-      expect(screen.getByText(/signed in successfully\./i)).toBeInTheDocument();
+      expect(window.location.href).toBe("/app");
    });

    // Verify API was called with correct credentials
@@ -111,6 +114,9 @@ describe("Email/Password Signup Workflow", () => {
  beforeEach(() => {
    jest.clearAllMocks();
    fetchSpy = jest.spyOn(global, "fetch");
+    // Mock window.location.href
+    delete (window as any).location;
+    window.location = { href: "" } as any;
  });

  afterEach(() => {
--- a/web/src/app/craft/components/ToggleWarningModal.tsx
+++ b/web/src/app/craft/components/ToggleWarningModal.tsx
@@ -39,7 +39,7 @@ export function ToggleWarningModal({
          {/* Message */}
          <div className="flex justify-center">
            <Text mainUiBody text04 className="text-center">
-              We recommend using <strong>Claude Opus 4.6</strong> for Crafting.
+              We recommend using <strong>Claude Opus 4.5</strong> for Crafting.
              <br />
              Other models may have reduced capabilities for code creation,
              <br />
--- a/web/src/app/craft/onboarding/components/OnboardingLlmSetup.tsx
+++ b/web/src/app/craft/onboarding/components/OnboardingLlmSetup.tsx
@@ -34,8 +34,8 @@ export const PROVIDERS: ProviderConfig[] = [
    providerName: LLMProviderName.ANTHROPIC,
    recommended: true,
    models: [
-      { name: "claude-opus-4-6", label: "Claude Opus 4.6", recommended: true },
-      { name: "claude-sonnet-4-6", label: "Claude Sonnet 4.6" },
+      { name: "claude-opus-4-5", label: "Claude Opus 4.5", recommended: true },
+      { name: "claude-sonnet-4-5", label: "Claude Sonnet 4.5" },
    ],
    apiKeyPlaceholder: "sk-ant-...",
    apiKeyUrl: "https://console.anthropic.com/dashboard",
--- a/web/src/app/craft/onboarding/constants.ts
+++ b/web/src/app/craft/onboarding/constants.ts
@@ -5,12 +5,12 @@
 export interface BuildLlmSelection {
  providerName: string; // e.g., "build-mode-anthropic" (LLMProviderDescriptor.name)
  provider: string; // e.g., "anthropic"
-  modelName: string; // e.g., "claude-opus-4-6"
+  modelName: string; // e.g., "claude-opus-4-5"
 }

 // Priority order for smart default LLM selection
 const LLM_SELECTION_PRIORITY = [
-  { provider: "anthropic", modelName: "claude-opus-4-6" },
+  { provider: "anthropic", modelName: "claude-opus-4-5" },
  { provider: "openai", modelName: "gpt-5.2" },
  { provider: "openrouter", modelName: "minimax/minimax-m2.1" },
 ] as const;
@@ -63,11 +63,11 @@ export function getDefaultLlmSelection(
 export const RECOMMENDED_BUILD_MODELS = {
  preferred: {
    provider: "anthropic",
-    modelName: "claude-opus-4-6",
-    displayName: "Claude Opus 4.6",
+    modelName: "claude-opus-4-5",
+    displayName: "Claude Opus 4.5",
  },
  alternatives: [
-    { provider: "anthropic", modelName: "claude-sonnet-4-6" },
+    { provider: "anthropic", modelName: "claude-sonnet-4-5" },
    { provider: "openai", modelName: "gpt-5.2" },
    { provider: "openai", modelName: "gpt-5.1-codex" },
    { provider: "openrouter", modelName: "minimax/minimax-m2.1" },
@@ -148,8 +148,8 @@ export const BUILD_MODE_PROVIDERS: BuildModeProvider[] = [
    providerName: "anthropic",
    recommended: true,
    models: [
-      { name: "claude-opus-4-6", label: "Claude Opus 4.6", recommended: true },
-      { name: "claude-sonnet-4-6", label: "Claude Sonnet 4.6" },
+      { name: "claude-opus-4-5", label: "Claude Opus 4.5", recommended: true },
+      { name: "claude-sonnet-4-5", label: "Claude Sonnet 4.5" },
    ],
    apiKeyPlaceholder: "sk-ant-...",
    apiKeyUrl: "https://console.anthropic.com/dashboard",
--- a/web/src/lib/llmConfig/cache.ts
+++ b/web/src/lib/llmConfig/cache.ts
@@ -1,21 +0,0 @@
-import { ScopedMutator } from "swr";
-import {
-  LLM_CHAT_PROVIDERS_URL,
-  LLM_PROVIDERS_ADMIN_URL,
-} from "@/lib/llmConfig/constants";
-
-const PERSONA_PROVIDER_ENDPOINT_PATTERN =
-  /^\/api\/llm\/persona\/\d+\/providers$/;
-
-export async function refreshLlmProviderCaches(
-  mutate: ScopedMutator
-): Promise<void> {
-  await Promise.all([
-    mutate(LLM_PROVIDERS_ADMIN_URL),
-    mutate(LLM_CHAT_PROVIDERS_URL),
-    mutate(
-      (key) =>
-        typeof key === "string" && PERSONA_PROVIDER_ENDPOINT_PATTERN.test(key)
-    ),
-  ]);
-}
--- a/web/src/lib/llmConfig/constants.ts
+++ b/web/src/lib/llmConfig/constants.ts
@@ -1,6 +1,5 @@
 export const LLM_ADMIN_URL = "/api/admin/llm";
 export const LLM_PROVIDERS_ADMIN_URL = `${LLM_ADMIN_URL}/provider`;
-export const LLM_CHAT_PROVIDERS_URL = "/api/llm/provider";

 export const LLM_CONTEXTUAL_COST_ADMIN_URL =
  "/api/admin/llm/provider-contextual-cost";
--- a/web/src/refresh-components/buttons/LineItem.tsx
+++ b/web/src/refresh-components/buttons/LineItem.tsx
@@ -57,12 +57,6 @@ export interface LineItemProps
    WithoutStyles<React.HTMLAttributes<HTMLDivElement>>,
    "children"
  > {
-  /**
-   * Whether the row should behave like a standalone interactive button.
-   * Set to false when nested inside another interactive primitive
-   * (e.g. Radix Select.Item) to avoid nested focus targets.
-   */
-  interactive?: boolean;
  // line-item variants
  strikethrough?: boolean;
  danger?: boolean;
@@ -137,7 +131,6 @@ export interface LineItemProps
 * - The component automatically adds a `data-selected="true"` attribute for custom styling
 */
 export default function LineItem({
-  interactive = true,
  selected,
  strikethrough,
  danger,
@@ -171,11 +164,6 @@ export default function LineItem({
  const emphasisKey = emphasized ? "emphasized" : "normal";

  const handleKeyDown = (e: React.KeyboardEvent<HTMLDivElement>) => {
-    if (!interactive) {
-      props.onKeyDown?.(e);
-      return;
-    }
-
    if (e.key === "Enter") {
      e.preventDefault();
      (e.currentTarget as HTMLDivElement).click();
@@ -186,11 +174,6 @@ export default function LineItem({
  };

  const handleKeyUp = (e: React.KeyboardEvent<HTMLDivElement>) => {
-    if (!interactive) {
-      props.onKeyUp?.(e);
-      return;
-    }
-
    if (e.key === " ") {
      e.preventDefault();
      (e.currentTarget as HTMLDivElement).click();
@@ -201,8 +184,8 @@ export default function LineItem({
  const content = (
    <div
      ref={ref}
-      role={interactive ? "button" : undefined}
-      tabIndex={interactive ? 0 : undefined}
+      role="button"
+      tabIndex={0}
      className={cn(
        "flex flex-row w-full items-start p-2 rounded-08 group/LineItem gap-2",
        !!(children && description) ? "items-start" : "items-center",
--- a/web/src/refresh-components/inputs/InputSelect.tsx
+++ b/web/src/refresh-components/inputs/InputSelect.tsx
@@ -369,7 +369,7 @@ function InputSelectItem({
    <SelectPrimitive.Item
      ref={ref}
      value={value}
-      className="outline-none focus:outline-none rounded-08 data-[highlighted]:bg-background-tint-02"
+      className="outline-none focus:outline-none"
      onSelect={onClick}
    >
      {/* Hidden ItemText for Radix to track selection */}
@@ -383,7 +383,7 @@ function InputSelectItem({
        selected={isSelected}
        emphasized
        description={description}
-        interactive={false}
+        onClick={noProp((event) => event.preventDefault())}
      >
        {children}
      </LineItem>
--- a/web/src/refresh-pages/AppPage.tsx
+++ b/web/src/refresh-pages/AppPage.tsx
@@ -193,7 +193,7 @@ export default function AppPage({ firstMessage }: ChatPageProps) {
      onSubmit({
        message,
        currentMessageFiles,
-        deepResearch: deepResearchEnabledForCurrentWorkflow,
+        deepResearch: deepResearchEnabled,
      });
    }
  }
@@ -218,8 +218,6 @@ export default function AppPage({ firstMessage }: ChatPageProps) {
    chatSessionId: currentChatSessionId,
    agentId: selectedAgent?.id,
  });
-  const deepResearchEnabledForCurrentWorkflow =
-    currentProjectId === null && deepResearchEnabled;

  const [presentingDocument, setPresentingDocument] =
    useState<MinimalOnyxDocument | null>(null);
@@ -437,15 +435,10 @@ export default function AppPage({ firstMessage }: ChatPageProps) {
    onSubmit({
      message: lastUserMsg.message,
      currentMessageFiles: currentMessageFiles,
-      deepResearch: deepResearchEnabledForCurrentWorkflow,
+      deepResearch: deepResearchEnabled,
      messageIdToResend: lastUserMsg.messageId,
    });
-  }, [
-    messageHistory,
-    onSubmit,
-    currentMessageFiles,
-    deepResearchEnabledForCurrentWorkflow,
-  ]);
+  }, [messageHistory, onSubmit, currentMessageFiles, deepResearchEnabled]);

  const toggleDocumentSidebar = useCallback(() => {
    if (!documentSidebarVisible) {
@@ -465,7 +458,7 @@ export default function AppPage({ firstMessage }: ChatPageProps) {
      onSubmit({
        message,
        currentMessageFiles,
-        deepResearch: deepResearchEnabledForCurrentWorkflow,
+        deepResearch: deepResearchEnabled,
      });
      if (showOnboarding || !onboardingDismissed) {
        finishOnboarding();
@@ -475,7 +468,7 @@ export default function AppPage({ firstMessage }: ChatPageProps) {
      resetInputBar,
      onSubmit,
      currentMessageFiles,
-      deepResearchEnabledForCurrentWorkflow,
+      deepResearchEnabled,
      showOnboarding,
      onboardingDismissed,
      finishOnboarding,
@@ -510,7 +503,7 @@ export default function AppPage({ firstMessage }: ChatPageProps) {
        onSubmit({
          message,
          currentMessageFiles,
-          deepResearch: deepResearchEnabledForCurrentWorkflow,
+          deepResearch: deepResearchEnabled,
        });
        if (showOnboarding || !onboardingDismissed) {
          finishOnboarding();
@@ -531,7 +524,7 @@ export default function AppPage({ firstMessage }: ChatPageProps) {
      resetInputBar,
      onSubmit,
      currentMessageFiles,
-      deepResearchEnabledForCurrentWorkflow,
+      deepResearchEnabled,
      showOnboarding,
      onboardingDismissed,
      finishOnboarding,
@@ -716,7 +709,7 @@ export default function AppPage({ firstMessage }: ChatPageProps) {
            >
              {/* Main content grid — 3 rows, animated */}
              <div
-                className="flex-1 w-full grid min-h-0 px-4 transition-[grid-template-rows] duration-150 ease-in-out"
+                className="flex-1 w-full grid min-h-0 transition-[grid-template-rows] duration-150 ease-in-out"
                style={gridStyle}
              >
                {/* ── Top row: ChatUI / WelcomeMessage / ProjectUI ── */}
@@ -739,9 +732,7 @@ export default function AppPage({ firstMessage }: ChatPageProps) {
                      <ChatUI
                        liveAgent={liveAgent!}
                        llmManager={llmManager}
-                        deepResearchEnabled={
-                          deepResearchEnabledForCurrentWorkflow
-                        }
+                        deepResearchEnabled={deepResearchEnabled}
                        currentMessageFiles={currentMessageFiles}
                        setPresentingDocument={setPresentingDocument}
                        onSubmit={onSubmit}
@@ -837,9 +828,7 @@ export default function AppPage({ firstMessage }: ChatPageProps) {
                      />
                      <AppInputBar
                        ref={chatInputBarRef}
-                        deepResearchEnabled={
-                          deepResearchEnabledForCurrentWorkflow
-                        }
+                        deepResearchEnabled={deepResearchEnabled}
                        toggleDeepResearch={toggleDeepResearch}
                        filterManager={filterManager}
                        llmManager={llmManager}
--- a/web/src/refresh-pages/admin/LLMConfigurationPage.tsx
+++ b/web/src/refresh-pages/admin/LLMConfigurationPage.tsx
@@ -20,7 +20,6 @@ import {
  getProviderIcon,
  getProviderProductName,
 } from "@/lib/llmConfig/providers";
-import { refreshLlmProviderCaches } from "@/lib/llmConfig/cache";
 import { deleteLlmProvider, setDefaultLlmModel } from "@/lib/llmConfig/svc";
 import Text from "@/refresh-components/texts/Text";
 import { Horizontal as HorizontalInput } from "@/layouts/input-layouts";
@@ -34,6 +33,7 @@ import {
  LLMProviderView,
  WellKnownLLMProviderDescriptor,
 } from "@/interfaces/llm";
+import { LLM_PROVIDERS_ADMIN_URL } from "@/lib/llmConfig/constants";
 import { getModalForExistingProvider } from "@/sections/modals/llmConfig/getModal";
 import { OpenAIModal } from "@/sections/modals/llmConfig/OpenAIModal";
 import { AnthropicModal } from "@/sections/modals/llmConfig/AnthropicModal";
@@ -140,7 +140,7 @@ function ExistingProviderCard({
  const handleDelete = async () => {
    try {
      await deleteLlmProvider(provider.id);
-      await refreshLlmProviderCaches(mutate);
+      mutate(LLM_PROVIDERS_ADMIN_URL);
      deleteModal.toggle(false);
      toast.success("Provider deleted successfully!");
    } catch (e) {
@@ -345,7 +345,7 @@ export default function LLMConfigurationPage() {

    try {
      await setDefaultLlmModel(providerId, modelName);
-      await refreshLlmProviderCaches(mutate);
+      mutate(LLM_PROVIDERS_ADMIN_URL);
      toast.success("Default model updated successfully!");
    } catch (e) {
      const message = e instanceof Error ? e.message : "Unknown error";
--- a/web/src/sections/input/AppInputBar.tsx
+++ b/web/src/sections/input/AppInputBar.tsx
@@ -148,7 +148,7 @@ const AppInputBar = React.memo(
      classification === "search";

    const { forcedToolIds, setForcedToolIds } = useForcedTools();
-    const { currentMessageFiles, setCurrentMessageFiles, currentProjectId } =
+    const { currentMessageFiles, setCurrentMessageFiles } =
      useProjectsContext();

    const currentIndexingFiles = useMemo(() => {
@@ -200,17 +200,9 @@ const AppInputBar = React.memo(
      const textarea = textAreaRef.current;
      if (!wrapper || !textarea) return;

-      // Reset so scrollHeight reflects actual content size
      wrapper.style.height = `${MIN_INPUT_HEIGHT}px`;
-
-      // scrollHeight doesn't include the wrapper's padding, so add it back
-      const wrapperStyle = getComputedStyle(wrapper);
-      const paddingTop = parseFloat(wrapperStyle.paddingTop);
-      const paddingBottom = parseFloat(wrapperStyle.paddingBottom);
-      const contentHeight = textarea.scrollHeight + paddingTop + paddingBottom;
-
      wrapper.style.height = `${Math.min(
-        Math.max(contentHeight, MIN_INPUT_HEIGHT),
+        Math.max(textarea.scrollHeight, MIN_INPUT_HEIGHT),
        MAX_INPUT_HEIGHT
      )}px`;
    }, [message, isSearchMode]);
@@ -366,19 +358,13 @@ const AppInputBar = React.memo(
    const showDeepResearch = useMemo(() => {
      const deepResearchGloballyEnabled =
        combinedSettings?.settings?.deep_research_enabled ?? true;
-      const isProjectWorkflow = currentProjectId !== null;
-
-      // TODO(@yuhong): Re-enable Deep Research in Projects workflow once it is fully supported.
-      // https://linear.app/onyx-app/issue/ENG-3818/re-enable-deep-research-in-projects
      return (
-        !isProjectWorkflow &&
        deepResearchGloballyEnabled &&
        hasSearchToolsAvailable(selectedAgent?.tools || [])
      );
    }, [
      selectedAgent?.tools,
      combinedSettings?.settings?.deep_research_enabled,
-      currentProjectId,
    ]);

    function handleKeyDownForPromptShortcuts(
--- a/web/src/sections/modals/llmConfig/BedrockModal.tsx
+++ b/web/src/sections/modals/llmConfig/BedrockModal.tsx
@@ -32,7 +32,6 @@ import Separator from "@/refresh-components/Separator";
 import Text from "@/refresh-components/texts/Text";
 import Tabs from "@/refresh-components/Tabs";
 import { cn } from "@/lib/utils";
-import { ScopedMutator } from "swr";

 export const BEDROCK_PROVIDER_NAME = "bedrock";
 const BEDROCK_DISPLAY_NAME = "AWS Bedrock";
@@ -84,7 +83,7 @@ interface BedrockModalInternalsProps {
  modelConfigurations: ModelConfiguration[];
  isTesting: boolean;
  testError: string;
-  mutate: ScopedMutator;
+  mutate: (key: string) => void;
  onClose: () => void;
 }

--- a/web/src/sections/modals/llmConfig/CustomModal.test.tsx
+++ b/web/src/sections/modals/llmConfig/CustomModal.test.tsx
@@ -164,18 +164,6 @@ describe("Custom LLM Provider Configuration Workflow", () => {

    // Verify SWR cache was invalidated
    expect(mockMutate).toHaveBeenCalledWith("/api/admin/llm/provider");
-    expect(mockMutate).toHaveBeenCalledWith("/api/llm/provider");
-
-    const personaProvidersMutateCall = mockMutate.mock.calls.find(
-      ([key]) => typeof key === "function"
-    );
-    expect(personaProvidersMutateCall).toBeDefined();
-
-    const personaProviderFilter = personaProvidersMutateCall?.[0] as (
-      key: unknown
-    ) => boolean;
-    expect(personaProviderFilter("/api/llm/persona/42/providers")).toBe(true);
-    expect(personaProviderFilter("/api/llm/provider")).toBe(false);
  });

  test("shows error when test configuration fails", async () => {
--- a/web/src/sections/modals/llmConfig/LMStudioForm.tsx
+++ b/web/src/sections/modals/llmConfig/LMStudioForm.tsx
@@ -27,7 +27,6 @@ import { DisplayModels } from "./components/DisplayModels";
 import { useCallback, useEffect, useMemo, useState } from "react";
 import { fetchModels } from "@/app/admin/configuration/llm/utils";
 import debounce from "lodash/debounce";
-import { ScopedMutator } from "swr";

 const DEFAULT_API_BASE = "http://localhost:1234";

@@ -47,7 +46,7 @@ interface LMStudioFormContentProps {
  setHasFetched: (value: boolean) => void;
  isTesting: boolean;
  testError: string;
-  mutate: ScopedMutator;
+  mutate: () => void;
  onClose: () => void;
  isFormValid: boolean;
 }
--- a/web/src/sections/modals/llmConfig/OllamaModal.tsx
+++ b/web/src/sections/modals/llmConfig/OllamaModal.tsx
@@ -26,7 +26,6 @@ import { DisplayModels } from "./components/DisplayModels";
 import { useCallback, useEffect, useMemo, useState } from "react";
 import { fetchOllamaModels } from "@/app/admin/configuration/llm/utils";
 import debounce from "lodash/debounce";
-import { ScopedMutator } from "swr";

 export const OLLAMA_PROVIDER_NAME = "ollama_chat";
 const DEFAULT_API_BASE = "http://127.0.0.1:11434";
@@ -45,7 +44,7 @@ interface OllamaModalContentProps {
  setFetchedModels: (models: ModelConfiguration[]) => void;
  isTesting: boolean;
  testError: string;
-  mutate: ScopedMutator;
+  mutate: () => void;
  onClose: () => void;
  isFormValid: boolean;
 }
--- a/web/src/sections/modals/llmConfig/components/FormActionButtons.tsx
+++ b/web/src/sections/modals/llmConfig/components/FormActionButtons.tsx
@@ -4,15 +4,14 @@ import Button from "@/refresh-components/buttons/Button";
 import { Button as OpalButton } from "@opal/components";
 import { SvgTrash } from "@opal/icons";
 import { LLMProviderView } from "@/interfaces/llm";
-import { refreshLlmProviderCaches } from "@/lib/llmConfig/cache";
+import { LLM_PROVIDERS_ADMIN_URL } from "@/lib/llmConfig/constants";
 import { deleteLlmProvider } from "@/lib/llmConfig/svc";
-import { ScopedMutator } from "swr";

 interface FormActionButtonsProps {
  isTesting: boolean;
  testError: string;
  existingLlmProvider?: LLMProviderView;
-  mutate: ScopedMutator;
+  mutate: (key: string) => void;
  onClose: () => void;
  isFormValid: boolean;
 }
@@ -30,7 +29,7 @@ export function FormActionButtons({

    try {
      await deleteLlmProvider(existingLlmProvider.id);
-      await refreshLlmProviderCaches(mutate);
+      mutate(LLM_PROVIDERS_ADMIN_URL);
      onClose();
    } catch (e) {
      const message = e instanceof Error ? e.message : "Unknown error";
--- a/web/src/sections/modals/llmConfig/components/FormWrapper.tsx
+++ b/web/src/sections/modals/llmConfig/components/FormWrapper.tsx
@@ -1,7 +1,7 @@
 "use client";

 import { useState, ReactNode } from "react";
-import useSWR, { useSWRConfig, ScopedMutator } from "swr";
+import useSWR, { useSWRConfig, KeyedMutator } from "swr";
 import { toast } from "@/hooks/useToast";
 import {
  LLMProviderView,
@@ -14,12 +14,12 @@ import { Button } from "@opal/components";
 import { SvgSettings } from "@opal/icons";
 import { Badge } from "@/components/ui/badge";
 import { cn } from "@/lib/utils";
-import { refreshLlmProviderCaches } from "@/lib/llmConfig/cache";
+import { LLM_PROVIDERS_ADMIN_URL } from "@/lib/llmConfig/constants";
 import { setDefaultLlmModel } from "@/lib/llmConfig/svc";

 export interface ProviderFormContext {
  onClose: () => void;
-  mutate: ScopedMutator;
+  mutate: KeyedMutator<any>;
  isTesting: boolean;
  setIsTesting: (testing: boolean) => void;
  testError: string;
@@ -95,7 +95,7 @@ export function ProviderFormEntrypointWrapper({

    try {
      await setDefaultLlmModel(existingLlmProvider.id, firstVisibleModel.name);
-      await refreshLlmProviderCaches(mutate);
+      await mutate(LLM_PROVIDERS_ADMIN_URL);
      toast.success("Provider set as default successfully!");
    } catch (e) {
      const message = e instanceof Error ? e.message : "Unknown error";
--- a/web/src/sections/modals/llmConfig/formUtils.ts
+++ b/web/src/sections/modals/llmConfig/formUtils.ts
@@ -7,11 +7,9 @@ import {
  LLM_ADMIN_URL,
  LLM_PROVIDERS_ADMIN_URL,
 } from "@/lib/llmConfig/constants";
-import { refreshLlmProviderCaches } from "@/lib/llmConfig/cache";
 import { toast } from "@/hooks/useToast";
 import * as Yup from "yup";
 import isEqual from "lodash/isEqual";
-import { ScopedMutator } from "swr";

 // Common class names for the Form component across all LLM provider forms
 export const LLM_FORM_CLASS_NAME = "flex flex-col gap-y-4 items-stretch mt-6";
@@ -107,7 +105,7 @@ export interface SubmitLLMProviderParams<
  hideSuccess?: boolean;
  setIsTesting: (testing: boolean) => void;
  setTestError: (error: string) => void;
-  mutate: ScopedMutator;
+  mutate: (key: string) => void;
  onClose: () => void;
  setSubmitting: (submitting: boolean) => void;
 }
@@ -289,7 +287,7 @@ export const submitLLMProvider = async <T extends BaseLLMFormValues>({
    }
  }

-  await refreshLlmProviderCaches(mutate);
+  mutate(LLM_PROVIDERS_ADMIN_URL);
  onClose();

  if (!hideSuccess) {
--- a/web/src/sections/onboarding/forms/tests/AnthropicOnboardingForm.test.tsx
+++ b/web/src/sections/onboarding/forms/tests/AnthropicOnboardingForm.test.tsx
@@ -10,6 +10,7 @@ import {
  createMockOnboardingActions,
  createMockFetchResponses,
  MOCK_PROVIDERS,
+  ANTHROPIC_DEFAULT_VISIBLE_MODELS,
 } from "./testHelpers";

 // Mock fetch
@@ -50,6 +51,8 @@ jest.mock("@/components/modals/ProviderModal", () => ({
  },
 }));

+// Mock fetchModels utility - returns the curated Anthropic visible models
+// that match ANTHROPIC_VISIBLE_MODEL_NAMES from backend
 const mockFetchModels = jest.fn().mockResolvedValue({
  models: [
    {
@@ -149,6 +152,71 @@ describe("AnthropicOnboardingForm", () => {
    });
  });

+  describe("Default Available Models", () => {
+    /**
+     * This test verifies that the exact curated list of Anthropic visible models
+     * matches what's returned from /api/admin/llm/built-in/options.
+     * The expected models are defined in ANTHROPIC_VISIBLE_MODEL_NAMES in
+     * backend/onyx/llm/llm_provider_options.py
+     */
+    test("llmDescriptor contains the correct default visible models from built-in options", () => {
+      const expectedModelNames = [
+        "claude-opus-4-5",
+        "claude-sonnet-4-5",
+        "claude-haiku-4-5",
+      ];
+
+      // Verify MOCK_PROVIDERS.anthropic has the correct model configurations
+      const actualModelNames = MOCK_PROVIDERS.anthropic.known_models.map(
+        (config) => config.name
+      );
+
+      // Check that all expected models are present
+      expect(actualModelNames).toEqual(
+        expect.arrayContaining(expectedModelNames)
+      );
+
+      // Check that only the expected models are present (no extras)
+      expect(actualModelNames).toHaveLength(expectedModelNames.length);
+
+      // Verify each model has is_visible set to true
+      MOCK_PROVIDERS.anthropic.known_models.forEach((config) => {
+        expect(config.is_visible).toBe(true);
+      });
+    });
+
+    test("ANTHROPIC_DEFAULT_VISIBLE_MODELS matches backend ANTHROPIC_VISIBLE_MODEL_NAMES", () => {
+      // These are the exact model names from backend/onyx/llm/llm_provider_options.py
+      // ANTHROPIC_VISIBLE_MODEL_NAMES = {"claude-opus-4-5", "claude-sonnet-4-5", "claude-haiku-4-5"}
+      const backendVisibleModelNames = new Set([
+        "claude-opus-4-5",
+        "claude-sonnet-4-5",
+        "claude-haiku-4-5",
+      ]);
+
+      const testHelperModelNames = new Set(
+        ANTHROPIC_DEFAULT_VISIBLE_MODELS.map((m) => m.name)
+      );
+
+      expect(testHelperModelNames).toEqual(backendVisibleModelNames);
+    });
+
+    test("all default models are marked as visible", () => {
+      ANTHROPIC_DEFAULT_VISIBLE_MODELS.forEach((model) => {
+        expect(model.is_visible).toBe(true);
+      });
+    });
+
+    test("default model claude-sonnet-4-5 is set correctly in component", () => {
+      // The AnthropicOnboardingForm sets DEFAULT_DEFAULT_MODEL_NAME = "claude-sonnet-4-5"
+      // Verify this model exists in the default visible models
+      const defaultModelExists = ANTHROPIC_DEFAULT_VISIBLE_MODELS.some(
+        (m) => m.name === "claude-sonnet-4-5"
+      );
+      expect(defaultModelExists).toBe(true);
+    });
+  });
+
  describe("Form Validation", () => {
    test("submit button is disabled when form is empty", () => {
      render(<AnthropicOnboardingForm {...defaultProps} />);
--- a/web/src/sections/onboarding/forms/tests/OpenAIOnboardingForm.test.tsx
+++ b/web/src/sections/onboarding/forms/tests/OpenAIOnboardingForm.test.tsx
@@ -10,6 +10,7 @@ import {
  createMockOnboardingActions,
  createMockFetchResponses,
  MOCK_PROVIDERS,
+  OPENAI_DEFAULT_VISIBLE_MODELS,
 } from "./testHelpers";

 // Mock fetch
@@ -53,6 +54,8 @@ jest.mock("@/components/modals/ProviderModal", () => ({
  },
 }));

+// Mock fetchModels utility - returns the curated OpenAI visible models
+// that match OPENAI_VISIBLE_MODEL_NAMES from backend
 const mockFetchModels = jest.fn().mockResolvedValue({
  models: [
    {
@@ -170,6 +173,77 @@ describe("OpenAIOnboardingForm", () => {
    });
  });

+  describe("Default Available Models", () => {
+    /**
+     * This test verifies that the exact curated list of OpenAI visible models
+     * matches what's returned from /api/admin/llm/built-in/options.
+     * The expected models are defined in OPENAI_VISIBLE_MODEL_NAMES in
+     * backend/onyx/llm/llm_provider_options.py
+     */
+    test("llmDescriptor contains the correct default visible models from built-in options", () => {
+      const expectedModelNames = [
+        "gpt-5.2",
+        "gpt-5-mini",
+        "o1",
+        "o3-mini",
+        "gpt-4o",
+        "gpt-4o-mini",
+      ];
+
+      // Verify MOCK_PROVIDERS.openai has the correct model configurations
+      const actualModelNames = MOCK_PROVIDERS.openai.known_models.map(
+        (config) => config.name
+      );
+
+      // Check that all expected models are present
+      expect(actualModelNames).toEqual(
+        expect.arrayContaining(expectedModelNames)
+      );
+
+      // Check that only the expected models are present (no extras)
+      expect(actualModelNames).toHaveLength(expectedModelNames.length);
+
+      // Verify each model has is_visible set to true
+      MOCK_PROVIDERS.openai.known_models.forEach((config) => {
+        expect(config.is_visible).toBe(true);
+      });
+    });
+
+    test("OPENAI_DEFAULT_VISIBLE_MODELS matches backend OPENAI_VISIBLE_MODEL_NAMES", () => {
+      // These are the exact model names from backend/onyx/llm/llm_provider_options.py
+      // OPENAI_VISIBLE_MODEL_NAMES = {"gpt-5.2", "gpt-5-mini", "o1", "o3-mini", "gpt-4o", "gpt-4o-mini"}
+      const backendVisibleModelNames = new Set([
+        "gpt-5.2",
+        "gpt-5-mini",
+        "o1",
+        "o3-mini",
+        "gpt-4o",
+        "gpt-4o-mini",
+      ]);
+
+      const testHelperModelNames = new Set(
+        OPENAI_DEFAULT_VISIBLE_MODELS.map((m) => m.name)
+      );
+
+      expect(testHelperModelNames).toEqual(backendVisibleModelNames);
+    });
+
+    test("all default models are marked as visible", () => {
+      OPENAI_DEFAULT_VISIBLE_MODELS.forEach((model) => {
+        expect(model.is_visible).toBe(true);
+      });
+    });
+
+    test("default model gpt-5.2 is set correctly in component", () => {
+      // The OpenAIOnboardingForm sets DEFAULT_DEFAULT_MODEL_NAME = "gpt-5.2"
+      // Verify this model exists in the default visible models
+      const defaultModelExists = OPENAI_DEFAULT_VISIBLE_MODELS.some(
+        (m) => m.name === "gpt-5.2"
+      );
+      expect(defaultModelExists).toBe(true);
+    });
+  });
+
  describe("Form Validation", () => {
    test("submit button is disabled when form is empty", () => {
      render(<OpenAIOnboardingForm {...defaultProps} />);
--- a/web/src/sections/onboarding/forms/tests/VertexAIOnboardingForm.test.tsx
+++ b/web/src/sections/onboarding/forms/tests/VertexAIOnboardingForm.test.tsx
@@ -10,6 +10,7 @@ import {
  createMockOnboardingActions,
  createMockFetchResponses,
  MOCK_PROVIDERS,
+  VERTEXAI_DEFAULT_VISIBLE_MODELS,
 } from "./testHelpers";

 // Mock fetch
@@ -50,6 +51,8 @@ jest.mock("@/components/modals/ProviderModal", () => ({
  },
 }));

+// Mock fetchModels utility - returns the curated Vertex AI visible models
+// that match VERTEXAI_VISIBLE_MODEL_NAMES from backend
 jest.mock("@/app/admin/configuration/llm/utils", () => ({
  canProviderFetchModels: jest.fn().mockReturnValue(true),
  fetchModels: jest.fn().mockResolvedValue({
@@ -151,6 +154,71 @@ describe("VertexAIOnboardingForm", () => {
    });
  });

+  describe("Default Available Models", () => {
+    /**
+     * This test verifies that the exact curated list of Vertex AI visible models
+     * matches what's returned from /api/admin/llm/built-in/options.
+     * The expected models are defined in VERTEXAI_VISIBLE_MODEL_NAMES in
+     * backend/onyx/llm/llm_provider_options.py
+     */
+    test("llmDescriptor contains the correct default visible models from built-in options", () => {
+      const expectedModelNames = [
+        "gemini-2.5-flash",
+        "gemini-2.5-flash-lite",
+        "gemini-2.5-pro",
+      ];
+
+      // Verify MOCK_PROVIDERS.vertexAi has the correct model configurations
+      const actualModelNames = MOCK_PROVIDERS.vertexAi.known_models.map(
+        (config) => config.name
+      );
+
+      // Check that all expected models are present
+      expect(actualModelNames).toEqual(
+        expect.arrayContaining(expectedModelNames)
+      );
+
+      // Check that only the expected models are present (no extras)
+      expect(actualModelNames).toHaveLength(expectedModelNames.length);
+
+      // Verify each model has is_visible set to true
+      MOCK_PROVIDERS.vertexAi.known_models.forEach((config) => {
+        expect(config.is_visible).toBe(true);
+      });
+    });
+
+    test("VERTEXAI_DEFAULT_VISIBLE_MODELS matches backend VERTEXAI_VISIBLE_MODEL_NAMES", () => {
+      // These are the exact model names from backend/onyx/llm/llm_provider_options.py
+      // VERTEXAI_VISIBLE_MODEL_NAMES = {"gemini-2.5-flash", "gemini-2.5-flash-lite", "gemini-2.5-pro"}
+      const backendVisibleModelNames = new Set([
+        "gemini-2.5-flash",
+        "gemini-2.5-flash-lite",
+        "gemini-2.5-pro",
+      ]);
+
+      const testHelperModelNames = new Set(
+        VERTEXAI_DEFAULT_VISIBLE_MODELS.map((m) => m.name)
+      );
+
+      expect(testHelperModelNames).toEqual(backendVisibleModelNames);
+    });
+
+    test("all default models are marked as visible", () => {
+      VERTEXAI_DEFAULT_VISIBLE_MODELS.forEach((model) => {
+        expect(model.is_visible).toBe(true);
+      });
+    });
+
+    test("default model gemini-2.5-pro is set correctly in component", () => {
+      // The VertexAIOnboardingForm sets DEFAULT_DEFAULT_MODEL_NAME = "gemini-2.5-pro"
+      // Verify this model exists in the default visible models
+      const defaultModelExists = VERTEXAI_DEFAULT_VISIBLE_MODELS.some(
+        (m) => m.name === "gemini-2.5-pro"
+      );
+      expect(defaultModelExists).toBe(true);
+    });
+  });
+
  describe("Form Validation", () => {
    test("submit button is disabled when form is empty", () => {
      render(<VertexAIOnboardingForm {...defaultProps} />);
--- a/web/src/sections/onboarding/forms/tests/testHelpers.tsx
+++ b/web/src/sections/onboarding/forms/tests/testHelpers.tsx
@@ -1,6 +1,7 @@
 /**
 * Shared test helpers and mocks for onboarding form tests
 */
+import React from "react";

 // Mock Element.prototype.scrollIntoView for JSDOM (not implemented in jsdom)
 Element.prototype.scrollIntoView = jest.fn();
@@ -160,6 +161,11 @@ export async function waitForModalOpen(screen: any, waitFor: any) {
 /**
 * Common provider descriptors for testing
 */
+/**
+ * The curated list of OpenAI visible models that are returned by
+ * /api/admin/llm/built-in/options. This must match OPENAI_VISIBLE_MODEL_NAMES
+ * in backend/onyx/llm/llm_provider_options.py
+ */
 export const OPENAI_DEFAULT_VISIBLE_MODELS = [
  {
    name: "gpt-5.2",
@@ -205,6 +211,11 @@ export const OPENAI_DEFAULT_VISIBLE_MODELS = [
  },
 ];

+/**
+ * The curated list of Anthropic visible models that are returned by
+ * /api/admin/llm/built-in/options. This must match ANTHROPIC_VISIBLE_MODEL_NAMES
+ * in backend/onyx/llm/llm_provider_options.py
+ */
 export const ANTHROPIC_DEFAULT_VISIBLE_MODELS = [
  {
    name: "claude-opus-4-5",
@@ -229,6 +240,11 @@ export const ANTHROPIC_DEFAULT_VISIBLE_MODELS = [
  },
 ];

+/**
+ * The curated list of Vertex AI visible models that are returned by
+ * /api/admin/llm/built-in/options. This must match VERTEXAI_VISIBLE_MODEL_NAMES
+ * in backend/onyx/llm/llm_provider_options.py
+ */
 export const VERTEXAI_DEFAULT_VISIBLE_MODELS = [
  {
    name: "gemini-2.5-flash",
--- a/web/tests/e2e/agents/create_and_edit_agent.spec.ts
+++ b/web/tests/e2e/agents/create_and_edit_agent.spec.ts
@@ -292,10 +292,7 @@ test.describe("Assistant Creation and Edit Verification", () => {
      expect(agentIdMatch).toBeTruthy();
      const agentId = agentIdMatch ? agentIdMatch[1] : null;
      expect(agentId).not.toBeNull();
-      await expectScreenshot(page, {
-        name: "welcome-page-with-assistant",
-        hide: ["[data-testid='AppInputBar/llm-popover-trigger']"],
-      });
+      await expectScreenshot(page, { name: "welcome-page-with-assistant" });

      // Store assistant ID for cleanup
      knowledgeAssistantId = Number(agentId);
Author	SHA1	Message	Date
Dane Urban	d7fdbaf772	Make better	2026-03-06 11:22:38 -08:00
Dane Urban	ddb2e069ed	.	2026-03-06 11:22:38 -08:00
Dane Urban	4b092cf252	.	2026-03-06 11:22:38 -08:00
Dane Urban	783c8a6481	Prevent the removal and hiding of default model	2026-03-06 11:22:38 -08:00