Merge main HEAD (segment 5) into sandboxed-agent-rebase

Resolve 83 conflicts: 10 backend, 62 frontend, 11 config/lock files.
Preserve sandbox/agent/collaboration features while adopting main's
UI refactorings (Dialog/AlertDialog/Popover), model provider updates,
and enterprise features.

Made-with: Cursor

api/tests/unit_tests/core/app/apps/advanced_chat/test_app_generator.py

@@ -234,6 +234,7 @@ class TestAdvancedChatAppGeneratorInternals:
         captured: dict[str, object] = {}
         prefill_calls: list[object] = []
         var_loader = SimpleNamespace(loader="draft")
+        workflow = SimpleNamespace(id="workflow-id")
 
         monkeypatch.setattr(
             "core.app.apps.advanced_chat.app_generator.AdvancedChatAppConfigManager.get_app_config",
@@ -260,8 +261,8 @@ class TestAdvancedChatAppGeneratorInternals:
             def __init__(self, session):
                 _ = session
 
-            def prefill_conversation_variable_default_values(self, workflow):
-                prefill_calls.append(workflow)
+            def prefill_conversation_variable_default_values(self, workflow, user_id):
+                prefill_calls.append((workflow, user_id))
 
         monkeypatch.setattr("core.app.apps.advanced_chat.app_generator.WorkflowDraftVariableService", _DraftVarService)
 
@@ -273,7 +274,7 @@ class TestAdvancedChatAppGeneratorInternals:
 
         result = generator.single_iteration_generate(
             app_model=SimpleNamespace(id="app", tenant_id="tenant"),
-            workflow=SimpleNamespace(id="workflow-id"),
+            workflow=workflow,
             node_id="node-1",
             user=SimpleNamespace(id="user-id"),
             args={"inputs": {"foo": "bar"}},
@@ -281,7 +282,7 @@ class TestAdvancedChatAppGeneratorInternals:
         )
 
         assert result == {"ok": True}
-        assert prefill_calls
+        assert prefill_calls == [(workflow, "user-id")]
         assert captured["variable_loader"] is var_loader
         assert captured["application_generate_entity"].single_iteration_run.node_id == "node-1"
 
@@ -291,6 +292,7 @@ class TestAdvancedChatAppGeneratorInternals:
         captured: dict[str, object] = {}
         prefill_calls: list[object] = []
         var_loader = SimpleNamespace(loader="draft")
+        workflow = SimpleNamespace(id="workflow-id")
 
         monkeypatch.setattr(
             "core.app.apps.advanced_chat.app_generator.AdvancedChatAppConfigManager.get_app_config",
@@ -317,8 +319,8 @@ class TestAdvancedChatAppGeneratorInternals:
             def __init__(self, session):
                 _ = session
 
-            def prefill_conversation_variable_default_values(self, workflow):
-                prefill_calls.append(workflow)
+            def prefill_conversation_variable_default_values(self, workflow, user_id):
+                prefill_calls.append((workflow, user_id))
 
         monkeypatch.setattr("core.app.apps.advanced_chat.app_generator.WorkflowDraftVariableService", _DraftVarService)
 
@@ -330,7 +332,7 @@ class TestAdvancedChatAppGeneratorInternals:
 
         result = generator.single_loop_generate(
             app_model=SimpleNamespace(id="app", tenant_id="tenant"),
-            workflow=SimpleNamespace(id="workflow-id"),
+            workflow=workflow,
             node_id="node-2",
             user=SimpleNamespace(id="user-id"),
             args=SimpleNamespace(inputs={"foo": "bar"}),
@@ -338,7 +340,7 @@ class TestAdvancedChatAppGeneratorInternals:
         )
 
         assert result == {"ok": True}
-        assert prefill_calls
+        assert prefill_calls == [(workflow, "user-id")]
         assert captured["variable_loader"] is var_loader
         assert captured["application_generate_entity"].single_loop_run.node_id == "node-2"
 

api/tests/unit_tests/core/app/apps/agent_chat/test_agent_chat_generate_response_converter.py

@@ -44,11 +44,22 @@ class TestAgentChatAppGenerateResponseConverterBlocking:
                 metadata={
                     "retriever_resources": [
                         {
+                            "dataset_id": "dataset-1",
+                            "dataset_name": "Dataset 1",
+                            "document_id": "document-1",
                             "segment_id": "s1",
                             "position": 1,
+                            "data_source_type": "file",
                             "document_name": "doc",
                             "score": 0.9,
+                            "hit_count": 2,
+                            "word_count": 128,
+                            "segment_position": 3,
+                            "index_node_hash": "abc1234",
                             "content": "content",
+                            "page": 5,
+                            "title": "Citation Title",
+                            "files": [{"id": "file-1"}],
                         }
                     ],
                     "annotation_reply": {"id": "a"},
@@ -107,11 +118,22 @@ class TestAgentChatAppGenerateResponseConverterStream:
                     metadata={
                         "retriever_resources": [
                             {
+                                "dataset_id": "dataset-1",
+                                "dataset_name": "Dataset 1",
+                                "document_id": "document-1",
                                 "segment_id": "s1",
                                 "position": 1,
+                                "data_source_type": "file",
                                 "document_name": "doc",
                                 "score": 0.9,
+                                "hit_count": 2,
+                                "word_count": 128,
+                                "segment_position": 3,
+                                "index_node_hash": "abc1234",
                                 "content": "content",
+                                "page": 5,
+                                "title": "Citation Title",
+                                "files": [{"id": "file-1"}],
                                 "summary": "summary",
                                 "extra": "ignored",
                             }
@@ -151,11 +173,22 @@ class TestAgentChatAppGenerateResponseConverterStream:
         assert "usage" not in metadata
         assert metadata["retriever_resources"] == [
             {
+                "dataset_id": "dataset-1",
+                "dataset_name": "Dataset 1",
+                "document_id": "document-1",
                 "segment_id": "s1",
                 "position": 1,
+                "data_source_type": "file",
                 "document_name": "doc",
                 "score": 0.9,
+                "hit_count": 2,
+                "word_count": 128,
+                "segment_position": 3,
+                "index_node_hash": "abc1234",
                 "content": "content",
+                "page": 5,
+                "title": "Citation Title",
+                "files": [{"id": "file-1"}],
                 "summary": "summary",
             }
         ]

api/tests/unit_tests/core/app/apps/common/test_workflow_response_converter_truncation.py

@@ -5,6 +5,7 @@ Unit tests for WorkflowResponseConverter focusing on process_data truncation fun
 import uuid
 from collections.abc import Mapping
 from dataclasses import dataclass
+from datetime import UTC, datetime
 from typing import Any
 from unittest.mock import Mock
 
@@ -234,6 +235,50 @@ class TestWorkflowResponseConverter:
         assert response.data.process_data == {}
         assert response.data.process_data_truncated is False
 
+    def test_workflow_node_finish_response_prefers_event_finished_at(
+        self,
+        monkeypatch: pytest.MonkeyPatch,
+    ) -> None:
+        """Finished timestamps should come from the event, not delayed queue processing time."""
+        converter = self.create_workflow_response_converter()
+        start_at = datetime(2024, 1, 1, 0, 0, 0, tzinfo=UTC).replace(tzinfo=None)
+        finished_at = datetime(2024, 1, 1, 0, 0, 2, tzinfo=UTC).replace(tzinfo=None)
+        delayed_processing_time = datetime(2024, 1, 1, 0, 0, 10, tzinfo=UTC).replace(tzinfo=None)
+
+        monkeypatch.setattr(
+            "core.app.apps.common.workflow_response_converter.naive_utc_now",
+            lambda: delayed_processing_time,
+        )
+        converter.workflow_start_to_stream_response(
+            task_id="bootstrap",
+            workflow_run_id="run-id",
+            workflow_id="wf-id",
+            reason=WorkflowStartReason.INITIAL,
+        )
+
+        event = QueueNodeSucceededEvent(
+            node_id="test-node-id",
+            node_type=BuiltinNodeTypes.CODE,
+            node_execution_id="node-exec-1",
+            start_at=start_at,
+            finished_at=finished_at,
+            in_iteration_id=None,
+            in_loop_id=None,
+            inputs={},
+            process_data={},
+            outputs={},
+            execution_metadata={},
+        )
+
+        response = converter.workflow_node_finish_to_stream_response(
+            event=event,
+            task_id="test-task-id",
+        )
+
+        assert response is not None
+        assert response.data.elapsed_time == 2.0
+        assert response.data.finished_at == int(finished_at.timestamp())
+
     def test_workflow_node_retry_response_uses_truncated_process_data(self):
         """Test that node retry response uses get_response_process_data()."""
         converter = self.create_workflow_response_converter()

api/tests/unit_tests/core/app/apps/completion/test_completion_generate_response_converter.py

@@ -38,11 +38,22 @@ class TestCompletionAppGenerateResponseConverter:
         metadata = {
             "retriever_resources": [
                 {
+                    "dataset_id": "dataset-1",
+                    "dataset_name": "Dataset 1",
+                    "document_id": "document-1",
                     "segment_id": "s",
                     "position": 1,
+                    "data_source_type": "file",
                     "document_name": "doc",
                     "score": 0.9,
+                    "hit_count": 2,
+                    "word_count": 128,
+                    "segment_position": 3,
+                    "index_node_hash": "abc1234",
                     "content": "c",
+                    "page": 5,
+                    "title": "Citation Title",
+                    "files": [{"id": "file-1"}],
                     "summary": "sum",
                     "extra": "x",
                 }
@@ -66,7 +77,12 @@ class TestCompletionAppGenerateResponseConverter:
 
         assert "annotation_reply" not in result["metadata"]
         assert "usage" not in result["metadata"]
+        assert result["metadata"]["retriever_resources"][0]["dataset_id"] == "dataset-1"
+        assert result["metadata"]["retriever_resources"][0]["document_id"] == "document-1"
         assert result["metadata"]["retriever_resources"][0]["segment_id"] == "s"
+        assert result["metadata"]["retriever_resources"][0]["data_source_type"] == "file"
+        assert result["metadata"]["retriever_resources"][0]["segment_position"] == 3
+        assert result["metadata"]["retriever_resources"][0]["index_node_hash"] == "abc1234"
         assert "extra" not in result["metadata"]["retriever_resources"][0]
 
     def test_convert_blocking_simple_response_metadata_not_dict(self):

api/tests/unit_tests/core/app/apps/test_advanced_chat_app_generator.py

@@ -11,6 +11,7 @@ from core.app.apps.advanced_chat.app_generator import AdvancedChatAppGenerator
 from core.app.entities.app_invoke_entities import AdvancedChatAppGenerateEntity, InvokeFrom
 from core.app.task_pipeline import message_cycle_manager
 from core.app.task_pipeline.message_cycle_manager import MessageCycleManager
+from models.enums import ConversationFromSource
 from models.model import AppMode, Conversation, Message
 
 
@@ -92,7 +93,7 @@ def test_init_generate_records_marks_existing_conversation():
         system_instruction_tokens=0,
         status="normal",
         invoke_from=InvokeFrom.WEB_APP.value,
-        from_source="api",
+        from_source=ConversationFromSource.API,
         from_end_user_id="user-id",
         from_account_id=None,
     )

api/tests/unit_tests/core/app/workflow/layers/test_persistence.py

@@ -0,0 +1,60 @@
+from datetime import UTC, datetime
+from unittest.mock import Mock
+
+import pytest
+
+from core.app.workflow.layers.persistence import (
+    PersistenceWorkflowInfo,
+    WorkflowPersistenceLayer,
+    _NodeRuntimeSnapshot,
+)
+from dify_graph.enums import WorkflowNodeExecutionStatus, WorkflowType
+from dify_graph.node_events import NodeRunResult
+
+
+def _build_layer() -> WorkflowPersistenceLayer:
+    application_generate_entity = Mock()
+    application_generate_entity.inputs = {}
+
+    return WorkflowPersistenceLayer(
+        application_generate_entity=application_generate_entity,
+        workflow_info=PersistenceWorkflowInfo(
+            workflow_id="workflow-id",
+            workflow_type=WorkflowType.WORKFLOW,
+            version="1",
+            graph_data={},
+        ),
+        workflow_execution_repository=Mock(),
+        workflow_node_execution_repository=Mock(),
+    )
+
+
+def test_update_node_execution_prefers_event_finished_at(monkeypatch: pytest.MonkeyPatch) -> None:
+    layer = _build_layer()
+    node_execution = Mock()
+    node_execution.id = "node-exec-1"
+    node_execution.created_at = datetime(2024, 1, 1, 0, 0, 0, tzinfo=UTC).replace(tzinfo=None)
+    node_execution.update_from_mapping = Mock()
+
+    layer._node_snapshots[node_execution.id] = _NodeRuntimeSnapshot(
+        node_id="node-id",
+        title="LLM",
+        predecessor_node_id=None,
+        iteration_id="iter-1",
+        loop_id=None,
+        created_at=node_execution.created_at,
+    )
+
+    event_finished_at = datetime(2024, 1, 1, 0, 0, 2, tzinfo=UTC).replace(tzinfo=None)
+    delayed_processing_time = datetime(2024, 1, 1, 0, 0, 10, tzinfo=UTC).replace(tzinfo=None)
+    monkeypatch.setattr("core.app.workflow.layers.persistence.naive_utc_now", lambda: delayed_processing_time)
+
+    layer._update_node_execution(
+        node_execution,
+        NodeRunResult(status=WorkflowNodeExecutionStatus.SUCCEEDED),
+        WorkflowNodeExecutionStatus.SUCCEEDED,
+        finished_at=event_finished_at,
+    )
+
+    assert node_execution.finished_at == event_finished_at
+    assert node_execution.elapsed_time == 2.0

api/tests/unit_tests/core/datasource/test_datasource_file_manager.py

@@ -166,6 +166,7 @@ class TestDatasourceFileManager:
         # Setup
         mock_guess_ext.return_value = None  # Cannot guess
         mock_uuid.return_value = MagicMock(hex="unique_hex")
+        mock_config.STORAGE_TYPE = "local"
 
         # Execute
         upload_file = DatasourceFileManager.create_file_by_raw(

api/tests/unit_tests/core/entities/test_entities_provider_configuration.py

@@ -35,6 +35,7 @@ from dify_graph.model_runtime.entities.provider_entities import (
     ProviderCredentialSchema,
     ProviderEntity,
 )
+from models.enums import CredentialSourceType
 from models.provider import ProviderType
 from models.provider_ids import ModelProviderID
 
@@ -409,7 +410,7 @@ def test_switch_preferred_provider_type_updates_existing_record_with_session() -
 
     configuration.switch_preferred_provider_type(ProviderType.SYSTEM, session=session)
 
-    assert existing_record.preferred_provider_type == ProviderType.SYSTEM.value
+    assert existing_record.preferred_provider_type == ProviderType.SYSTEM
     session.commit.assert_called_once()
 
 
@@ -514,7 +515,7 @@ def test_get_custom_provider_models_sets_status_for_removed_credentials_and_inva
                         id="lb-base",
                         name="LB Base",
                         credentials={},
-                        credential_source_type="provider",
+                        credential_source_type=CredentialSourceType.PROVIDER,
                     )
                 ],
             ),
@@ -528,7 +529,7 @@ def test_get_custom_provider_models_sets_status_for_removed_credentials_and_inva
                         id="lb-custom",
                         name="LB Custom",
                         credentials={},
-                        credential_source_type="custom_model",
+                        credential_source_type=CredentialSourceType.CUSTOM_MODEL,
                     )
                 ],
             ),
@@ -734,7 +735,7 @@ def test_create_provider_credential_creates_provider_record_when_missing() -> No
 def test_create_provider_credential_marks_existing_provider_as_valid() -> None:
     configuration = _build_provider_configuration()
     session = Mock()
-    provider_record = SimpleNamespace(is_valid=False)
+    provider_record = SimpleNamespace(id="provider-1", is_valid=False, credential_id="existing-cred")
 
     with _patched_session(session):
         with patch.object(ProviderConfiguration, "_check_provider_credential_name_exists", return_value=False):
@@ -743,6 +744,25 @@ def test_create_provider_credential_marks_existing_provider_as_valid() -> None:
                     configuration.create_provider_credential({"api_key": "raw"}, "Main")
 
     assert provider_record.is_valid is True
+    assert provider_record.credential_id == "existing-cred"
+    session.commit.assert_called_once()
+
+
+def test_create_provider_credential_auto_activates_when_no_active_credential() -> None:
+    configuration = _build_provider_configuration()
+    session = Mock()
+    provider_record = SimpleNamespace(id="provider-1", is_valid=False, credential_id=None, updated_at=None)
+
+    with _patched_session(session):
+        with patch.object(ProviderConfiguration, "_check_provider_credential_name_exists", return_value=False):
+            with patch.object(ProviderConfiguration, "validate_provider_credentials", return_value={"api_key": "enc"}):
+                with patch.object(ProviderConfiguration, "_get_provider_record", return_value=provider_record):
+                    with patch("core.entities.provider_configuration.ProviderCredentialsCache"):
+                        with patch.object(ProviderConfiguration, "switch_preferred_provider_type"):
+                            configuration.create_provider_credential({"api_key": "raw"}, "Main")
+
+    assert provider_record.is_valid is True
+    assert provider_record.credential_id is not None
     session.commit.assert_called_once()
 
 
@@ -807,7 +827,7 @@ def test_update_load_balancing_configs_updates_all_matching_configs() -> None:
         configuration._update_load_balancing_configs_with_credential(
             credential_id="cred-1",
             credential_record=credential_record,
-            credential_source="provider",
+            credential_source=CredentialSourceType.PROVIDER,
             session=session,
         )
 
@@ -825,7 +845,7 @@ def test_update_load_balancing_configs_returns_when_no_matching_configs() -> Non
     configuration._update_load_balancing_configs_with_credential(
         credential_id="cred-1",
         credential_record=SimpleNamespace(encrypted_config="{}", credential_name="Main"),
-        credential_source="provider",
+        credential_source=CredentialSourceType.PROVIDER,
         session=session,
     )
 

api/tests/unit_tests/core/mcp/auth/test_auth_flow.py

@@ -801,6 +801,27 @@ class TestAuthOrchestration:
         urls = build_protected_resource_metadata_discovery_urls(None, "https://api.example.com")
         assert urls == ["https://api.example.com/.well-known/oauth-protected-resource"]
 
+    def test_build_protected_resource_metadata_discovery_urls_with_relative_hint(self):
+        urls = build_protected_resource_metadata_discovery_urls(
+            "/.well-known/oauth-protected-resource/tenant/mcp",
+            "https://api.example.com/tenant/mcp",
+        )
+        assert urls == [
+            "https://api.example.com/.well-known/oauth-protected-resource/tenant/mcp",
+            "https://api.example.com/.well-known/oauth-protected-resource",
+        ]
+
+    def test_build_protected_resource_metadata_discovery_urls_ignores_scheme_less_hint(self):
+        urls = build_protected_resource_metadata_discovery_urls(
+            "/openapi-mcp.cn-hangzhou.aliyuncs.com/.well-known/oauth-protected-resource/tenant/mcp",
+            "https://openapi-mcp.cn-hangzhou.aliyuncs.com/tenant/mcp",
+        )
+
+        assert urls == [
+            "https://openapi-mcp.cn-hangzhou.aliyuncs.com/.well-known/oauth-protected-resource/tenant/mcp",
+            "https://openapi-mcp.cn-hangzhou.aliyuncs.com/.well-known/oauth-protected-resource",
+        ]
+
     def test_build_oauth_authorization_server_metadata_discovery_urls(self):
         # Case 1: with auth_server_url
         urls = build_oauth_authorization_server_metadata_discovery_urls(

api/tests/unit_tests/core/moderation/api/test_api.py

@@ -0,0 +1,181 @@
+from unittest.mock import MagicMock, patch
+
+import pytest
+from pydantic import ValidationError
+
+from core.extension.api_based_extension_requestor import APIBasedExtensionPoint
+from core.moderation.api.api import ApiModeration, ModerationInputParams, ModerationOutputParams
+from core.moderation.base import ModerationAction, ModerationInputsResult, ModerationOutputsResult
+from models.api_based_extension import APIBasedExtension
+
+
+class TestApiModeration:
+    @pytest.fixture
+    def api_config(self):
+        return {
+            "inputs_config": {
+                "enabled": True,
+            },
+            "outputs_config": {
+                "enabled": True,
+            },
+            "api_based_extension_id": "test-extension-id",
+        }
+
+    @pytest.fixture
+    def api_moderation(self, api_config):
+        return ApiModeration(app_id="test-app-id", tenant_id="test-tenant-id", config=api_config)
+
+    def test_moderation_input_params(self):
+        params = ModerationInputParams(app_id="app-1", inputs={"key": "val"}, query="test query")
+        assert params.app_id == "app-1"
+        assert params.inputs == {"key": "val"}
+        assert params.query == "test query"
+
+        # Test defaults
+        params_default = ModerationInputParams()
+        assert params_default.app_id == ""
+        assert params_default.inputs == {}
+        assert params_default.query == ""
+
+    def test_moderation_output_params(self):
+        params = ModerationOutputParams(app_id="app-1", text="test text")
+        assert params.app_id == "app-1"
+        assert params.text == "test text"
+
+        with pytest.raises(ValidationError):
+            ModerationOutputParams()
+
+    @patch("core.moderation.api.api.ApiModeration._get_api_based_extension")
+    def test_validate_config_success(self, mock_get_extension, api_config):
+        mock_get_extension.return_value = MagicMock(spec=APIBasedExtension)
+        ApiModeration.validate_config("test-tenant-id", api_config)
+        mock_get_extension.assert_called_once_with("test-tenant-id", "test-extension-id")
+
+    def test_validate_config_missing_extension_id(self):
+        config = {
+            "inputs_config": {"enabled": True},
+            "outputs_config": {"enabled": True},
+        }
+        with pytest.raises(ValueError, match="api_based_extension_id is required"):
+            ApiModeration.validate_config("test-tenant-id", config)
+
+    @patch("core.moderation.api.api.ApiModeration._get_api_based_extension")
+    def test_validate_config_extension_not_found(self, mock_get_extension, api_config):
+        mock_get_extension.return_value = None
+        with pytest.raises(ValueError, match="API-based Extension not found"):
+            ApiModeration.validate_config("test-tenant-id", api_config)
+
+    @patch("core.moderation.api.api.ApiModeration._get_config_by_requestor")
+    def test_moderation_for_inputs_enabled(self, mock_get_config, api_moderation):
+        mock_get_config.return_value = {"flagged": True, "action": "direct_output", "preset_response": "Blocked by API"}
+
+        result = api_moderation.moderation_for_inputs(inputs={"q": "a"}, query="hello")
+
+        assert isinstance(result, ModerationInputsResult)
+        assert result.flagged is True
+        assert result.action == ModerationAction.DIRECT_OUTPUT
+        assert result.preset_response == "Blocked by API"
+
+        mock_get_config.assert_called_once_with(
+            APIBasedExtensionPoint.APP_MODERATION_INPUT,
+            {"app_id": "test-app-id", "inputs": {"q": "a"}, "query": "hello"},
+        )
+
+    def test_moderation_for_inputs_disabled(self):
+        config = {
+            "inputs_config": {"enabled": False},
+            "outputs_config": {"enabled": True},
+            "api_based_extension_id": "ext-id",
+        }
+        moderation = ApiModeration("app-id", "tenant-id", config)
+        result = moderation.moderation_for_inputs(inputs={}, query="")
+
+        assert result.flagged is False
+        assert result.action == ModerationAction.DIRECT_OUTPUT
+        assert result.preset_response == ""
+
+    def test_moderation_for_inputs_no_config(self):
+        moderation = ApiModeration("app-id", "tenant-id", None)
+        with pytest.raises(ValueError, match="The config is not set"):
+            moderation.moderation_for_inputs({}, "")
+
+    @patch("core.moderation.api.api.ApiModeration._get_config_by_requestor")
+    def test_moderation_for_outputs_enabled(self, mock_get_config, api_moderation):
+        mock_get_config.return_value = {"flagged": False, "action": "direct_output", "preset_response": ""}
+
+        result = api_moderation.moderation_for_outputs(text="hello world")
+
+        assert isinstance(result, ModerationOutputsResult)
+        assert result.flagged is False
+
+        mock_get_config.assert_called_once_with(
+            APIBasedExtensionPoint.APP_MODERATION_OUTPUT, {"app_id": "test-app-id", "text": "hello world"}
+        )
+
+    def test_moderation_for_outputs_disabled(self):
+        config = {
+            "inputs_config": {"enabled": True},
+            "outputs_config": {"enabled": False},
+            "api_based_extension_id": "ext-id",
+        }
+        moderation = ApiModeration("app-id", "tenant-id", config)
+        result = moderation.moderation_for_outputs(text="test")
+
+        assert result.flagged is False
+        assert result.action == ModerationAction.DIRECT_OUTPUT
+
+    def test_moderation_for_outputs_no_config(self):
+        moderation = ApiModeration("app-id", "tenant-id", None)
+        with pytest.raises(ValueError, match="The config is not set"):
+            moderation.moderation_for_outputs("test")
+
+    @patch("core.moderation.api.api.ApiModeration._get_api_based_extension")
+    @patch("core.moderation.api.api.decrypt_token")
+    @patch("core.moderation.api.api.APIBasedExtensionRequestor")
+    def test_get_config_by_requestor_success(self, mock_requestor_cls, mock_decrypt, mock_get_ext, api_moderation):
+        mock_ext = MagicMock(spec=APIBasedExtension)
+        mock_ext.api_endpoint = "http://api.test"
+        mock_ext.api_key = "encrypted-key"
+        mock_get_ext.return_value = mock_ext
+
+        mock_decrypt.return_value = "decrypted-key"
+
+        mock_requestor = MagicMock()
+        mock_requestor.request.return_value = {"flagged": True}
+        mock_requestor_cls.return_value = mock_requestor
+
+        params = {"some": "params"}
+        result = api_moderation._get_config_by_requestor(APIBasedExtensionPoint.APP_MODERATION_INPUT, params)
+
+        assert result == {"flagged": True}
+        mock_get_ext.assert_called_once_with("test-tenant-id", "test-extension-id")
+        mock_decrypt.assert_called_once_with("test-tenant-id", "encrypted-key")
+        mock_requestor_cls.assert_called_once_with("http://api.test", "decrypted-key")
+        mock_requestor.request.assert_called_once_with(APIBasedExtensionPoint.APP_MODERATION_INPUT, params)
+
+    def test_get_config_by_requestor_no_config(self):
+        moderation = ApiModeration("app-id", "tenant-id", None)
+        with pytest.raises(ValueError, match="The config is not set"):
+            moderation._get_config_by_requestor(APIBasedExtensionPoint.APP_MODERATION_INPUT, {})
+
+    @patch("core.moderation.api.api.ApiModeration._get_api_based_extension")
+    def test_get_config_by_requestor_extension_not_found(self, mock_get_ext, api_moderation):
+        mock_get_ext.return_value = None
+        with pytest.raises(ValueError, match="API-based Extension not found"):
+            api_moderation._get_config_by_requestor(APIBasedExtensionPoint.APP_MODERATION_INPUT, {})
+
+    @patch("core.moderation.api.api.db.session.scalar")
+    def test_get_api_based_extension(self, mock_scalar):
+        mock_ext = MagicMock(spec=APIBasedExtension)
+        mock_scalar.return_value = mock_ext
+
+        result = ApiModeration._get_api_based_extension("tenant-1", "ext-1")
+
+        assert result == mock_ext
+        mock_scalar.assert_called_once()
+        # Verify the call has the correct filters
+        args, kwargs = mock_scalar.call_args
+        stmt = args[0]
+        # We can't easily inspect the statement without complex sqlalchemy tricks,
+        # but calling it is usually enough for unit tests if we mock the result.

api/tests/unit_tests/core/moderation/test_input_moderation.py

@@ -0,0 +1,207 @@
+from unittest.mock import MagicMock, patch
+
+import pytest
+
+from core.app.app_config.entities import AppConfig, SensitiveWordAvoidanceEntity
+from core.moderation.base import ModerationAction, ModerationError, ModerationInputsResult
+from core.moderation.input_moderation import InputModeration
+from core.ops.entities.trace_entity import TraceTaskName
+from core.ops.ops_trace_manager import TraceQueueManager
+
+
+class TestInputModeration:
+    @pytest.fixture
+    def app_config(self):
+        config = MagicMock(spec=AppConfig)
+        config.sensitive_word_avoidance = None
+        return config
+
+    @pytest.fixture
+    def input_moderation(self):
+        return InputModeration()
+
+    def test_check_no_sensitive_word_avoidance(self, app_config, input_moderation):
+        app_id = "test_app_id"
+        tenant_id = "test_tenant_id"
+        inputs = {"input_key": "input_value"}
+        query = "test query"
+        message_id = "test_message_id"
+
+        flagged, final_inputs, final_query = input_moderation.check(
+            app_id=app_id, tenant_id=tenant_id, app_config=app_config, inputs=inputs, query=query, message_id=message_id
+        )
+
+        assert flagged is False
+        assert final_inputs == inputs
+        assert final_query == query
+
+    @patch("core.moderation.input_moderation.ModerationFactory")
+    def test_check_not_flagged(self, mock_factory_cls, app_config, input_moderation):
+        app_id = "test_app_id"
+        tenant_id = "test_tenant_id"
+        inputs = {"input_key": "input_value"}
+        query = "test query"
+        message_id = "test_message_id"
+
+        # Setup config
+        sensitive_word_config = MagicMock(spec=SensitiveWordAvoidanceEntity)
+        sensitive_word_config.type = "keywords"
+        sensitive_word_config.config = {"keywords": ["bad"]}
+        app_config.sensitive_word_avoidance = sensitive_word_config
+
+        # Setup factory mock
+        mock_factory = mock_factory_cls.return_value
+        mock_result = ModerationInputsResult(flagged=False, action=ModerationAction.DIRECT_OUTPUT)
+        mock_factory.moderation_for_inputs.return_value = mock_result
+
+        flagged, final_inputs, final_query = input_moderation.check(
+            app_id=app_id, tenant_id=tenant_id, app_config=app_config, inputs=inputs, query=query, message_id=message_id
+        )
+
+        assert flagged is False
+        assert final_inputs == inputs
+        assert final_query == query
+        mock_factory_cls.assert_called_once_with(
+            name="keywords", app_id=app_id, tenant_id=tenant_id, config={"keywords": ["bad"]}
+        )
+        mock_factory.moderation_for_inputs.assert_called_once_with(dict(inputs), query)
+
+    @patch("core.moderation.input_moderation.ModerationFactory")
+    @patch("core.moderation.input_moderation.TraceTask")
+    def test_check_with_trace_manager(self, mock_trace_task, mock_factory_cls, app_config, input_moderation):
+        app_id = "test_app_id"
+        tenant_id = "test_tenant_id"
+        inputs = {"input_key": "input_value"}
+        query = "test query"
+        message_id = "test_message_id"
+        trace_manager = MagicMock(spec=TraceQueueManager)
+
+        # Setup config
+        sensitive_word_config = MagicMock(spec=SensitiveWordAvoidanceEntity)
+        sensitive_word_config.type = "keywords"
+        sensitive_word_config.config = {}
+        app_config.sensitive_word_avoidance = sensitive_word_config
+
+        # Setup factory mock
+        mock_factory = mock_factory_cls.return_value
+        mock_result = ModerationInputsResult(flagged=False, action=ModerationAction.DIRECT_OUTPUT)
+        mock_factory.moderation_for_inputs.return_value = mock_result
+
+        input_moderation.check(
+            app_id=app_id,
+            tenant_id=tenant_id,
+            app_config=app_config,
+            inputs=inputs,
+            query=query,
+            message_id=message_id,
+            trace_manager=trace_manager,
+        )
+
+        trace_manager.add_trace_task.assert_called_once_with(mock_trace_task.return_value)
+        mock_trace_task.assert_called_once()
+        call_kwargs = mock_trace_task.call_args.kwargs
+        call_args = mock_trace_task.call_args.args
+        assert call_args[0] == TraceTaskName.MODERATION_TRACE
+        assert call_kwargs["message_id"] == message_id
+        assert call_kwargs["moderation_result"] == mock_result
+        assert call_kwargs["inputs"] == inputs
+        assert "timer" in call_kwargs
+
+    @patch("core.moderation.input_moderation.ModerationFactory")
+    def test_check_flagged_direct_output(self, mock_factory_cls, app_config, input_moderation):
+        app_id = "test_app_id"
+        tenant_id = "test_tenant_id"
+        inputs = {"input_key": "input_value"}
+        query = "test query"
+        message_id = "test_message_id"
+
+        # Setup config
+        sensitive_word_config = MagicMock(spec=SensitiveWordAvoidanceEntity)
+        sensitive_word_config.type = "keywords"
+        sensitive_word_config.config = {}
+        app_config.sensitive_word_avoidance = sensitive_word_config
+
+        # Setup factory mock
+        mock_factory = mock_factory_cls.return_value
+        mock_result = ModerationInputsResult(
+            flagged=True, action=ModerationAction.DIRECT_OUTPUT, preset_response="Blocked content"
+        )
+        mock_factory.moderation_for_inputs.return_value = mock_result
+
+        with pytest.raises(ModerationError) as excinfo:
+            input_moderation.check(
+                app_id=app_id,
+                tenant_id=tenant_id,
+                app_config=app_config,
+                inputs=inputs,
+                query=query,
+                message_id=message_id,
+            )
+
+        assert str(excinfo.value) == "Blocked content"
+
+    @patch("core.moderation.input_moderation.ModerationFactory")
+    def test_check_flagged_overridden(self, mock_factory_cls, app_config, input_moderation):
+        app_id = "test_app_id"
+        tenant_id = "test_tenant_id"
+        inputs = {"input_key": "input_value"}
+        query = "test query"
+        message_id = "test_message_id"
+
+        # Setup config
+        sensitive_word_config = MagicMock(spec=SensitiveWordAvoidanceEntity)
+        sensitive_word_config.type = "keywords"
+        sensitive_word_config.config = {}
+        app_config.sensitive_word_avoidance = sensitive_word_config
+
+        # Setup factory mock
+        mock_factory = mock_factory_cls.return_value
+        mock_result = ModerationInputsResult(
+            flagged=True,
+            action=ModerationAction.OVERRIDDEN,
+            inputs={"input_key": "overridden_value"},
+            query="overridden query",
+        )
+        mock_factory.moderation_for_inputs.return_value = mock_result
+
+        flagged, final_inputs, final_query = input_moderation.check(
+            app_id=app_id, tenant_id=tenant_id, app_config=app_config, inputs=inputs, query=query, message_id=message_id
+        )
+
+        assert flagged is True
+        assert final_inputs == {"input_key": "overridden_value"}
+        assert final_query == "overridden query"
+
+    @patch("core.moderation.input_moderation.ModerationFactory")
+    def test_check_flagged_other_action(self, mock_factory_cls, app_config, input_moderation):
+        app_id = "test_app_id"
+        tenant_id = "test_tenant_id"
+        inputs = {"input_key": "input_value"}
+        query = "test query"
+        message_id = "test_message_id"
+
+        # Setup config
+        sensitive_word_config = MagicMock(spec=SensitiveWordAvoidanceEntity)
+        sensitive_word_config.type = "keywords"
+        sensitive_word_config.config = {}
+        app_config.sensitive_word_avoidance = sensitive_word_config
+
+        # Setup factory mock
+        mock_factory = mock_factory_cls.return_value
+        mock_result = MagicMock()
+        mock_result.flagged = True
+        mock_result.action = "NONE"  # Some other action
+        mock_factory.moderation_for_inputs.return_value = mock_result
+
+        flagged, final_inputs, final_query = input_moderation.check(
+            app_id=app_id,
+            tenant_id=tenant_id,
+            app_config=app_config,
+            inputs=inputs,
+            query=query,
+            message_id=message_id,
+        )
+
+        assert flagged is True
+        assert final_inputs == inputs
+        assert final_query == query

api/tests/unit_tests/core/moderation/test_output_moderation.py

@@ -0,0 +1,234 @@
+from unittest.mock import MagicMock, patch
+
+import pytest
+from flask import Flask
+
+from core.app.apps.base_app_queue_manager import AppQueueManager, PublishFrom
+from core.app.entities.queue_entities import QueueMessageReplaceEvent
+from core.moderation.base import ModerationAction, ModerationOutputsResult
+from core.moderation.output_moderation import ModerationRule, OutputModeration
+
+
+class TestOutputModeration:
+    @pytest.fixture
+    def mock_queue_manager(self):
+        return MagicMock(spec=AppQueueManager)
+
+    @pytest.fixture
+    def moderation_rule(self):
+        return ModerationRule(type="keywords", config={"keywords": "badword"})
+
+    @pytest.fixture
+    def output_moderation(self, mock_queue_manager, moderation_rule):
+        return OutputModeration(
+            tenant_id="test_tenant", app_id="test_app", rule=moderation_rule, queue_manager=mock_queue_manager
+        )
+
+    def test_should_direct_output(self, output_moderation):
+        assert output_moderation.should_direct_output() is False
+        output_moderation.final_output = "blocked"
+        assert output_moderation.should_direct_output() is True
+
+    def test_get_final_output(self, output_moderation):
+        assert output_moderation.get_final_output() == ""
+        output_moderation.final_output = "blocked"
+        assert output_moderation.get_final_output() == "blocked"
+
+    def test_append_new_token(self, output_moderation):
+        with patch.object(OutputModeration, "start_thread") as mock_start:
+            output_moderation.append_new_token("hello")
+            assert output_moderation.buffer == "hello"
+            mock_start.assert_called_once()
+
+            output_moderation.thread = MagicMock()
+            output_moderation.append_new_token(" world")
+            assert output_moderation.buffer == "hello world"
+            assert mock_start.call_count == 1
+
+    def test_moderation_completion_no_flag(self, output_moderation):
+        with patch.object(OutputModeration, "moderation") as mock_moderation:
+            mock_moderation.return_value = ModerationOutputsResult(flagged=False, action=ModerationAction.DIRECT_OUTPUT)
+
+            output, flagged = output_moderation.moderation_completion("safe content")
+
+            assert output == "safe content"
+            assert flagged is False
+            assert output_moderation.is_final_chunk is True
+
+    def test_moderation_completion_flagged_direct_output(self, output_moderation, mock_queue_manager):
+        with patch.object(OutputModeration, "moderation") as mock_moderation:
+            mock_moderation.return_value = ModerationOutputsResult(
+                flagged=True, action=ModerationAction.DIRECT_OUTPUT, preset_response="preset"
+            )
+
+            output, flagged = output_moderation.moderation_completion("badword content", public_event=True)
+
+            assert output == "preset"
+            assert flagged is True
+            mock_queue_manager.publish.assert_called_once()
+            args, _ = mock_queue_manager.publish.call_args
+            assert isinstance(args[0], QueueMessageReplaceEvent)
+            assert args[0].text == "preset"
+            assert args[1] == PublishFrom.TASK_PIPELINE
+
+    def test_moderation_completion_flagged_overridden(self, output_moderation, mock_queue_manager):
+        with patch.object(OutputModeration, "moderation") as mock_moderation:
+            mock_moderation.return_value = ModerationOutputsResult(
+                flagged=True, action=ModerationAction.OVERRIDDEN, text="masked content"
+            )
+
+            output, flagged = output_moderation.moderation_completion("badword content", public_event=True)
+
+            assert output == "masked content"
+            assert flagged is True
+            mock_queue_manager.publish.assert_called_once()
+            args, _ = mock_queue_manager.publish.call_args
+            assert args[0].text == "masked content"
+
+    def test_start_thread(self, output_moderation):
+        mock_app = MagicMock(spec=Flask)
+        with patch("core.moderation.output_moderation.current_app") as mock_current_app:
+            mock_current_app._get_current_object.return_value = mock_app
+            with patch("threading.Thread") as mock_thread_class:
+                mock_thread_instance = MagicMock()
+                mock_thread_class.return_value = mock_thread_instance
+
+                thread = output_moderation.start_thread()
+
+                assert thread == mock_thread_instance
+                mock_thread_class.assert_called_once()
+                mock_thread_instance.start.assert_called_once()
+
+    def test_stop_thread(self, output_moderation):
+        mock_thread = MagicMock()
+        mock_thread.is_alive.return_value = True
+        output_moderation.thread = mock_thread
+
+        output_moderation.stop_thread()
+        assert output_moderation.thread_running is False
+
+        output_moderation.thread_running = True
+        mock_thread.is_alive.return_value = False
+        output_moderation.stop_thread()
+        assert output_moderation.thread_running is True
+
+    @patch("core.moderation.output_moderation.ModerationFactory")
+    def test_moderation_success(self, mock_factory_class, output_moderation):
+        mock_factory = mock_factory_class.return_value
+        mock_result = ModerationOutputsResult(flagged=False, action=ModerationAction.DIRECT_OUTPUT)
+        mock_factory.moderation_for_outputs.return_value = mock_result
+
+        result = output_moderation.moderation("tenant", "app", "buffer")
+
+        assert result == mock_result
+        mock_factory_class.assert_called_once_with(
+            name="keywords", app_id="app", tenant_id="tenant", config={"keywords": "badword"}
+        )
+
+    @patch("core.moderation.output_moderation.ModerationFactory")
+    def test_moderation_exception(self, mock_factory_class, output_moderation):
+        mock_factory_class.side_effect = Exception("error")
+
+        result = output_moderation.moderation("tenant", "app", "buffer")
+        assert result is None
+
+    def test_worker_loop_and_exit(self, output_moderation, mock_queue_manager):
+        mock_app = MagicMock(spec=Flask)
+
+        # Test exit on thread_running=False
+        output_moderation.thread_running = False
+        output_moderation.worker(mock_app, 10)
+        # Should exit immediately
+
+    def test_worker_no_flag(self, output_moderation):
+        mock_app = MagicMock(spec=Flask)
+
+        with patch.object(OutputModeration, "moderation") as mock_moderation:
+            mock_moderation.return_value = ModerationOutputsResult(flagged=False, action=ModerationAction.DIRECT_OUTPUT)
+
+            output_moderation.buffer = "safe"
+            output_moderation.is_final_chunk = True
+
+            # To avoid infinite loop, we'll set thread_running to False after one iteration
+            def side_effect(*args, **kwargs):
+                output_moderation.thread_running = False
+                return mock_moderation.return_value
+
+            mock_moderation.side_effect = side_effect
+
+            output_moderation.worker(mock_app, 10)
+
+            assert mock_moderation.called
+
+    def test_worker_flagged_direct_output(self, output_moderation, mock_queue_manager):
+        mock_app = MagicMock(spec=Flask)
+
+        with patch.object(OutputModeration, "moderation") as mock_moderation:
+            mock_moderation.return_value = ModerationOutputsResult(
+                flagged=True, action=ModerationAction.DIRECT_OUTPUT, preset_response="preset"
+            )
+
+            output_moderation.buffer = "badword"
+            output_moderation.is_final_chunk = True
+
+            output_moderation.worker(mock_app, 10)
+
+            assert output_moderation.final_output == "preset"
+            mock_queue_manager.publish.assert_called_once()
+            # It breaks on DIRECT_OUTPUT
+
+    def test_worker_flagged_overridden(self, output_moderation, mock_queue_manager):
+        mock_app = MagicMock(spec=Flask)
+
+        with patch.object(OutputModeration, "moderation") as mock_moderation:
+            # Use side_effect to change thread_running on second call
+            def side_effect(*args, **kwargs):
+                if mock_moderation.call_count > 1:
+                    output_moderation.thread_running = False
+                    return None
+                return ModerationOutputsResult(flagged=True, action=ModerationAction.OVERRIDDEN, text="masked")
+
+            mock_moderation.side_effect = side_effect
+
+            output_moderation.buffer = "badword"
+            output_moderation.is_final_chunk = True
+
+            output_moderation.worker(mock_app, 10)
+
+            mock_queue_manager.publish.assert_called_once()
+            args, _ = mock_queue_manager.publish.call_args
+            assert args[0].text == "masked"
+
+    def test_worker_chunk_too_small(self, output_moderation):
+        mock_app = MagicMock(spec=Flask)
+        with patch("time.sleep") as mock_sleep:
+            # chunk_length < buffer_size and not is_final_chunk
+            output_moderation.buffer = "123"  # length 3
+            output_moderation.is_final_chunk = False
+
+            def sleep_side_effect(seconds):
+                output_moderation.thread_running = False
+
+            mock_sleep.side_effect = sleep_side_effect
+
+            output_moderation.worker(mock_app, 10)  # buffer_size 10
+
+            mock_sleep.assert_called_once_with(1)
+
+    def test_worker_empty_not_flagged(self, output_moderation, mock_queue_manager):
+        mock_app = MagicMock(spec=Flask)
+        with patch.object(OutputModeration, "moderation") as mock_moderation:
+            # Return None (exception or no rule)
+            mock_moderation.return_value = None
+
+            def side_effect(*args, **kwargs):
+                output_moderation.thread_running = False
+
+            mock_moderation.side_effect = side_effect
+
+            output_moderation.buffer = "something"
+            output_moderation.is_final_chunk = True
+
+            output_moderation.worker(mock_app, 10)
+
+            mock_queue_manager.publish.assert_not_called()

api/tests/unit_tests/core/rag/datasource/vdb/tidb_on_qdrant/test_tidb_on_qdrant_vector.py

@@ -0,0 +1,160 @@
+from unittest.mock import patch
+
+import httpx
+import pytest
+from qdrant_client.http import models as rest
+from qdrant_client.http.exceptions import UnexpectedResponse
+
+from core.rag.datasource.vdb.tidb_on_qdrant.tidb_on_qdrant_vector import (
+    TidbOnQdrantConfig,
+    TidbOnQdrantVector,
+)
+
+
+class TestTidbOnQdrantVectorDeleteByIds:
+    """Unit tests for TidbOnQdrantVector.delete_by_ids method."""
+
+    @pytest.fixture
+    def vector_instance(self):
+        """Create a TidbOnQdrantVector instance for testing."""
+        config = TidbOnQdrantConfig(
+            endpoint="http://localhost:6333",
+            api_key="test_api_key",
+        )
+
+        with patch("core.rag.datasource.vdb.tidb_on_qdrant.tidb_on_qdrant_vector.qdrant_client.QdrantClient"):
+            vector = TidbOnQdrantVector(
+                collection_name="test_collection",
+                group_id="test_group",
+                config=config,
+            )
+            return vector
+
+    def test_delete_by_ids_with_multiple_ids(self, vector_instance):
+        """Test batch deletion with multiple document IDs."""
+        ids = ["doc1", "doc2", "doc3"]
+
+        vector_instance.delete_by_ids(ids)
+
+        # Verify that delete was called once with MatchAny filter
+        vector_instance._client.delete.assert_called_once()
+        call_args = vector_instance._client.delete.call_args
+
+        # Check collection name
+        assert call_args[1]["collection_name"] == "test_collection"
+
+        # Verify filter uses MatchAny with all IDs
+        filter_selector = call_args[1]["points_selector"]
+        filter_obj = filter_selector.filter
+        assert len(filter_obj.must) == 1
+
+        field_condition = filter_obj.must[0]
+        assert field_condition.key == "metadata.doc_id"
+        assert isinstance(field_condition.match, rest.MatchAny)
+        assert set(field_condition.match.any) == {"doc1", "doc2", "doc3"}
+
+    def test_delete_by_ids_with_single_id(self, vector_instance):
+        """Test deletion with a single document ID."""
+        ids = ["doc1"]
+
+        vector_instance.delete_by_ids(ids)
+
+        # Verify that delete was called once
+        vector_instance._client.delete.assert_called_once()
+        call_args = vector_instance._client.delete.call_args
+
+        # Verify filter uses MatchAny with single ID
+        filter_selector = call_args[1]["points_selector"]
+        filter_obj = filter_selector.filter
+        field_condition = filter_obj.must[0]
+        assert isinstance(field_condition.match, rest.MatchAny)
+        assert field_condition.match.any == ["doc1"]
+
+    def test_delete_by_ids_with_empty_list(self, vector_instance):
+        """Test deletion with empty ID list returns early without API call."""
+        vector_instance.delete_by_ids([])
+
+        # Verify that delete was NOT called
+        vector_instance._client.delete.assert_not_called()
+
+    def test_delete_by_ids_with_404_error(self, vector_instance):
+        """Test that 404 errors (collection not found) are handled gracefully."""
+        ids = ["doc1", "doc2"]
+
+        # Mock a 404 error
+        error = UnexpectedResponse(
+            status_code=404,
+            reason_phrase="Not Found",
+            content=b"Collection not found",
+            headers=httpx.Headers(),
+        )
+        vector_instance._client.delete.side_effect = error
+
+        # Should not raise an exception
+        vector_instance.delete_by_ids(ids)
+
+        # Verify delete was called
+        vector_instance._client.delete.assert_called_once()
+
+    def test_delete_by_ids_with_unexpected_error(self, vector_instance):
+        """Test that non-404 errors are re-raised."""
+        ids = ["doc1", "doc2"]
+
+        # Mock a 500 error
+        error = UnexpectedResponse(
+            status_code=500,
+            reason_phrase="Internal Server Error",
+            content=b"Server error",
+            headers=httpx.Headers(),
+        )
+        vector_instance._client.delete.side_effect = error
+
+        # Should re-raise the exception
+        with pytest.raises(UnexpectedResponse) as exc_info:
+            vector_instance.delete_by_ids(ids)
+
+        assert exc_info.value.status_code == 500
+
+    def test_delete_by_ids_with_large_batch(self, vector_instance):
+        """Test deletion with a large batch of IDs."""
+        # Create 1000 IDs
+        ids = [f"doc_{i}" for i in range(1000)]
+
+        vector_instance.delete_by_ids(ids)
+
+        # Verify single delete call with all IDs
+        vector_instance._client.delete.assert_called_once()
+        call_args = vector_instance._client.delete.call_args
+
+        filter_selector = call_args[1]["points_selector"]
+        filter_obj = filter_selector.filter
+        field_condition = filter_obj.must[0]
+
+        # Verify all 1000 IDs are in the batch
+        assert len(field_condition.match.any) == 1000
+        assert "doc_0" in field_condition.match.any
+        assert "doc_999" in field_condition.match.any
+
+    def test_delete_by_ids_filter_structure(self, vector_instance):
+        """Test that the filter structure is correctly constructed."""
+        ids = ["doc1", "doc2"]
+
+        vector_instance.delete_by_ids(ids)
+
+        call_args = vector_instance._client.delete.call_args
+        filter_selector = call_args[1]["points_selector"]
+        filter_obj = filter_selector.filter
+
+        # Verify Filter structure
+        assert isinstance(filter_obj, rest.Filter)
+        assert filter_obj.must is not None
+        assert len(filter_obj.must) == 1
+
+        # Verify FieldCondition structure
+        field_condition = filter_obj.must[0]
+        assert isinstance(field_condition, rest.FieldCondition)
+        assert field_condition.key == "metadata.doc_id"
+
+        # Verify MatchAny structure
+        assert isinstance(field_condition.match, rest.MatchAny)
+        assert field_condition.match.any == ids

api/tests/unit_tests/core/rag/datasource/vdb/weaviate/test_weavaite.py

@@ -0,0 +1,33 @@
+from unittest.mock import MagicMock, patch
+
+from core.rag.datasource.vdb.weaviate.weaviate_vector import WeaviateConfig, WeaviateVector
+
+
+def test_init_client_with_valid_config():
+    """Test successful client initialization with valid configuration."""
+    config = WeaviateConfig(
+        endpoint="http://localhost:8080",
+        api_key="WVF5YThaHlkYwhGUSmCRgsX3tD5ngdN8pkih",
+    )
+
+    with patch("weaviate.connect_to_custom") as mock_connect:
+        mock_client = MagicMock()
+        mock_client.is_ready.return_value = True
+        mock_connect.return_value = mock_client
+
+        vector = WeaviateVector(
+            collection_name="test_collection",
+            config=config,
+            attributes=["doc_id"],
+        )
+
+        assert vector._client == mock_client
+        mock_connect.assert_called_once()
+        call_kwargs = mock_connect.call_args[1]
+        assert call_kwargs["http_host"] == "localhost"
+        assert call_kwargs["http_port"] == 8080
+        assert call_kwargs["http_secure"] is False
+        assert call_kwargs["grpc_host"] == "localhost"
+        assert call_kwargs["grpc_port"] == 50051
+        assert call_kwargs["grpc_secure"] is False
+        assert call_kwargs["auth_credentials"] is not None

api/tests/unit_tests/core/rag/datasource/vdb/weaviate/test_weaviate_vector.py

@@ -0,0 +1,335 @@
+"""Unit tests for Weaviate vector database implementation.
+
+Focuses on verifying that doc_type is properly handled in:
+- Collection schema creation (_create_collection)
+- Property migration (_ensure_properties)
+- Vector search result metadata (search_by_vector)
+- Full-text search result metadata (search_by_full_text)
+"""
+
+import unittest
+from types import SimpleNamespace
+from unittest.mock import MagicMock, patch
+
+from core.rag.datasource.vdb.weaviate import weaviate_vector as weaviate_vector_module
+from core.rag.datasource.vdb.weaviate.weaviate_vector import WeaviateConfig, WeaviateVector
+from core.rag.models.document import Document
+
+
+class TestWeaviateVector(unittest.TestCase):
+    """Tests for WeaviateVector class with focus on doc_type metadata handling."""
+
+    def setUp(self):
+        weaviate_vector_module._weaviate_client = None
+        self.config = WeaviateConfig(
+            endpoint="http://localhost:8080",
+            api_key="test-key",
+            batch_size=100,
+        )
+        self.collection_name = "Test_Collection_Node"
+        self.attributes = ["doc_id", "dataset_id", "document_id", "doc_hash", "doc_type"]
+
+    def tearDown(self):
+        weaviate_vector_module._weaviate_client = None
+
+    @patch("core.rag.datasource.vdb.weaviate.weaviate_vector.weaviate")
+    def _create_weaviate_vector(self, mock_weaviate_module):
+        """Helper to create a WeaviateVector instance with mocked client."""
+        mock_client = MagicMock()
+        mock_client.is_ready.return_value = True
+        mock_weaviate_module.connect_to_custom.return_value = mock_client
+
+        wv = WeaviateVector(
+            collection_name=self.collection_name,
+            config=self.config,
+            attributes=self.attributes,
+        )
+        return wv, mock_client
+
+    @patch("core.rag.datasource.vdb.weaviate.weaviate_vector.weaviate")
+    def test_init(self, mock_weaviate_module):
+        """Test WeaviateVector initialization stores attributes including doc_type."""
+        mock_client = MagicMock()
+        mock_client.is_ready.return_value = True
+        mock_weaviate_module.connect_to_custom.return_value = mock_client
+
+        wv = WeaviateVector(
+            collection_name=self.collection_name,
+            config=self.config,
+            attributes=self.attributes,
+        )
+
+        assert wv._collection_name == self.collection_name
+        assert "doc_type" in wv._attributes
+
+    @patch("core.rag.datasource.vdb.weaviate.weaviate_vector.redis_client")
+    @patch("core.rag.datasource.vdb.weaviate.weaviate_vector.dify_config")
+    @patch("core.rag.datasource.vdb.weaviate.weaviate_vector.weaviate")
+    def test_create_collection_includes_doc_type_property(self, mock_weaviate_module, mock_dify_config, mock_redis):
+        """Test that _create_collection defines doc_type in the schema properties."""
+        # Mock Redis
+        mock_lock = MagicMock()
+        mock_lock.__enter__ = MagicMock()
+        mock_lock.__exit__ = MagicMock()
+        mock_redis.lock.return_value = mock_lock
+        mock_redis.get.return_value = None
+        mock_redis.set.return_value = None
+
+        # Mock dify_config
+        mock_dify_config.WEAVIATE_TOKENIZATION = None
+
+        # Mock client
+        mock_client = MagicMock()
+        mock_client.is_ready.return_value = True
+        mock_weaviate_module.connect_to_custom.return_value = mock_client
+        mock_client.collections.exists.return_value = False
+
+        # Mock _ensure_properties to avoid side effects
+        mock_col = MagicMock()
+        mock_client.collections.use.return_value = mock_col
+        mock_cfg = MagicMock()
+        mock_cfg.properties = []
+        mock_col.config.get.return_value = mock_cfg
+
+        wv = WeaviateVector(
+            collection_name=self.collection_name,
+            config=self.config,
+            attributes=self.attributes,
+        )
+        wv._create_collection()
+
+        # Verify collections.create was called
+        mock_client.collections.create.assert_called_once()
+
+        # Extract properties from the create call
+        call_kwargs = mock_client.collections.create.call_args
+        properties = call_kwargs.kwargs.get("properties")
+
+        # Verify doc_type is among the defined properties
+        property_names = [p.name for p in properties]
+        assert "doc_type" in property_names, (
+            f"doc_type should be in collection schema properties, got: {property_names}"
+        )
+
+    @patch("core.rag.datasource.vdb.weaviate.weaviate_vector.weaviate")
+    def test_ensure_properties_adds_missing_doc_type(self, mock_weaviate_module):
+        """Test that _ensure_properties adds doc_type when it's missing from existing schema."""
+        mock_client = MagicMock()
+        mock_client.is_ready.return_value = True
+        mock_weaviate_module.connect_to_custom.return_value = mock_client
+
+        # Collection exists but doc_type property is missing
+        mock_client.collections.exists.return_value = True
+        mock_col = MagicMock()
+        mock_client.collections.use.return_value = mock_col
+
+        # Simulate existing properties WITHOUT doc_type
+        existing_props = [
+            SimpleNamespace(name="text"),
+            SimpleNamespace(name="document_id"),
+            SimpleNamespace(name="doc_id"),
+            SimpleNamespace(name="chunk_index"),
+        ]
+        mock_cfg = MagicMock()
+        mock_cfg.properties = existing_props
+        mock_col.config.get.return_value = mock_cfg
+
+        wv = WeaviateVector(
+            collection_name=self.collection_name,
+            config=self.config,
+            attributes=self.attributes,
+        )
+        wv._ensure_properties()
+
+        # Verify add_property was called and includes doc_type
+        add_calls = mock_col.config.add_property.call_args_list
+        added_names = [call.args[0].name for call in add_calls]
+        assert "doc_type" in added_names, f"doc_type should be added to existing collection, added: {added_names}"
+
+    @patch("core.rag.datasource.vdb.weaviate.weaviate_vector.weaviate")
+    def test_ensure_properties_skips_existing_doc_type(self, mock_weaviate_module):
+        """Test that _ensure_properties does not add doc_type when it already exists."""
+        mock_client = MagicMock()
+        mock_client.is_ready.return_value = True
+        mock_weaviate_module.connect_to_custom.return_value = mock_client
+
+        mock_client.collections.exists.return_value = True
+        mock_col = MagicMock()
+        mock_client.collections.use.return_value = mock_col
+
+        # Simulate existing properties WITH doc_type already present
+        existing_props = [
+            SimpleNamespace(name="text"),
+            SimpleNamespace(name="document_id"),
+            SimpleNamespace(name="doc_id"),
+            SimpleNamespace(name="doc_type"),
+            SimpleNamespace(name="chunk_index"),
+        ]
+        mock_cfg = MagicMock()
+        mock_cfg.properties = existing_props
+        mock_col.config.get.return_value = mock_cfg
+
+        wv = WeaviateVector(
+            collection_name=self.collection_name,
+            config=self.config,
+            attributes=self.attributes,
+        )
+        wv._ensure_properties()
+
+        # No properties should be added
+        mock_col.config.add_property.assert_not_called()
+
+    @patch("core.rag.datasource.vdb.weaviate.weaviate_vector.weaviate")
+    def test_search_by_vector_returns_doc_type_in_metadata(self, mock_weaviate_module):
+        """Test that search_by_vector returns doc_type in document metadata.
+
+        This is the core bug fix verification: when doc_type is in _attributes,
+        it should appear in return_properties and thus be included in results.
+        """
+        mock_client = MagicMock()
+        mock_client.is_ready.return_value = True
+        mock_weaviate_module.connect_to_custom.return_value = mock_client
+
+        mock_client.collections.exists.return_value = True
+        mock_col = MagicMock()
+        mock_client.collections.use.return_value = mock_col
+
+        # Simulate search result with doc_type in properties
+        mock_obj = MagicMock()
+        mock_obj.properties = {
+            "text": "image content description",
+            "doc_id": "upload_file_id_123",
+            "dataset_id": "dataset_1",
+            "document_id": "doc_1",
+            "doc_hash": "hash_abc",
+            "doc_type": "image",
+        }
+        mock_obj.metadata.distance = 0.1
+
+        mock_result = MagicMock()
+        mock_result.objects = [mock_obj]
+        mock_col.query.near_vector.return_value = mock_result
+
+        wv = WeaviateVector(
+            collection_name=self.collection_name,
+            config=self.config,
+            attributes=self.attributes,
+        )
+        docs = wv.search_by_vector(query_vector=[0.1] * 128, top_k=1)
+
+        # Verify doc_type is in return_properties
+        call_kwargs = mock_col.query.near_vector.call_args
+        return_props = call_kwargs.kwargs.get("return_properties")
+        assert "doc_type" in return_props, f"doc_type should be in return_properties, got: {return_props}"
+
+        # Verify doc_type is in result metadata
+        assert len(docs) == 1
+        assert docs[0].metadata.get("doc_type") == "image"
+
+    @patch("core.rag.datasource.vdb.weaviate.weaviate_vector.weaviate")
+    def test_search_by_full_text_returns_doc_type_in_metadata(self, mock_weaviate_module):
+        """Test that search_by_full_text also returns doc_type in document metadata."""
+        mock_client = MagicMock()
+        mock_client.is_ready.return_value = True
+        mock_weaviate_module.connect_to_custom.return_value = mock_client
+
+        mock_client.collections.exists.return_value = True
+        mock_col = MagicMock()
+        mock_client.collections.use.return_value = mock_col
+
+        # Simulate BM25 search result with doc_type
+        mock_obj = MagicMock()
+        mock_obj.properties = {
+            "text": "image content description",
+            "doc_id": "upload_file_id_456",
+            "doc_type": "image",
+        }
+        mock_obj.vector = {"default": [0.1] * 128}
+
+        mock_result = MagicMock()
+        mock_result.objects = [mock_obj]
+        mock_col.query.bm25.return_value = mock_result
+
+        wv = WeaviateVector(
+            collection_name=self.collection_name,
+            config=self.config,
+            attributes=self.attributes,
+        )
+        docs = wv.search_by_full_text(query="image", top_k=1)
+
+        # Verify doc_type is in return_properties
+        call_kwargs = mock_col.query.bm25.call_args
+        return_props = call_kwargs.kwargs.get("return_properties")
+        assert "doc_type" in return_props, (
+            f"doc_type should be in return_properties for BM25 search, got: {return_props}"
+        )
+
+        # Verify doc_type is in result metadata
+        assert len(docs) == 1
+        assert docs[0].metadata.get("doc_type") == "image"
+
+    @patch("core.rag.datasource.vdb.weaviate.weaviate_vector.weaviate")
+    def test_add_texts_stores_doc_type_in_properties(self, mock_weaviate_module):
+        """Test that add_texts includes doc_type from document metadata in stored properties."""
+        mock_client = MagicMock()
+        mock_client.is_ready.return_value = True
+        mock_weaviate_module.connect_to_custom.return_value = mock_client
+
+        mock_col = MagicMock()
+        mock_client.collections.use.return_value = mock_col
+
+        # Create a document with doc_type metadata (as produced by multimodal indexing)
+        doc = Document(
+            page_content="an image of a cat",
+            metadata={
+                "doc_id": "upload_file_123",
+                "doc_type": "image",
+                "dataset_id": "ds_1",
+                "document_id": "doc_1",
+                "doc_hash": "hash_xyz",
+            },
+        )
+
+        wv = WeaviateVector(
+            collection_name=self.collection_name,
+            config=self.config,
+            attributes=self.attributes,
+        )
+
+        # Mock batch context manager
+        mock_batch = MagicMock()
+        mock_batch.__enter__ = MagicMock(return_value=mock_batch)
+        mock_batch.__exit__ = MagicMock(return_value=False)
+        mock_col.batch.dynamic.return_value = mock_batch
+
+        wv.add_texts(documents=[doc], embeddings=[[0.1] * 128])
+
+        # Verify batch.add_object was called with doc_type in properties
+        mock_batch.add_object.assert_called_once()
+        call_kwargs = mock_batch.add_object.call_args
+        stored_props = call_kwargs.kwargs.get("properties")
+        assert stored_props.get("doc_type") == "image", f"doc_type should be stored in properties, got: {stored_props}"
+
+
+class TestVectorDefaultAttributes(unittest.TestCase):
+    """Tests for Vector class default attributes list."""
+
+    @patch("core.rag.datasource.vdb.vector_factory.Vector._get_embeddings")
+    @patch("core.rag.datasource.vdb.vector_factory.Vector._init_vector")
+    def test_default_attributes_include_doc_type(self, mock_init_vector, mock_get_embeddings):
+        """Test that Vector class default attributes include doc_type."""
+        from core.rag.datasource.vdb.vector_factory import Vector
+
+        mock_get_embeddings.return_value = MagicMock()
+        mock_init_vector.return_value = MagicMock()
+
+        mock_dataset = MagicMock()
+        mock_dataset.index_struct_dict = None
+
+        vector = Vector(dataset=mock_dataset)
+
+        assert "doc_type" in vector._attributes, f"doc_type should be in default attributes, got: {vector._attributes}"
+
+
+if __name__ == "__main__":
+    unittest.main()

api/tests/unit_tests/core/rag/extractor/firecrawl/test_firecrawl.py

@@ -104,10 +104,11 @@ class TestFirecrawlApp:
 
     def test_map_known_error(self, mocker: MockerFixture):
         app = FirecrawlApp(api_key="fc-key", base_url="https://custom.firecrawl.dev")
-        mock_handle = mocker.patch.object(app, "_handle_error")
+        mock_handle = mocker.patch.object(app, "_handle_error", side_effect=Exception("map error"))
         mocker.patch("httpx.post", return_value=_response(409, {"error": "conflict"}))
 
-        assert app.map("https://example.com") == {}
+        with pytest.raises(Exception, match="map error"):
+            app.map("https://example.com")
         mock_handle.assert_called_once()
 
     def test_map_unknown_error_raises(self, mocker: MockerFixture):
@@ -177,10 +178,11 @@ class TestFirecrawlApp:
 
     def test_check_crawl_status_non_200_uses_error_handler(self, mocker: MockerFixture):
         app = FirecrawlApp(api_key="fc-key", base_url="https://custom.firecrawl.dev")
-        mock_handle = mocker.patch.object(app, "_handle_error")
+        mock_handle = mocker.patch.object(app, "_handle_error", side_effect=Exception("crawl error"))
         mocker.patch("httpx.get", return_value=_response(500, {"error": "server"}))
 
-        assert app.check_crawl_status("job-1") == {}
+        with pytest.raises(Exception, match="crawl error"):
+            app.check_crawl_status("job-1")
         mock_handle.assert_called_once()
 
     def test_check_crawl_status_save_failure_raises(self, mocker: MockerFixture):
@@ -272,9 +274,10 @@ class TestFirecrawlApp:
 
     def test_search_known_http_error(self, mocker: MockerFixture):
         app = FirecrawlApp(api_key="fc-key", base_url="https://custom.firecrawl.dev")
-        mock_handle = mocker.patch.object(app, "_handle_error")
+        mock_handle = mocker.patch.object(app, "_handle_error", side_effect=Exception("search error"))
         mocker.patch("httpx.post", return_value=_response(408, {"error": "timeout"}))
-        assert app.search("python") == {}
+        with pytest.raises(Exception, match="search error"):
+            app.search("python")
         mock_handle.assert_called_once()
 
     def test_search_unknown_http_error(self, mocker: MockerFixture):

api/tests/unit_tests/core/rag/indexing/processor/test_paragraph_index_processor.py

@@ -236,7 +236,8 @@ class TestParagraphIndexProcessor:
             "core.rag.index_processor.processor.paragraph_index_processor.RetrievalService.retrieve"
         ) as mock_retrieve:
             mock_retrieve.return_value = [accepted, rejected]
-            docs = processor.retrieve("semantic_search", "query", dataset, 5, 0.5, {})
+            reranking_model = {"reranking_provider_name": "", "reranking_model_name": ""}
+            docs = processor.retrieve("semantic_search", "query", dataset, 5, 0.5, reranking_model)
 
         assert len(docs) == 1
         assert docs[0].metadata["score"] == 0.9

api/tests/unit_tests/core/rag/indexing/processor/test_parent_child_index_processor.py

@@ -307,7 +307,8 @@ class TestParentChildIndexProcessor:
             "core.rag.index_processor.processor.parent_child_index_processor.RetrievalService.retrieve"
         ) as mock_retrieve:
             mock_retrieve.return_value = [ok_result, low_result]
-            docs = processor.retrieve("semantic_search", "query", dataset, 3, 0.5, {})
+            reranking_model = {"reranking_provider_name": "", "reranking_model_name": ""}
+            docs = processor.retrieve("semantic_search", "query", dataset, 3, 0.5, reranking_model)
 
         assert len(docs) == 1
         assert docs[0].page_content == "keep"

api/tests/unit_tests/core/rag/indexing/processor/test_qa_index_processor.py

@@ -262,7 +262,8 @@ class TestQAIndexProcessor:
 
         with patch("core.rag.index_processor.processor.qa_index_processor.RetrievalService.retrieve") as mock_retrieve:
             mock_retrieve.return_value = [result_ok, result_low]
-            docs = processor.retrieve("semantic_search", "query", dataset, 5, 0.5, {})
+            reranking_model = {"reranking_provider_name": "", "reranking_model_name": ""}
+            docs = processor.retrieve("semantic_search", "query", dataset, 5, 0.5, reranking_model)
 
         assert len(docs) == 1
         assert docs[0].page_content == "accepted"

api/tests/unit_tests/core/rag/retrieval/test_dataset_retrieval.py

@@ -25,6 +25,7 @@ from core.app.app_config.entities import ModelConfig as WorkflowModelConfig
 from core.app.entities.app_invoke_entities import InvokeFrom, ModelConfigWithCredentialsEntity
 from core.entities.agent_entities import PlanningStrategy
 from core.entities.model_entities import ModelStatus
+from core.rag.data_post_processor.data_post_processor import WeightsDict
 from core.rag.datasource.retrieval_service import RetrievalService
 from core.rag.index_processor.constant.doc_type import DocType
 from core.rag.index_processor.constant.index_type import IndexStructureType
@@ -4686,7 +4687,10 @@ class TestSingleAndMultipleRetrieveCoverage:
             extra={"dataset_name": "Ext", "title": "Ext"},
         )
         app = Flask(__name__)
-        weights = {"vector_setting": {}}
+        weights: WeightsDict = {
+            "vector_setting": {"vector_weight": 0.5, "embedding_provider_name": "", "embedding_model_name": ""},
+            "keyword_setting": {"keyword_weight": 0.5},
+        }
 
         def fake_multiple_thread(**kwargs):
             if kwargs["query"]:

api/tests/unit_tests/core/repositories/test_human_input_repository.py

@@ -0,0 +1,677 @@
+from __future__ import annotations
+
+import dataclasses
+import json
+from collections.abc import Sequence
+from datetime import datetime, timedelta
+from types import SimpleNamespace
+from typing import Any
+from unittest.mock import MagicMock
+
+import pytest
+
+from core.repositories.human_input_repository import (
+    HumanInputFormRecord,
+    HumanInputFormRepositoryImpl,
+    HumanInputFormSubmissionRepository,
+    _HumanInputFormEntityImpl,
+    _HumanInputFormRecipientEntityImpl,
+    _InvalidTimeoutStatusError,
+    _WorkspaceMemberInfo,
+)
+from dify_graph.nodes.human_input.entities import (
+    EmailDeliveryConfig,
+    EmailDeliveryMethod,
+    EmailRecipients,
+    ExternalRecipient,
+    HumanInputNodeData,
+    MemberRecipient,
+    UserAction,
+    WebAppDeliveryMethod,
+)
+from dify_graph.nodes.human_input.enums import HumanInputFormKind, HumanInputFormStatus
+from dify_graph.repositories.human_input_form_repository import FormCreateParams, FormNotFoundError
+from libs.datetime_utils import naive_utc_now
+from models.human_input import HumanInputFormRecipient, RecipientType
+
+
+@pytest.fixture(autouse=True)
+def _stub_select(monkeypatch: pytest.MonkeyPatch) -> None:
+    class _FakeSelect:
+        def join(self, *_args: Any, **_kwargs: Any) -> _FakeSelect:
+            return self
+
+        def where(self, *_args: Any, **_kwargs: Any) -> _FakeSelect:
+            return self
+
+        def options(self, *_args: Any, **_kwargs: Any) -> _FakeSelect:
+            return self
+
+    monkeypatch.setattr("core.repositories.human_input_repository.select", lambda *_args, **_kwargs: _FakeSelect())
+    monkeypatch.setattr("core.repositories.human_input_repository.selectinload", lambda *_args, **_kwargs: "_loader")
+
+
+def _make_form_definition_json(*, include_expiration_time: bool) -> str:
+    payload: dict[str, Any] = {
+        "form_content": "hi",
+        "inputs": [],
+        "user_actions": [{"id": "submit", "title": "Submit"}],
+        "rendered_content": "<p>hi</p>",
+    }
+    if include_expiration_time:
+        payload["expiration_time"] = naive_utc_now()
+    return json.dumps(payload, default=str)
+
+
+@dataclasses.dataclass
+class _DummyForm:
+    id: str
+    workflow_run_id: str | None
+    node_id: str
+    tenant_id: str
+    app_id: str
+    form_definition: str
+    rendered_content: str
+    expiration_time: datetime
+    form_kind: HumanInputFormKind = HumanInputFormKind.RUNTIME
+    created_at: datetime = dataclasses.field(default_factory=naive_utc_now)
+    selected_action_id: str | None = None
+    submitted_data: str | None = None
+    submitted_at: datetime | None = None
+    submission_user_id: str | None = None
+    submission_end_user_id: str | None = None
+    completed_by_recipient_id: str | None = None
+    status: HumanInputFormStatus = HumanInputFormStatus.WAITING
+
+
+@dataclasses.dataclass
+class _DummyRecipient:
+    id: str
+    form_id: str
+    recipient_type: RecipientType
+    access_token: str | None
+
+
+class _FakeScalarResult:
+    def __init__(self, obj: Any):
+        self._obj = obj
+
+    def first(self) -> Any:
+        if isinstance(self._obj, list):
+            return self._obj[0] if self._obj else None
+        return self._obj
+
+    def all(self) -> list[Any]:
+        if self._obj is None:
+            return []
+        if isinstance(self._obj, list):
+            return list(self._obj)
+        return [self._obj]
+
+
+class _FakeExecuteResult:
+    def __init__(self, rows: Sequence[tuple[Any, ...]]):
+        self._rows = list(rows)
+
+    def all(self) -> list[tuple[Any, ...]]:
+        return list(self._rows)
+
+
+class _FakeSession:
+    def __init__(
+        self,
+        *,
+        scalars_result: Any = None,
+        scalars_results: list[Any] | None = None,
+        forms: dict[str, _DummyForm] | None = None,
+        recipients: dict[str, _DummyRecipient] | None = None,
+        execute_rows: Sequence[tuple[Any, ...]] = (),
+    ):
+        if scalars_results is not None:
+            self._scalars_queue = list(scalars_results)
+        else:
+            self._scalars_queue = [scalars_result]
+        self._forms = forms or {}
+        self._recipients = recipients or {}
+        self._execute_rows = list(execute_rows)
+        self.added: list[Any] = []
+
+    def scalars(self, _query: Any) -> _FakeScalarResult:
+        if self._scalars_queue:
+            value = self._scalars_queue.pop(0)
+        else:
+            value = None
+        return _FakeScalarResult(value)
+
+    def execute(self, _stmt: Any) -> _FakeExecuteResult:
+        return _FakeExecuteResult(self._execute_rows)
+
+    def get(self, model_cls: Any, obj_id: str) -> Any:
+        name = getattr(model_cls, "__name__", "")
+        if name == "HumanInputForm":
+            return self._forms.get(obj_id)
+        if name == "HumanInputFormRecipient":
+            return self._recipients.get(obj_id)
+        return None
+
+    def add(self, obj: Any) -> None:
+        self.added.append(obj)
+
+    def add_all(self, objs: Sequence[Any]) -> None:
+        self.added.extend(list(objs))
+
+    def flush(self) -> None:
+        # Simulate DB default population for attributes referenced in entity wrappers.
+        for obj in self.added:
+            if hasattr(obj, "id") and obj.id in (None, ""):
+                obj.id = f"gen-{len(str(self.added))}"
+            if isinstance(obj, HumanInputFormRecipient) and obj.access_token is None:
+                if obj.recipient_type == RecipientType.CONSOLE:
+                    obj.access_token = "token-console"
+                elif obj.recipient_type == RecipientType.BACKSTAGE:
+                    obj.access_token = "token-backstage"
+                else:
+                    obj.access_token = "token-webapp"
+
+    def refresh(self, _obj: Any) -> None:
+        return None
+
+    def begin(self) -> _FakeSession:
+        return self
+
+    def __enter__(self) -> _FakeSession:
+        return self
+
+    def __exit__(self, exc_type, exc, tb) -> None:
+        return None
+
+
+class _SessionFactoryStub:
+    def __init__(self, session: _FakeSession):
+        self._session = session
+
+    def create_session(self) -> _FakeSession:
+        return self._session
+
+
+def _patch_session_factory(monkeypatch: pytest.MonkeyPatch, session: _FakeSession) -> None:
+    monkeypatch.setattr("core.repositories.human_input_repository.session_factory", _SessionFactoryStub(session))
+
+
+def test_recipient_entity_token_raises_when_missing() -> None:
+    recipient = SimpleNamespace(id="r1", access_token=None)
+    entity = _HumanInputFormRecipientEntityImpl(recipient)  # type: ignore[arg-type]
+    with pytest.raises(AssertionError, match="access_token should not be None"):
+        _ = entity.token
+
+
+def test_recipient_entity_id_and_token_success() -> None:
+    recipient = SimpleNamespace(id="r1", access_token="tok")
+    entity = _HumanInputFormRecipientEntityImpl(recipient)  # type: ignore[arg-type]
+    assert entity.id == "r1"
+    assert entity.token == "tok"
+
+
+def test_form_entity_web_app_token_prefers_console_then_webapp_then_none() -> None:
+    form = _DummyForm(
+        id="f1",
+        workflow_run_id="run",
+        node_id="node",
+        tenant_id="tenant",
+        app_id="app",
+        form_definition=_make_form_definition_json(include_expiration_time=True),
+        rendered_content="<p>x</p>",
+        expiration_time=naive_utc_now(),
+    )
+    console = _DummyRecipient(id="c1", form_id=form.id, recipient_type=RecipientType.CONSOLE, access_token="ctok")
+    webapp = _DummyRecipient(
+        id="w1", form_id=form.id, recipient_type=RecipientType.STANDALONE_WEB_APP, access_token="wtok"
+    )
+
+    entity = _HumanInputFormEntityImpl(form_model=form, recipient_models=[webapp, console])  # type: ignore[arg-type]
+    assert entity.web_app_token == "ctok"
+
+    entity = _HumanInputFormEntityImpl(form_model=form, recipient_models=[webapp])  # type: ignore[arg-type]
+    assert entity.web_app_token == "wtok"
+
+    entity = _HumanInputFormEntityImpl(form_model=form, recipient_models=[])  # type: ignore[arg-type]
+    assert entity.web_app_token is None
+
+
+def test_form_entity_submitted_data_parsed() -> None:
+    form = _DummyForm(
+        id="f1",
+        workflow_run_id="run",
+        node_id="node",
+        tenant_id="tenant",
+        app_id="app",
+        form_definition=_make_form_definition_json(include_expiration_time=True),
+        rendered_content="<p>x</p>",
+        expiration_time=naive_utc_now(),
+        submitted_data='{"a": 1}',
+        submitted_at=naive_utc_now(),
+    )
+    entity = _HumanInputFormEntityImpl(form_model=form, recipient_models=[])  # type: ignore[arg-type]
+    assert entity.submitted is True
+    assert entity.submitted_data == {"a": 1}
+    assert entity.rendered_content == "<p>x</p>"
+    assert entity.selected_action_id is None
+    assert entity.status == HumanInputFormStatus.WAITING
+
+
+def test_form_record_from_models_injects_expiration_time_when_missing() -> None:
+    expiration = naive_utc_now()
+    form = _DummyForm(
+        id="f1",
+        workflow_run_id=None,
+        node_id="node",
+        tenant_id="tenant",
+        app_id="app",
+        form_definition=_make_form_definition_json(include_expiration_time=False),
+        rendered_content="<p>x</p>",
+        expiration_time=expiration,
+        submitted_data='{"k": "v"}',
+    )
+    record = HumanInputFormRecord.from_models(form, None)  # type: ignore[arg-type]
+    assert record.definition.expiration_time == expiration
+    assert record.submitted_data == {"k": "v"}
+    assert record.submitted is False
+
+
+def test_create_email_recipients_from_resolved_dedupes_and_skips_blank(monkeypatch: pytest.MonkeyPatch) -> None:
+    created: list[SimpleNamespace] = []
+
+    def fake_new(cls, form_id: str, delivery_id: str, payload: Any):  # type: ignore[no-untyped-def]
+        recipient = SimpleNamespace(
+            id=f"{payload.TYPE}-{len(created)}",
+            form_id=form_id,
+            delivery_id=delivery_id,
+            recipient_type=payload.TYPE,
+            recipient_payload=payload.model_dump_json(),
+            access_token="tok",
+        )
+        created.append(recipient)
+        return recipient
+
+    monkeypatch.setattr("core.repositories.human_input_repository.HumanInputFormRecipient.new", classmethod(fake_new))
+
+    repo = HumanInputFormRepositoryImpl(tenant_id="tenant")
+    recipients = repo._create_email_recipients_from_resolved(  # type: ignore[attr-defined]
+        form_id="f",
+        delivery_id="d",
+        members=[
+            _WorkspaceMemberInfo(user_id="u1", email=""),
+            _WorkspaceMemberInfo(user_id="u2", email="a@example.com"),
+            _WorkspaceMemberInfo(user_id="u3", email="a@example.com"),
+        ],
+        external_emails=["", "a@example.com", "b@example.com", "b@example.com"],
+    )
+    assert [r.recipient_type for r in recipients] == [RecipientType.EMAIL_MEMBER, RecipientType.EMAIL_EXTERNAL]
+
+
+def test_query_workspace_members_by_ids_empty_returns_empty() -> None:
+    repo = HumanInputFormRepositoryImpl(tenant_id="tenant")
+    assert repo._query_workspace_members_by_ids(session=MagicMock(), restrict_to_user_ids=["", ""]) == []
+
+
+def test_query_workspace_members_by_ids_maps_rows() -> None:
+    session = _FakeSession(execute_rows=[("u1", "a@example.com"), ("u2", "b@example.com")])
+    repo = HumanInputFormRepositoryImpl(tenant_id="tenant")
+    rows = repo._query_workspace_members_by_ids(session=session, restrict_to_user_ids=["u1", "u2"])
+    assert rows == [
+        _WorkspaceMemberInfo(user_id="u1", email="a@example.com"),
+        _WorkspaceMemberInfo(user_id="u2", email="b@example.com"),
+    ]
+
+
+def test_query_all_workspace_members_maps_rows() -> None:
+    session = _FakeSession(execute_rows=[("u1", "a@example.com")])
+    repo = HumanInputFormRepositoryImpl(tenant_id="tenant")
+    rows = repo._query_all_workspace_members(session=session)
+    assert rows == [_WorkspaceMemberInfo(user_id="u1", email="a@example.com")]
+
+
+def test_repository_init_sets_tenant_id() -> None:
+    repo = HumanInputFormRepositoryImpl(tenant_id="tenant")
+    assert repo._tenant_id == "tenant"
+
+
+def test_delivery_method_to_model_webapp_creates_delivery_and_recipient(monkeypatch: pytest.MonkeyPatch) -> None:
+    repo = HumanInputFormRepositoryImpl(tenant_id="tenant")
+    monkeypatch.setattr("core.repositories.human_input_repository.uuidv7", lambda: "del-1")
+    result = repo._delivery_method_to_model(
+        session=MagicMock(), form_id="form-1", delivery_method=WebAppDeliveryMethod()
+    )
+    assert result.delivery.id == "del-1"
+    assert result.delivery.form_id == "form-1"
+    assert len(result.recipients) == 1
+    assert result.recipients[0].recipient_type == RecipientType.STANDALONE_WEB_APP
+
+
+def test_delivery_method_to_model_email_uses_build_email_recipients(monkeypatch: pytest.MonkeyPatch) -> None:
+    repo = HumanInputFormRepositoryImpl(tenant_id="tenant")
+    monkeypatch.setattr("core.repositories.human_input_repository.uuidv7", lambda: "del-1")
+    called: dict[str, Any] = {}
+
+    def fake_build(*, session: Any, form_id: str, delivery_id: str, recipients_config: Any) -> list[Any]:
+        called.update(
+            {"session": session, "form_id": form_id, "delivery_id": delivery_id, "recipients_config": recipients_config}
+        )
+        return ["r"]
+
+    monkeypatch.setattr(repo, "_build_email_recipients", fake_build)
+
+    method = EmailDeliveryMethod(
+        config=EmailDeliveryConfig(
+            recipients=EmailRecipients(
+                whole_workspace=False,
+                items=[MemberRecipient(user_id="u1"), ExternalRecipient(email="e@example.com")],
+            ),
+            subject="s",
+            body="b",
+        )
+    )
+    result = repo._delivery_method_to_model(session="sess", form_id="form-1", delivery_method=method)
+    assert result.recipients == ["r"]
+    assert called["delivery_id"] == "del-1"
+
+
+def test_build_email_recipients_uses_all_members_when_whole_workspace(monkeypatch: pytest.MonkeyPatch) -> None:
+    repo = HumanInputFormRepositoryImpl(tenant_id="tenant")
+    monkeypatch.setattr(
+        repo,
+        "_query_all_workspace_members",
+        lambda *, session: [_WorkspaceMemberInfo(user_id="u", email="a@example.com")],
+    )
+    monkeypatch.setattr(repo, "_create_email_recipients_from_resolved", lambda **_: ["ok"])
+    recipients = repo._build_email_recipients(
+        session=MagicMock(),
+        form_id="f",
+        delivery_id="d",
+        recipients_config=EmailRecipients(whole_workspace=True, items=[ExternalRecipient(email="e@example.com")]),
+    )
+    assert recipients == ["ok"]
+
+
+def test_build_email_recipients_uses_selected_members_when_not_whole_workspace(monkeypatch: pytest.MonkeyPatch) -> None:
+    repo = HumanInputFormRepositoryImpl(tenant_id="tenant")
+
+    def fake_query(*, session: Any, restrict_to_user_ids: Sequence[str]) -> list[_WorkspaceMemberInfo]:
+        assert restrict_to_user_ids == ["u1"]
+        return [_WorkspaceMemberInfo(user_id="u1", email="a@example.com")]
+
+    monkeypatch.setattr(repo, "_query_workspace_members_by_ids", fake_query)
+    monkeypatch.setattr(repo, "_create_email_recipients_from_resolved", lambda **_: ["ok"])
+    recipients = repo._build_email_recipients(
+        session=MagicMock(),
+        form_id="f",
+        delivery_id="d",
+        recipients_config=EmailRecipients(
+            whole_workspace=False,
+            items=[MemberRecipient(user_id="u1"), ExternalRecipient(email="e@example.com")],
+        ),
+    )
+    assert recipients == ["ok"]
+
+
+def test_get_form_returns_entity_and_none_when_missing(monkeypatch: pytest.MonkeyPatch) -> None:
+    _patch_session_factory(monkeypatch, _FakeSession(scalars_results=[None]))
+    repo = HumanInputFormRepositoryImpl(tenant_id="tenant")
+    assert repo.get_form("run", "node") is None
+
+    form = _DummyForm(
+        id="f1",
+        workflow_run_id="run",
+        node_id="node",
+        tenant_id="tenant",
+        app_id="app",
+        form_definition=_make_form_definition_json(include_expiration_time=True),
+        rendered_content="<p>x</p>",
+        expiration_time=naive_utc_now(),
+    )
+    recipient = _DummyRecipient(
+        id="r1",
+        form_id=form.id,
+        recipient_type=RecipientType.STANDALONE_WEB_APP,
+        access_token="tok",
+    )
+    session = _FakeSession(scalars_results=[form, [recipient]])
+    _patch_session_factory(monkeypatch, session)
+    repo = HumanInputFormRepositoryImpl(tenant_id="tenant")
+    entity = repo.get_form("run", "node")
+    assert entity is not None
+    assert entity.id == "f1"
+    assert entity.recipients[0].id == "r1"
+    assert entity.recipients[0].token == "tok"
+
+
+def test_create_form_adds_console_and_backstage_recipients(monkeypatch: pytest.MonkeyPatch) -> None:
+    fixed_now = datetime(2024, 1, 1, 0, 0, 0)
+    monkeypatch.setattr("core.repositories.human_input_repository.naive_utc_now", lambda: fixed_now)
+
+    ids = iter(["form-id", "del-web", "del-console", "del-backstage"])
+    monkeypatch.setattr("core.repositories.human_input_repository.uuidv7", lambda: next(ids))
+
+    session = _FakeSession()
+    _patch_session_factory(monkeypatch, session)
+    repo = HumanInputFormRepositoryImpl(tenant_id="tenant")
+
+    form_config = HumanInputNodeData(
+        title="Title",
+        delivery_methods=[],
+        form_content="hello",
+        inputs=[],
+        user_actions=[UserAction(id="submit", title="Submit")],
+    )
+    params = FormCreateParams(
+        app_id="app",
+        workflow_execution_id="run",
+        node_id="node",
+        form_config=form_config,
+        rendered_content="<p>hello</p>",
+        delivery_methods=[WebAppDeliveryMethod()],
+        display_in_ui=True,
+        resolved_default_values={},
+        form_kind=HumanInputFormKind.RUNTIME,
+        console_recipient_required=True,
+        console_creator_account_id="acc-1",
+        backstage_recipient_required=True,
+    )
+
+    entity = repo.create_form(params)
+    assert entity.id == "form-id"
+    assert entity.expiration_time == fixed_now + timedelta(hours=form_config.timeout)
+    # Console token should take precedence when console recipient is present.
+    assert entity.web_app_token == "token-console"
+    assert len(entity.recipients) == 3
+
+
+def test_submission_get_by_token_returns_none_when_missing_or_form_missing(monkeypatch: pytest.MonkeyPatch) -> None:
+    _patch_session_factory(monkeypatch, _FakeSession(scalars_result=None))
+    repo = HumanInputFormSubmissionRepository()
+    assert repo.get_by_token("tok") is None
+
+    recipient = SimpleNamespace(form=None)
+    _patch_session_factory(monkeypatch, _FakeSession(scalars_result=recipient))
+    repo = HumanInputFormSubmissionRepository()
+    assert repo.get_by_token("tok") is None
+
+
+def test_submission_repository_init_no_args() -> None:
+    repo = HumanInputFormSubmissionRepository()
+    assert isinstance(repo, HumanInputFormSubmissionRepository)
+
+
+def test_submission_get_by_token_and_get_by_form_id_success_paths(monkeypatch: pytest.MonkeyPatch) -> None:
+    form = _DummyForm(
+        id="f1",
+        workflow_run_id=None,
+        node_id="node",
+        tenant_id="tenant",
+        app_id="app",
+        form_definition=_make_form_definition_json(include_expiration_time=True),
+        rendered_content="<p>x</p>",
+        expiration_time=naive_utc_now(),
+    )
+    recipient = SimpleNamespace(
+        id="r1",
+        form_id=form.id,
+        recipient_type=RecipientType.STANDALONE_WEB_APP,
+        access_token="tok",
+        form=form,
+    )
+
+    _patch_session_factory(monkeypatch, _FakeSession(scalars_result=recipient))
+    repo = HumanInputFormSubmissionRepository()
+    record = repo.get_by_token("tok")
+    assert record is not None
+    assert record.access_token == "tok"
+
+    _patch_session_factory(monkeypatch, _FakeSession(scalars_result=recipient))
+    repo = HumanInputFormSubmissionRepository()
+    record = repo.get_by_form_id_and_recipient_type(form_id=form.id, recipient_type=RecipientType.STANDALONE_WEB_APP)
+    assert record is not None
+    assert record.recipient_id == "r1"
+
+
+def test_submission_get_by_form_id_returns_none_on_missing(monkeypatch: pytest.MonkeyPatch) -> None:
+    _patch_session_factory(monkeypatch, _FakeSession(scalars_result=None))
+    repo = HumanInputFormSubmissionRepository()
+    assert repo.get_by_form_id_and_recipient_type(form_id="f", recipient_type=RecipientType.CONSOLE) is None
+
+
+def test_mark_submitted_updates_and_raises_when_missing(monkeypatch: pytest.MonkeyPatch) -> None:
+    fixed_now = datetime(2024, 1, 1, 0, 0, 0)
+    monkeypatch.setattr("core.repositories.human_input_repository.naive_utc_now", lambda: fixed_now)
+
+    missing_session = _FakeSession(forms={})
+    _patch_session_factory(monkeypatch, missing_session)
+    repo = HumanInputFormSubmissionRepository()
+    with pytest.raises(FormNotFoundError, match="form not found"):
+        repo.mark_submitted(
+            form_id="missing",
+            recipient_id=None,
+            selected_action_id="a",
+            form_data={},
+            submission_user_id=None,
+            submission_end_user_id=None,
+        )
+
+    form = _DummyForm(
+        id="f",
+        workflow_run_id=None,
+        node_id="node",
+        tenant_id="tenant",
+        app_id="app",
+        form_definition=_make_form_definition_json(include_expiration_time=True),
+        rendered_content="<p>x</p>",
+        expiration_time=fixed_now,
+    )
+    recipient = _DummyRecipient(id="r", form_id=form.id, recipient_type=RecipientType.CONSOLE, access_token="tok")
+    session = _FakeSession(forms={form.id: form}, recipients={recipient.id: recipient})
+    _patch_session_factory(monkeypatch, session)
+    repo = HumanInputFormSubmissionRepository()
+    record = repo.mark_submitted(
+        form_id=form.id,
+        recipient_id=recipient.id,
+        selected_action_id="approve",
+        form_data={"k": "v"},
+        submission_user_id="u",
+        submission_end_user_id="eu",
+    )
+    assert form.status == HumanInputFormStatus.SUBMITTED
+    assert form.submitted_at == fixed_now
+    assert record.submitted_data == {"k": "v"}
+
+
+def test_mark_timeout_invalid_status_raises(monkeypatch: pytest.MonkeyPatch) -> None:
+    form = _DummyForm(
+        id="f",
+        workflow_run_id=None,
+        node_id="node",
+        tenant_id="tenant",
+        app_id="app",
+        form_definition=_make_form_definition_json(include_expiration_time=True),
+        rendered_content="<p>x</p>",
+        expiration_time=naive_utc_now(),
+    )
+    session = _FakeSession(forms={form.id: form})
+    _patch_session_factory(monkeypatch, session)
+    repo = HumanInputFormSubmissionRepository()
+    with pytest.raises(_InvalidTimeoutStatusError, match="invalid timeout status"):
+        repo.mark_timeout(form_id=form.id, timeout_status=HumanInputFormStatus.SUBMITTED)  # type: ignore[arg-type]
+
+
+def test_mark_timeout_already_timed_out_returns_record(monkeypatch: pytest.MonkeyPatch) -> None:
+    form = _DummyForm(
+        id="f",
+        workflow_run_id=None,
+        node_id="node",
+        tenant_id="tenant",
+        app_id="app",
+        form_definition=_make_form_definition_json(include_expiration_time=True),
+        rendered_content="<p>x</p>",
+        expiration_time=naive_utc_now(),
+        status=HumanInputFormStatus.TIMEOUT,
+    )
+    session = _FakeSession(forms={form.id: form})
+    _patch_session_factory(monkeypatch, session)
+    repo = HumanInputFormSubmissionRepository()
+    record = repo.mark_timeout(form_id=form.id, timeout_status=HumanInputFormStatus.TIMEOUT, reason="r")
+    assert record.status == HumanInputFormStatus.TIMEOUT
+
+
+def test_mark_timeout_submitted_raises_form_not_found(monkeypatch: pytest.MonkeyPatch) -> None:
+    form = _DummyForm(
+        id="f",
+        workflow_run_id=None,
+        node_id="node",
+        tenant_id="tenant",
+        app_id="app",
+        form_definition=_make_form_definition_json(include_expiration_time=True),
+        rendered_content="<p>x</p>",
+        expiration_time=naive_utc_now(),
+        status=HumanInputFormStatus.SUBMITTED,
+    )
+    session = _FakeSession(forms={form.id: form})
+    _patch_session_factory(monkeypatch, session)
+    repo = HumanInputFormSubmissionRepository()
+    with pytest.raises(FormNotFoundError, match="form already submitted"):
+        repo.mark_timeout(form_id=form.id, timeout_status=HumanInputFormStatus.EXPIRED)
+
+
+def test_mark_timeout_updates_fields(monkeypatch: pytest.MonkeyPatch) -> None:
+    form = _DummyForm(
+        id="f",
+        workflow_run_id=None,
+        node_id="node",
+        tenant_id="tenant",
+        app_id="app",
+        form_definition=_make_form_definition_json(include_expiration_time=True),
+        rendered_content="<p>x</p>",
+        expiration_time=naive_utc_now(),
+        selected_action_id="a",
+        submitted_data="{}",
+        submission_user_id="u",
+        submission_end_user_id="eu",
+        completed_by_recipient_id="r",
+        status=HumanInputFormStatus.WAITING,
+    )
+    session = _FakeSession(forms={form.id: form})
+    _patch_session_factory(monkeypatch, session)
+    repo = HumanInputFormSubmissionRepository()
+    record = repo.mark_timeout(form_id=form.id, timeout_status=HumanInputFormStatus.EXPIRED)
+    assert form.status == HumanInputFormStatus.EXPIRED
+    assert form.selected_action_id is None
+    assert form.submitted_data is None
+    assert form.submission_user_id is None
+    assert form.submission_end_user_id is None
+    assert form.completed_by_recipient_id is None
+    assert record.status == HumanInputFormStatus.EXPIRED
+
+
+def test_mark_timeout_raises_when_form_missing(monkeypatch: pytest.MonkeyPatch) -> None:
+    _patch_session_factory(monkeypatch, _FakeSession(forms={}))
+    repo = HumanInputFormSubmissionRepository()
+    with pytest.raises(FormNotFoundError, match="form not found"):
+        repo.mark_timeout(form_id="missing", timeout_status=HumanInputFormStatus.TIMEOUT)

api/tests/unit_tests/core/repositories/test_sqlalchemy_workflow_execution_repository.py

@@ -1,84 +1,291 @@
-from datetime import datetime
+from datetime import UTC, datetime
 from unittest.mock import MagicMock
 from uuid import uuid4
 
-from sqlalchemy import create_engine
+import pytest
+from sqlalchemy.engine import Engine
 from sqlalchemy.orm import sessionmaker
 
 from core.repositories.sqlalchemy_workflow_execution_repository import SQLAlchemyWorkflowExecutionRepository
-from dify_graph.entities.workflow_execution import WorkflowExecution, WorkflowType
-from models import Account, WorkflowRun
+from dify_graph.entities.workflow_execution import WorkflowExecution, WorkflowExecutionStatus, WorkflowType
+from models import Account, CreatorUserRole, EndUser, WorkflowRun
 from models.enums import WorkflowRunTriggeredFrom
 
 
-def _build_repository_with_mocked_session(session: MagicMock) -> SQLAlchemyWorkflowExecutionRepository:
-    engine = create_engine("sqlite:///:memory:")
-    real_session_factory = sessionmaker(bind=engine, expire_on_commit=False)
-
-    user = MagicMock(spec=Account)
-    user.id = str(uuid4())
-    user.current_tenant_id = str(uuid4())
-
-    repository = SQLAlchemyWorkflowExecutionRepository(
-        session_factory=real_session_factory,
-        user=user,
-        app_id="app-id",
-        triggered_from=WorkflowRunTriggeredFrom.APP_RUN,
-    )
-
-    session_context = MagicMock()
-    session_context.__enter__.return_value = session
-    session_context.__exit__.return_value = False
-    repository._session_factory = MagicMock(return_value=session_context)
-    return repository
-
-
-def _build_execution(*, execution_id: str, started_at: datetime) -> WorkflowExecution:
-    return WorkflowExecution.new(
-        id_=execution_id,
-        workflow_id="workflow-id",
-        workflow_type=WorkflowType.WORKFLOW,
-        workflow_version="1.0.0",
-        graph={"nodes": [], "edges": []},
-        inputs={"query": "hello"},
-        started_at=started_at,
-    )
-
-
-def test_save_uses_execution_started_at_when_record_does_not_exist():
+@pytest.fixture
+def mock_session_factory():
+    """Mock SQLAlchemy session factory."""
+    session_factory = MagicMock(spec=sessionmaker)
     session = MagicMock()
     session.get.return_value = None
-    repository = _build_repository_with_mocked_session(session)
-
-    started_at = datetime(2026, 1, 1, 12, 0, 0)
-    execution = _build_execution(execution_id=str(uuid4()), started_at=started_at)
-
-    repository.save(execution)
-
-    saved_model = session.merge.call_args.args[0]
-    assert saved_model.created_at == started_at
-    session.commit.assert_called_once()
+    session_factory.return_value.__enter__.return_value = session
+    return session_factory
 
 
-def test_save_preserves_existing_created_at_when_record_already_exists():
-    session = MagicMock()
-    repository = _build_repository_with_mocked_session(session)
+@pytest.fixture
+def mock_engine():
+    """Mock SQLAlchemy Engine."""
+    return MagicMock(spec=Engine)
 
-    execution_id = str(uuid4())
-    existing_created_at = datetime(2026, 1, 1, 12, 0, 0)
-    existing_run = WorkflowRun()
-    existing_run.id = execution_id
-    existing_run.tenant_id = repository._tenant_id
-    existing_run.created_at = existing_created_at
-    session.get.return_value = existing_run
 
-    execution = _build_execution(
-        execution_id=execution_id,
-        started_at=datetime(2026, 1, 1, 12, 30, 0),
+@pytest.fixture
+def mock_account():
+    """Mock Account user."""
+    account = MagicMock(spec=Account)
+    account.id = str(uuid4())
+    account.current_tenant_id = str(uuid4())
+    return account
+
+
+@pytest.fixture
+def mock_end_user():
+    """Mock EndUser."""
+    user = MagicMock(spec=EndUser)
+    user.id = str(uuid4())
+    user.tenant_id = str(uuid4())
+    return user
+
+
+@pytest.fixture
+def sample_workflow_execution():
+    """Sample WorkflowExecution for testing."""
+    return WorkflowExecution(
+        id_=str(uuid4()),
+        workflow_id=str(uuid4()),
+        workflow_type=WorkflowType.WORKFLOW,
+        workflow_version="1.0",
+        graph={"nodes": [], "edges": []},
+        inputs={"input1": "value1"},
+        outputs={"output1": "result1"},
+        status=WorkflowExecutionStatus.SUCCEEDED,
+        error_message="",
+        total_tokens=100,
+        total_steps=5,
+        exceptions_count=0,
+        started_at=datetime.now(UTC),
+        finished_at=datetime.now(UTC),
     )
 
-    repository.save(execution)
 
-    saved_model = session.merge.call_args.args[0]
-    assert saved_model.created_at == existing_created_at
-    session.commit.assert_called_once()
+class TestSQLAlchemyWorkflowExecutionRepository:
+    def test_init_with_sessionmaker(self, mock_session_factory, mock_account):
+        app_id = "test_app_id"
+        triggered_from = WorkflowRunTriggeredFrom.APP_RUN
+
+        repo = SQLAlchemyWorkflowExecutionRepository(
+            session_factory=mock_session_factory, user=mock_account, app_id=app_id, triggered_from=triggered_from
+        )
+
+        assert repo._session_factory == mock_session_factory
+        assert repo._tenant_id == mock_account.current_tenant_id
+        assert repo._app_id == app_id
+        assert repo._triggered_from == triggered_from
+        assert repo._creator_user_id == mock_account.id
+        assert repo._creator_user_role == CreatorUserRole.ACCOUNT
+
+    def test_init_with_engine(self, mock_engine, mock_account):
+        repo = SQLAlchemyWorkflowExecutionRepository(
+            session_factory=mock_engine,
+            user=mock_account,
+            app_id="test_app_id",
+            triggered_from=WorkflowRunTriggeredFrom.APP_RUN,
+        )
+
+        assert isinstance(repo._session_factory, sessionmaker)
+        assert repo._session_factory.kw["bind"] == mock_engine
+
+    def test_init_invalid_session_factory(self, mock_account):
+        with pytest.raises(ValueError, match="Invalid session_factory type"):
+            SQLAlchemyWorkflowExecutionRepository(
+                session_factory="invalid", user=mock_account, app_id=None, triggered_from=None
+            )
+
+    def test_init_no_tenant_id(self, mock_session_factory):
+        user = MagicMock(spec=Account)
+        user.current_tenant_id = None
+
+        with pytest.raises(ValueError, match="User must have a tenant_id"):
+            SQLAlchemyWorkflowExecutionRepository(
+                session_factory=mock_session_factory, user=user, app_id=None, triggered_from=None
+            )
+
+    def test_init_with_end_user(self, mock_session_factory, mock_end_user):
+        repo = SQLAlchemyWorkflowExecutionRepository(
+            session_factory=mock_session_factory, user=mock_end_user, app_id=None, triggered_from=None
+        )
+        assert repo._tenant_id == mock_end_user.tenant_id
+        assert repo._creator_user_role == CreatorUserRole.END_USER
+
+    def test_to_domain_model(self, mock_session_factory, mock_account):
+        repo = SQLAlchemyWorkflowExecutionRepository(
+            session_factory=mock_session_factory, user=mock_account, app_id=None, triggered_from=None
+        )
+
+        db_model = MagicMock(spec=WorkflowRun)
+        db_model.id = str(uuid4())
+        db_model.workflow_id = str(uuid4())
+        db_model.type = "workflow"
+        db_model.version = "1.0"
+        db_model.inputs_dict = {"in": "val"}
+        db_model.outputs_dict = {"out": "val"}
+        db_model.graph_dict = {"nodes": []}
+        db_model.status = "succeeded"
+        db_model.error = "some error"
+        db_model.total_tokens = 50
+        db_model.total_steps = 3
+        db_model.exceptions_count = 1
+        db_model.created_at = datetime.now(UTC)
+        db_model.finished_at = datetime.now(UTC)
+
+        domain_model = repo._to_domain_model(db_model)
+
+        assert domain_model.id_ == db_model.id
+        assert domain_model.workflow_id == db_model.workflow_id
+        assert domain_model.status == WorkflowExecutionStatus.SUCCEEDED
+        assert domain_model.inputs == db_model.inputs_dict
+        assert domain_model.error_message == "some error"
+
+    def test_to_db_model(self, mock_session_factory, mock_account, sample_workflow_execution):
+        repo = SQLAlchemyWorkflowExecutionRepository(
+            session_factory=mock_session_factory,
+            user=mock_account,
+            app_id="test_app",
+            triggered_from=WorkflowRunTriggeredFrom.DEBUGGING,
+        )
+
+        # Make elapsed time deterministic to avoid flaky tests
+        sample_workflow_execution.started_at = datetime(2023, 1, 1, 0, 0, 0, tzinfo=UTC)
+        sample_workflow_execution.finished_at = datetime(2023, 1, 1, 0, 0, 10, tzinfo=UTC)
+
+        db_model = repo._to_db_model(sample_workflow_execution)
+
+        assert db_model.id == sample_workflow_execution.id_
+        assert db_model.tenant_id == repo._tenant_id
+        assert db_model.app_id == "test_app"
+        assert db_model.triggered_from == WorkflowRunTriggeredFrom.DEBUGGING
+        assert db_model.status == sample_workflow_execution.status.value
+        assert db_model.total_tokens == sample_workflow_execution.total_tokens
+        assert db_model.elapsed_time == 10.0
+
+    def test_to_db_model_edge_cases(self, mock_session_factory, mock_account, sample_workflow_execution):
+        repo = SQLAlchemyWorkflowExecutionRepository(
+            session_factory=mock_session_factory,
+            user=mock_account,
+            app_id="test_app",
+            triggered_from=WorkflowRunTriggeredFrom.DEBUGGING,
+        )
+        # Test with empty/None fields
+        sample_workflow_execution.graph = None
+        sample_workflow_execution.inputs = None
+        sample_workflow_execution.outputs = None
+        sample_workflow_execution.error_message = None
+        sample_workflow_execution.finished_at = None
+
+        db_model = repo._to_db_model(sample_workflow_execution)
+
+        assert db_model.graph is None
+        assert db_model.inputs is None
+        assert db_model.outputs is None
+        assert db_model.error is None
+        assert db_model.elapsed_time == 0
+
+    def test_to_db_model_app_id_none(self, mock_session_factory, mock_account, sample_workflow_execution):
+        repo = SQLAlchemyWorkflowExecutionRepository(
+            session_factory=mock_session_factory,
+            user=mock_account,
+            app_id=None,
+            triggered_from=WorkflowRunTriggeredFrom.APP_RUN,
+        )
+
+        db_model = repo._to_db_model(sample_workflow_execution)
+        assert not hasattr(db_model, "app_id") or db_model.app_id is None
+        assert db_model.tenant_id == repo._tenant_id
+
+    def test_to_db_model_missing_context(self, mock_session_factory, mock_account, sample_workflow_execution):
+        repo = SQLAlchemyWorkflowExecutionRepository(
+            session_factory=mock_session_factory, user=mock_account, app_id=None, triggered_from=None
+        )
+
+        # Test triggered_from missing
+        with pytest.raises(ValueError, match="triggered_from is required"):
+            repo._to_db_model(sample_workflow_execution)
+
+        repo._triggered_from = WorkflowRunTriggeredFrom.APP_RUN
+        repo._creator_user_id = None
+        with pytest.raises(ValueError, match="created_by is required"):
+            repo._to_db_model(sample_workflow_execution)
+
+        repo._creator_user_id = "some_id"
+        repo._creator_user_role = None
+        with pytest.raises(ValueError, match="created_by_role is required"):
+            repo._to_db_model(sample_workflow_execution)
+
+    def test_save(self, mock_session_factory, mock_account, sample_workflow_execution):
+        repo = SQLAlchemyWorkflowExecutionRepository(
+            session_factory=mock_session_factory,
+            user=mock_account,
+            app_id="test_app",
+            triggered_from=WorkflowRunTriggeredFrom.APP_RUN,
+        )
+
+        repo.save(sample_workflow_execution)
+
+        session = mock_session_factory.return_value.__enter__.return_value
+        session.merge.assert_called_once()
+        session.commit.assert_called_once()
+
+        # Check cache
+        assert sample_workflow_execution.id_ in repo._execution_cache
+        cached_model = repo._execution_cache[sample_workflow_execution.id_]
+        assert cached_model.id == sample_workflow_execution.id_
+
+    def test_save_uses_execution_started_at_when_record_does_not_exist(
+        self, mock_session_factory, mock_account, sample_workflow_execution
+    ):
+        repo = SQLAlchemyWorkflowExecutionRepository(
+            session_factory=mock_session_factory,
+            user=mock_account,
+            app_id="test_app",
+            triggered_from=WorkflowRunTriggeredFrom.APP_RUN,
+        )
+
+        started_at = datetime(2026, 1, 1, 12, 0, 0, tzinfo=UTC)
+        sample_workflow_execution.started_at = started_at
+
+        session = mock_session_factory.return_value.__enter__.return_value
+        session.get.return_value = None
+
+        repo.save(sample_workflow_execution)
+
+        saved_model = session.merge.call_args.args[0]
+        assert saved_model.created_at == started_at
+        session.commit.assert_called_once()
+
+    def test_save_preserves_existing_created_at_when_record_already_exists(
+        self, mock_session_factory, mock_account, sample_workflow_execution
+    ):
+        repo = SQLAlchemyWorkflowExecutionRepository(
+            session_factory=mock_session_factory,
+            user=mock_account,
+            app_id="test_app",
+            triggered_from=WorkflowRunTriggeredFrom.APP_RUN,
+        )
+
+        execution_id = sample_workflow_execution.id_
+        existing_created_at = datetime(2026, 1, 1, 12, 0, 0, tzinfo=UTC)
+
+        existing_run = WorkflowRun()
+        existing_run.id = execution_id
+        existing_run.tenant_id = repo._tenant_id
+        existing_run.created_at = existing_created_at
+
+        session = mock_session_factory.return_value.__enter__.return_value
+        session.get.return_value = existing_run
+
+        sample_workflow_execution.started_at = datetime(2026, 1, 1, 12, 30, 0, tzinfo=UTC)
+
+        repo.save(sample_workflow_execution)
+
+        saved_model = session.merge.call_args.args[0]
+        assert saved_model.created_at == existing_created_at
+        session.commit.assert_called_once()

api/tests/unit_tests/core/repositories/test_sqlalchemy_workflow_node_execution_repository.py

@@ -0,0 +1,772 @@
+from __future__ import annotations
+
+import json
+import logging
+from collections.abc import Mapping
+from datetime import UTC, datetime
+from types import SimpleNamespace
+from typing import Any
+from unittest.mock import MagicMock, Mock
+
+import psycopg2.errors
+import pytest
+from sqlalchemy import Engine, create_engine
+from sqlalchemy.exc import IntegrityError
+from sqlalchemy.orm import sessionmaker
+
+from configs import dify_config
+from core.repositories.sqlalchemy_workflow_node_execution_repository import (
+    SQLAlchemyWorkflowNodeExecutionRepository,
+    _deterministic_json_dump,
+    _filter_by_offload_type,
+    _find_first,
+    _replace_or_append_offload,
+)
+from dify_graph.entities import WorkflowNodeExecution
+from dify_graph.enums import (
+    NodeType,
+    WorkflowNodeExecutionMetadataKey,
+    WorkflowNodeExecutionStatus,
+)
+from dify_graph.repositories.workflow_node_execution_repository import OrderConfig
+from models import Account, EndUser
+from models.enums import ExecutionOffLoadType
+from models.workflow import WorkflowNodeExecutionModel, WorkflowNodeExecutionOffload, WorkflowNodeExecutionTriggeredFrom
+
+
+def _mock_account(*, tenant_id: str = "tenant", user_id: str = "user") -> Account:
+    user = Mock(spec=Account)
+    user.id = user_id
+    user.current_tenant_id = tenant_id
+    return user
+
+
+def _mock_end_user(*, tenant_id: str = "tenant", user_id: str = "user") -> EndUser:
+    user = Mock(spec=EndUser)
+    user.id = user_id
+    user.tenant_id = tenant_id
+    return user
+
+
+def _execution(
+    *,
+    execution_id: str = "exec-id",
+    node_execution_id: str = "node-exec-id",
+    workflow_run_id: str = "run-id",
+    status: WorkflowNodeExecutionStatus = WorkflowNodeExecutionStatus.SUCCEEDED,
+    inputs: Mapping[str, Any] | None = None,
+    outputs: Mapping[str, Any] | None = None,
+    process_data: Mapping[str, Any] | None = None,
+    metadata: Mapping[WorkflowNodeExecutionMetadataKey, Any] | None = None,
+) -> WorkflowNodeExecution:
+    return WorkflowNodeExecution(
+        id=execution_id,
+        node_execution_id=node_execution_id,
+        workflow_id="workflow-id",
+        workflow_execution_id=workflow_run_id,
+        index=1,
+        predecessor_node_id=None,
+        node_id="node-id",
+        node_type=NodeType.LLM,
+        title="Title",
+        inputs=inputs,
+        outputs=outputs,
+        process_data=process_data,
+        status=status,
+        error=None,
+        elapsed_time=1.0,
+        metadata=metadata,
+        created_at=datetime.now(UTC),
+        finished_at=None,
+    )
+
+
+class _SessionCtx:
+    def __init__(self, session: Any):
+        self._session = session
+
+    def __enter__(self) -> Any:
+        return self._session
+
+    def __exit__(self, exc_type, exc, tb) -> None:
+        return None
+
+
+def _session_factory(session: Any) -> sessionmaker:
+    factory = Mock(spec=sessionmaker)
+    factory.return_value = _SessionCtx(session)
+    return factory
+
+
+def test_init_accepts_engine_and_sessionmaker_and_sets_role(monkeypatch: pytest.MonkeyPatch) -> None:
+    monkeypatch.setattr(
+        "core.repositories.sqlalchemy_workflow_node_execution_repository.FileService",
+        lambda *_: SimpleNamespace(upload_file=Mock()),
+    )
+
+    engine: Engine = create_engine("sqlite:///:memory:")
+    repo = SQLAlchemyWorkflowNodeExecutionRepository(
+        session_factory=engine,
+        user=_mock_account(),
+        app_id=None,
+        triggered_from=WorkflowNodeExecutionTriggeredFrom.WORKFLOW_RUN,
+    )
+    assert isinstance(repo._session_factory, sessionmaker)
+
+    sm = Mock(spec=sessionmaker)
+    repo = SQLAlchemyWorkflowNodeExecutionRepository(
+        session_factory=sm,
+        user=_mock_end_user(),
+        app_id="app",
+        triggered_from=WorkflowNodeExecutionTriggeredFrom.SINGLE_STEP,
+    )
+    assert repo._creator_user_role.value == "end_user"
+
+
+def test_init_rejects_invalid_session_factory_type(monkeypatch: pytest.MonkeyPatch) -> None:
+    monkeypatch.setattr(
+        "core.repositories.sqlalchemy_workflow_node_execution_repository.FileService",
+        lambda *_: SimpleNamespace(upload_file=Mock()),
+    )
+    with pytest.raises(ValueError, match="Invalid session_factory type"):
+        SQLAlchemyWorkflowNodeExecutionRepository(  # type: ignore[arg-type]
+            session_factory=object(),
+            user=_mock_account(),
+            app_id=None,
+            triggered_from=WorkflowNodeExecutionTriggeredFrom.WORKFLOW_RUN,
+        )
+
+
+def test_init_requires_tenant_id(monkeypatch: pytest.MonkeyPatch) -> None:
+    monkeypatch.setattr(
+        "core.repositories.sqlalchemy_workflow_node_execution_repository.FileService",
+        lambda *_: SimpleNamespace(upload_file=Mock()),
+    )
+    user = _mock_account()
+    user.current_tenant_id = None
+    with pytest.raises(ValueError, match="User must have a tenant_id"):
+        SQLAlchemyWorkflowNodeExecutionRepository(
+            session_factory=Mock(spec=sessionmaker),
+            user=user,
+            app_id=None,
+            triggered_from=WorkflowNodeExecutionTriggeredFrom.WORKFLOW_RUN,
+        )
+
+
+def test_create_truncator_uses_config(monkeypatch: pytest.MonkeyPatch) -> None:
+    created: dict[str, Any] = {}
+
+    class FakeTruncator:
+        def __init__(self, *, max_size_bytes: int, array_element_limit: int, string_length_limit: int):
+            created.update(
+                {
+                    "max_size_bytes": max_size_bytes,
+                    "array_element_limit": array_element_limit,
+                    "string_length_limit": string_length_limit,
+                }
+            )
+
+    monkeypatch.setattr(
+        "core.repositories.sqlalchemy_workflow_node_execution_repository.VariableTruncator",
+        FakeTruncator,
+    )
+    monkeypatch.setattr(
+        "core.repositories.sqlalchemy_workflow_node_execution_repository.FileService",
+        lambda *_: SimpleNamespace(upload_file=Mock()),
+    )
+
+    repo = SQLAlchemyWorkflowNodeExecutionRepository(
+        session_factory=Mock(spec=sessionmaker),
+        user=_mock_account(),
+        app_id=None,
+        triggered_from=WorkflowNodeExecutionTriggeredFrom.WORKFLOW_RUN,
+    )
+    _ = repo._create_truncator()
+    assert created["max_size_bytes"] == dify_config.WORKFLOW_VARIABLE_TRUNCATION_MAX_SIZE
+
+
+def test_helpers_find_first_and_replace_or_append_and_filter() -> None:
+    assert _deterministic_json_dump({"b": 1, "a": 2}) == '{"a": 2, "b": 1}'
+    assert _find_first([], lambda _: True) is None
+    assert _find_first([1, 2, 3], lambda x: x > 1) == 2
+
+    off1 = WorkflowNodeExecutionOffload(type_=ExecutionOffLoadType.INPUTS)
+    off2 = WorkflowNodeExecutionOffload(type_=ExecutionOffLoadType.OUTPUTS)
+    assert _find_first([off1, off2], _filter_by_offload_type(ExecutionOffLoadType.OUTPUTS)) is off2
+
+    replaced = _replace_or_append_offload([off1, off2], WorkflowNodeExecutionOffload(type_=ExecutionOffLoadType.INPUTS))
+    assert len(replaced) == 2
+    assert [o.type_ for o in replaced] == [ExecutionOffLoadType.OUTPUTS, ExecutionOffLoadType.INPUTS]
+
+
+def test_to_db_model_requires_constructor_context(monkeypatch: pytest.MonkeyPatch) -> None:
+    monkeypatch.setattr(
+        "core.repositories.sqlalchemy_workflow_node_execution_repository.FileService",
+        lambda *_: SimpleNamespace(upload_file=Mock()),
+    )
+    repo = SQLAlchemyWorkflowNodeExecutionRepository(
+        session_factory=Mock(spec=sessionmaker),
+        user=_mock_account(),
+        app_id=None,
+        triggered_from=WorkflowNodeExecutionTriggeredFrom.WORKFLOW_RUN,
+    )
+    execution = _execution(inputs={"b": 1, "a": 2}, metadata={WorkflowNodeExecutionMetadataKey.TOTAL_TOKENS: 1})
+
+    # Happy path: deterministic json dump should be sorted
+    db_model = repo._to_db_model(execution)
+    assert json.loads(db_model.inputs or "{}") == {"a": 2, "b": 1}
+    assert json.loads(db_model.execution_metadata or "{}")["total_tokens"] == 1
+
+    repo._triggered_from = None
+    with pytest.raises(ValueError, match="triggered_from is required"):
+        repo._to_db_model(execution)
+
+
+def test_to_db_model_requires_creator_user_id_and_role(monkeypatch: pytest.MonkeyPatch) -> None:
+    monkeypatch.setattr(
+        "core.repositories.sqlalchemy_workflow_node_execution_repository.FileService",
+        lambda *_: SimpleNamespace(upload_file=Mock()),
+    )
+    repo = SQLAlchemyWorkflowNodeExecutionRepository(
+        session_factory=Mock(spec=sessionmaker),
+        user=_mock_account(),
+        app_id="app",
+        triggered_from=WorkflowNodeExecutionTriggeredFrom.WORKFLOW_RUN,
+    )
+    execution = _execution()
+    db_model = repo._to_db_model(execution)
+    assert db_model.app_id == "app"
+
+    repo._creator_user_id = None
+    with pytest.raises(ValueError, match="created_by is required"):
+        repo._to_db_model(execution)
+
+    repo._creator_user_id = "user"
+    repo._creator_user_role = None
+    with pytest.raises(ValueError, match="created_by_role is required"):
+        repo._to_db_model(execution)
+
+
+def test_is_duplicate_key_error_and_regenerate_id(
+    monkeypatch: pytest.MonkeyPatch, caplog: pytest.LogCaptureFixture
+) -> None:
+    monkeypatch.setattr(
+        "core.repositories.sqlalchemy_workflow_node_execution_repository.FileService",
+        lambda *_: SimpleNamespace(upload_file=Mock()),
+    )
+    repo = SQLAlchemyWorkflowNodeExecutionRepository(
+        session_factory=Mock(spec=sessionmaker),
+        user=_mock_account(),
+        app_id=None,
+        triggered_from=WorkflowNodeExecutionTriggeredFrom.WORKFLOW_RUN,
+    )
+
+    unique = Mock(spec=psycopg2.errors.UniqueViolation)
+    duplicate_error = IntegrityError("dup", params=None, orig=unique)
+    assert repo._is_duplicate_key_error(duplicate_error) is True
+    assert repo._is_duplicate_key_error(IntegrityError("other", params=None, orig=None)) is False
+
+    execution = _execution(execution_id="old-id")
+    db_model = WorkflowNodeExecutionModel()
+    db_model.id = "old-id"
+    monkeypatch.setattr("core.repositories.sqlalchemy_workflow_node_execution_repository.uuidv7", lambda: "new-id")
+    caplog.set_level(logging.WARNING)
+    repo._regenerate_id_on_duplicate(execution, db_model)
+    assert execution.id == "new-id"
+    assert db_model.id == "new-id"
+    assert any("Duplicate key conflict" in r.message for r in caplog.records)
+
+
+def test_persist_to_database_updates_existing_and_inserts_new(monkeypatch: pytest.MonkeyPatch) -> None:
+    monkeypatch.setattr(
+        "core.repositories.sqlalchemy_workflow_node_execution_repository.FileService",
+        lambda *_: SimpleNamespace(upload_file=Mock()),
+    )
+    session = MagicMock()
+    repo = SQLAlchemyWorkflowNodeExecutionRepository(
+        session_factory=_session_factory(session),
+        user=_mock_account(),
+        app_id=None,
+        triggered_from=WorkflowNodeExecutionTriggeredFrom.WORKFLOW_RUN,
+    )
+
+    db_model = WorkflowNodeExecutionModel()
+    db_model.id = "id1"
+    db_model.node_execution_id = "node1"
+    db_model.foo = "bar"  # type: ignore[attr-defined]
+    db_model.__dict__["_private"] = "x"
+
+    existing = SimpleNamespace()
+    session.get.return_value = existing
+    repo._persist_to_database(db_model)
+    assert existing.foo == "bar"
+    session.add.assert_not_called()
+    assert repo._node_execution_cache["node1"] is db_model
+
+    session.reset_mock()
+    session.get.return_value = None
+    repo._node_execution_cache.clear()
+    repo._persist_to_database(db_model)
+    session.add.assert_called_once_with(db_model)
+    assert repo._node_execution_cache["node1"] is db_model
+
+
+def test_truncate_and_upload_returns_none_when_no_values_or_not_truncated(monkeypatch: pytest.MonkeyPatch) -> None:
+    monkeypatch.setattr(
+        "core.repositories.sqlalchemy_workflow_node_execution_repository.FileService",
+        lambda *_: SimpleNamespace(upload_file=Mock()),
+    )
+    repo = SQLAlchemyWorkflowNodeExecutionRepository(
+        session_factory=Mock(spec=sessionmaker),
+        user=_mock_account(),
+        app_id="app",
+        triggered_from=WorkflowNodeExecutionTriggeredFrom.WORKFLOW_RUN,
+    )
+
+    assert repo._truncate_and_upload(None, "e", ExecutionOffLoadType.INPUTS) is None
+
+    class FakeTruncator:
+        def truncate_variable_mapping(self, value: Any):  # type: ignore[no-untyped-def]
+            return value, False
+
+    monkeypatch.setattr(repo, "_create_truncator", lambda: FakeTruncator())
+    assert repo._truncate_and_upload({"a": 1}, "e", ExecutionOffLoadType.INPUTS) is None
+
+
+def test_truncate_and_upload_uploads_and_builds_offload(monkeypatch: pytest.MonkeyPatch) -> None:
+    uploaded: dict[str, Any] = {}
+
+    class FakeFileService:
+        def upload_file(self, *, filename: str, content: bytes, mimetype: str, user: Any):  # type: ignore[no-untyped-def]
+            uploaded.update({"filename": filename, "content": content, "mimetype": mimetype, "user": user})
+            return SimpleNamespace(id="file-id", key="file-key")
+
+    monkeypatch.setattr(
+        "core.repositories.sqlalchemy_workflow_node_execution_repository.FileService", lambda *_: FakeFileService()
+    )
+    monkeypatch.setattr("core.repositories.sqlalchemy_workflow_node_execution_repository.uuidv7", lambda: "offload-id")
+
+    repo = SQLAlchemyWorkflowNodeExecutionRepository(
+        session_factory=Mock(spec=sessionmaker),
+        user=_mock_account(),
+        app_id="app",
+        triggered_from=WorkflowNodeExecutionTriggeredFrom.WORKFLOW_RUN,
+    )
+
+    class FakeTruncator:
+        def truncate_variable_mapping(self, value: Any):  # type: ignore[no-untyped-def]
+            return {"truncated": True}, True
+
+    monkeypatch.setattr(repo, "_create_truncator", lambda: FakeTruncator())
+
+    result = repo._truncate_and_upload({"a": 1}, "exec", ExecutionOffLoadType.INPUTS)
+    assert result is not None
+    assert result.truncated_value == {"truncated": True}
+    assert uploaded["filename"].startswith("node_execution_exec_inputs.json")
+    assert result.offload.file_id == "file-id"
+    assert result.offload.type_ == ExecutionOffLoadType.INPUTS
+
+
+def test_to_domain_model_loads_offloaded_files(monkeypatch: pytest.MonkeyPatch) -> None:
+    monkeypatch.setattr(
+        "core.repositories.sqlalchemy_workflow_node_execution_repository.FileService",
+        lambda *_: SimpleNamespace(upload_file=Mock()),
+    )
+    repo = SQLAlchemyWorkflowNodeExecutionRepository(
+        session_factory=Mock(spec=sessionmaker),
+        user=_mock_account(),
+        app_id=None,
+        triggered_from=WorkflowNodeExecutionTriggeredFrom.WORKFLOW_RUN,
+    )
+
+    db_model = WorkflowNodeExecutionModel()
+    db_model.id = "id"
+    db_model.node_execution_id = "node-exec"
+    db_model.workflow_id = "wf"
+    db_model.workflow_run_id = "run"
+    db_model.index = 1
+    db_model.predecessor_node_id = None
+    db_model.node_id = "node"
+    db_model.node_type = NodeType.LLM
+    db_model.title = "t"
+    db_model.inputs = json.dumps({"trunc": "i"})
+    db_model.process_data = json.dumps({"trunc": "p"})
+    db_model.outputs = json.dumps({"trunc": "o"})
+    db_model.status = WorkflowNodeExecutionStatus.SUCCEEDED
+    db_model.error = None
+    db_model.elapsed_time = 0.1
+    db_model.execution_metadata = json.dumps({"total_tokens": 3})
+    db_model.created_at = datetime.now(UTC)
+    db_model.finished_at = None
+
+    off_in = WorkflowNodeExecutionOffload(type_=ExecutionOffLoadType.INPUTS)
+    off_out = WorkflowNodeExecutionOffload(type_=ExecutionOffLoadType.OUTPUTS)
+    off_proc = WorkflowNodeExecutionOffload(type_=ExecutionOffLoadType.PROCESS_DATA)
+    off_in.file = SimpleNamespace(key="k-in")
+    off_out.file = SimpleNamespace(key="k-out")
+    off_proc.file = SimpleNamespace(key="k-proc")
+    db_model.offload_data = [off_out, off_in, off_proc]
+
+    def fake_load(key: str) -> bytes:
+        return json.dumps({"full": key}).encode()
+
+    monkeypatch.setattr("core.repositories.sqlalchemy_workflow_node_execution_repository.storage.load", fake_load)
+
+    domain = repo._to_domain_model(db_model)
+    assert domain.inputs == {"full": "k-in"}
+    assert domain.outputs == {"full": "k-out"}
+    assert domain.process_data == {"full": "k-proc"}
+    assert domain.get_truncated_inputs() == {"trunc": "i"}
+    assert domain.get_truncated_outputs() == {"trunc": "o"}
+    assert domain.get_truncated_process_data() == {"trunc": "p"}
+
+
+def test_to_domain_model_returns_early_when_no_offload_data(monkeypatch: pytest.MonkeyPatch) -> None:
+    monkeypatch.setattr(
+        "core.repositories.sqlalchemy_workflow_node_execution_repository.FileService",
+        lambda *_: SimpleNamespace(upload_file=Mock()),
+    )
+    repo = SQLAlchemyWorkflowNodeExecutionRepository(
+        session_factory=Mock(spec=sessionmaker),
+        user=_mock_account(),
+        app_id=None,
+        triggered_from=WorkflowNodeExecutionTriggeredFrom.WORKFLOW_RUN,
+    )
+
+    db_model = WorkflowNodeExecutionModel()
+    db_model.id = "id"
+    db_model.node_execution_id = "node-exec"
+    db_model.workflow_id = "wf"
+    db_model.workflow_run_id = "run"
+    db_model.index = 1
+    db_model.predecessor_node_id = None
+    db_model.node_id = "node"
+    db_model.node_type = NodeType.LLM
+    db_model.title = "t"
+    db_model.inputs = json.dumps({"i": 1})
+    db_model.process_data = json.dumps({"p": 2})
+    db_model.outputs = json.dumps({"o": 3})
+    db_model.status = WorkflowNodeExecutionStatus.SUCCEEDED
+    db_model.error = None
+    db_model.elapsed_time = 0.1
+    db_model.execution_metadata = "{}"
+    db_model.created_at = datetime.now(UTC)
+    db_model.finished_at = None
+    db_model.offload_data = []
+
+    domain = repo._to_domain_model(db_model)
+    assert domain.inputs == {"i": 1}
+    assert domain.outputs == {"o": 3}
+
+
+def test_json_encode_uses_runtime_converter(monkeypatch: pytest.MonkeyPatch) -> None:
+    class FakeConverter:
+        def to_json_encodable(self, values: Mapping[str, Any]) -> Mapping[str, Any]:
+            return {"wrapped": values["a"]}
+
+    monkeypatch.setattr(
+        "core.repositories.sqlalchemy_workflow_node_execution_repository.WorkflowRuntimeTypeConverter",
+        FakeConverter,
+    )
+    assert SQLAlchemyWorkflowNodeExecutionRepository._json_encode({"a": 1}) == '{"wrapped": 1}'
+
+
+def test_save_execution_data_handles_existing_db_model_and_truncation(monkeypatch: pytest.MonkeyPatch) -> None:
+    monkeypatch.setattr(
+        "core.repositories.sqlalchemy_workflow_node_execution_repository.FileService",
+        lambda *_: SimpleNamespace(upload_file=Mock()),
+    )
+    session = MagicMock()
+    session.execute.return_value.scalars.return_value.first.return_value = SimpleNamespace(
+        id="id",
+        offload_data=[WorkflowNodeExecutionOffload(type_=ExecutionOffLoadType.INPUTS)],
+        inputs=None,
+        outputs=None,
+        process_data=None,
+    )
+    session.merge = Mock()
+    session.flush = Mock()
+    session.begin.return_value.__enter__ = Mock(return_value=session)
+    session.begin.return_value.__exit__ = Mock(return_value=None)
+
+    repo = SQLAlchemyWorkflowNodeExecutionRepository(
+        session_factory=_session_factory(session),
+        user=_mock_account(),
+        app_id="app",
+        triggered_from=WorkflowNodeExecutionTriggeredFrom.WORKFLOW_RUN,
+    )
+
+    execution = _execution(inputs={"a": 1}, outputs={"b": 2}, process_data={"c": 3})
+
+    trunc_result = SimpleNamespace(
+        truncated_value={"trunc": True},
+        offload=WorkflowNodeExecutionOffload(type_=ExecutionOffLoadType.INPUTS, file_id="f1"),
+    )
+    monkeypatch.setattr(
+        repo, "_truncate_and_upload", lambda values, *_args, **_kwargs: trunc_result if values == {"a": 1} else None
+    )
+    monkeypatch.setattr(repo, "_json_encode", lambda values: json.dumps(values, sort_keys=True))
+
+    repo.save_execution_data(execution)
+    # Inputs should be truncated, outputs/process_data encoded directly
+    db_model = session.merge.call_args.args[0]
+    assert json.loads(db_model.inputs) == {"trunc": True}
+    assert json.loads(db_model.outputs) == {"b": 2}
+    assert json.loads(db_model.process_data) == {"c": 3}
+    assert any(off.type_ == ExecutionOffLoadType.INPUTS for off in db_model.offload_data)
+    assert execution.get_truncated_inputs() == {"trunc": True}
+
+
+def test_save_execution_data_truncates_outputs_and_process_data(monkeypatch: pytest.MonkeyPatch) -> None:
+    monkeypatch.setattr(
+        "core.repositories.sqlalchemy_workflow_node_execution_repository.FileService",
+        lambda *_: SimpleNamespace(upload_file=Mock()),
+    )
+    existing = SimpleNamespace(
+        id="id",
+        offload_data=[],
+        inputs=None,
+        outputs=None,
+        process_data=None,
+    )
+    session = MagicMock()
+    session.execute.return_value.scalars.return_value.first.return_value = existing
+    session.merge = Mock()
+    session.flush = Mock()
+    session.begin.return_value.__enter__ = Mock(return_value=session)
+    session.begin.return_value.__exit__ = Mock(return_value=None)
+
+    repo = SQLAlchemyWorkflowNodeExecutionRepository(
+        session_factory=_session_factory(session),
+        user=_mock_account(),
+        app_id="app",
+        triggered_from=WorkflowNodeExecutionTriggeredFrom.WORKFLOW_RUN,
+    )
+
+    execution = _execution(inputs={"a": 1}, outputs={"b": 2}, process_data={"c": 3})
+
+    def trunc(values: Mapping[str, Any], *_args: Any, **_kwargs: Any) -> Any:
+        if values == {"b": 2}:
+            return SimpleNamespace(
+                truncated_value={"b": "trunc"},
+                offload=WorkflowNodeExecutionOffload(type_=ExecutionOffLoadType.OUTPUTS, file_id="f2"),
+            )
+        if values == {"c": 3}:
+            return SimpleNamespace(
+                truncated_value={"c": "trunc"},
+                offload=WorkflowNodeExecutionOffload(type_=ExecutionOffLoadType.PROCESS_DATA, file_id="f3"),
+            )
+        return None
+
+    monkeypatch.setattr(repo, "_truncate_and_upload", trunc)
+    monkeypatch.setattr(repo, "_json_encode", lambda values: json.dumps(values, sort_keys=True))
+
+    repo.save_execution_data(execution)
+    db_model = session.merge.call_args.args[0]
+    assert json.loads(db_model.outputs) == {"b": "trunc"}
+    assert json.loads(db_model.process_data) == {"c": "trunc"}
+    assert execution.get_truncated_outputs() == {"b": "trunc"}
+    assert execution.get_truncated_process_data() == {"c": "trunc"}
+
+
+def test_save_execution_data_handles_missing_db_model(monkeypatch: pytest.MonkeyPatch) -> None:
+    monkeypatch.setattr(
+        "core.repositories.sqlalchemy_workflow_node_execution_repository.FileService",
+        lambda *_: SimpleNamespace(upload_file=Mock()),
+    )
+    session = MagicMock()
+    session.execute.return_value.scalars.return_value.first.return_value = None
+    session.merge = Mock()
+    session.flush = Mock()
+    session.begin.return_value.__enter__ = Mock(return_value=session)
+    session.begin.return_value.__exit__ = Mock(return_value=None)
+
+    repo = SQLAlchemyWorkflowNodeExecutionRepository(
+        session_factory=_session_factory(session),
+        user=_mock_account(),
+        app_id=None,
+        triggered_from=WorkflowNodeExecutionTriggeredFrom.WORKFLOW_RUN,
+    )
+
+    execution = _execution(inputs={"a": 1})
+    fake_db_model = SimpleNamespace(id=execution.id, offload_data=[], inputs=None, outputs=None, process_data=None)
+    monkeypatch.setattr(repo, "_to_db_model", lambda *_: fake_db_model)
+    monkeypatch.setattr(repo, "_truncate_and_upload", lambda *_args, **_kwargs: None)
+    monkeypatch.setattr(repo, "_json_encode", lambda values: json.dumps(values))
+
+    repo.save_execution_data(execution)
+    merged = session.merge.call_args.args[0]
+    assert merged.inputs == '{"a": 1}'
+
+
+def test_save_retries_duplicate_and_logs_non_duplicate(
+    monkeypatch: pytest.MonkeyPatch, caplog: pytest.LogCaptureFixture
+) -> None:
+    monkeypatch.setattr(
+        "core.repositories.sqlalchemy_workflow_node_execution_repository.FileService",
+        lambda *_: SimpleNamespace(upload_file=Mock()),
+    )
+    repo = SQLAlchemyWorkflowNodeExecutionRepository(
+        session_factory=Mock(spec=sessionmaker),
+        user=_mock_account(),
+        app_id=None,
+        triggered_from=WorkflowNodeExecutionTriggeredFrom.WORKFLOW_RUN,
+    )
+
+    execution = _execution(execution_id="id")
+    unique = Mock(spec=psycopg2.errors.UniqueViolation)
+    duplicate_error = IntegrityError("dup", params=None, orig=unique)
+    other_error = IntegrityError("other", params=None, orig=None)
+
+    calls = {"n": 0}
+
+    def persist(_db_model: Any) -> None:
+        calls["n"] += 1
+        if calls["n"] == 1:
+            raise duplicate_error
+
+    monkeypatch.setattr(repo, "_persist_to_database", persist)
+    monkeypatch.setattr("core.repositories.sqlalchemy_workflow_node_execution_repository.uuidv7", lambda: "new-id")
+    repo.save(execution)
+    assert execution.id == "new-id"
+    assert repo._node_execution_cache[execution.node_execution_id] is not None
+
+    caplog.set_level(logging.ERROR)
+    monkeypatch.setattr(repo, "_persist_to_database", lambda _db: (_ for _ in ()).throw(other_error))
+    with pytest.raises(IntegrityError):
+        repo.save(_execution(execution_id="id2", node_execution_id="node2"))
+    assert any("Non-duplicate key integrity error" in r.message for r in caplog.records)
+
+
+def test_save_logs_and_reraises_on_unexpected_error(
+    monkeypatch: pytest.MonkeyPatch, caplog: pytest.LogCaptureFixture
+) -> None:
+    monkeypatch.setattr(
+        "core.repositories.sqlalchemy_workflow_node_execution_repository.FileService",
+        lambda *_: SimpleNamespace(upload_file=Mock()),
+    )
+    repo = SQLAlchemyWorkflowNodeExecutionRepository(
+        session_factory=Mock(spec=sessionmaker),
+        user=_mock_account(),
+        app_id=None,
+        triggered_from=WorkflowNodeExecutionTriggeredFrom.WORKFLOW_RUN,
+    )
+    caplog.set_level(logging.ERROR)
+    monkeypatch.setattr(repo, "_persist_to_database", lambda _db: (_ for _ in ()).throw(RuntimeError("boom")))
+    with pytest.raises(RuntimeError, match="boom"):
+        repo.save(_execution(execution_id="id3", node_execution_id="node3"))
+    assert any("Failed to save workflow node execution" in r.message for r in caplog.records)
+
+
+def test_get_db_models_by_workflow_run_orders_and_caches(monkeypatch: pytest.MonkeyPatch) -> None:
+    monkeypatch.setattr(
+        "core.repositories.sqlalchemy_workflow_node_execution_repository.FileService",
+        lambda *_: SimpleNamespace(upload_file=Mock()),
+    )
+
+    class FakeStmt:
+        def __init__(self) -> None:
+            self.where_calls = 0
+            self.order_by_args: tuple[Any, ...] | None = None
+
+        def where(self, *_args: Any) -> FakeStmt:
+            self.where_calls += 1
+            return self
+
+        def order_by(self, *args: Any) -> FakeStmt:
+            self.order_by_args = args
+            return self
+
+    stmt = FakeStmt()
+    monkeypatch.setattr(
+        "core.repositories.sqlalchemy_workflow_node_execution_repository.WorkflowNodeExecutionModel.preload_offload_data_and_files",
+        lambda _q: stmt,
+    )
+    monkeypatch.setattr("core.repositories.sqlalchemy_workflow_node_execution_repository.select", lambda *_: "select")
+
+    model1 = SimpleNamespace(node_execution_id="n1")
+    model2 = SimpleNamespace(node_execution_id=None)
+    session = MagicMock()
+    session.scalars.return_value.all.return_value = [model1, model2]
+
+    repo = SQLAlchemyWorkflowNodeExecutionRepository(
+        session_factory=_session_factory(session),
+        user=_mock_account(),
+        app_id="app",
+        triggered_from=WorkflowNodeExecutionTriggeredFrom.WORKFLOW_RUN,
+    )
+
+    order = OrderConfig(order_by=["index", "missing"], order_direction="desc")
+    db_models = repo.get_db_models_by_workflow_run("run", order)
+    assert db_models == [model1, model2]
+    assert repo._node_execution_cache["n1"] is model1
+    assert stmt.order_by_args is not None
+
+
+def test_get_db_models_by_workflow_run_uses_asc_order(monkeypatch: pytest.MonkeyPatch) -> None:
+    monkeypatch.setattr(
+        "core.repositories.sqlalchemy_workflow_node_execution_repository.FileService",
+        lambda *_: SimpleNamespace(upload_file=Mock()),
+    )
+
+    class FakeStmt:
+        def where(self, *_args: Any) -> FakeStmt:
+            return self
+
+        def order_by(self, *args: Any) -> FakeStmt:
+            self.args = args  # type: ignore[attr-defined]
+            return self
+
+    stmt = FakeStmt()
+    monkeypatch.setattr(
+        "core.repositories.sqlalchemy_workflow_node_execution_repository.WorkflowNodeExecutionModel.preload_offload_data_and_files",
+        lambda _q: stmt,
+    )
+    monkeypatch.setattr("core.repositories.sqlalchemy_workflow_node_execution_repository.select", lambda *_: "select")
+
+    session = MagicMock()
+    session.scalars.return_value.all.return_value = []
+    repo = SQLAlchemyWorkflowNodeExecutionRepository(
+        session_factory=_session_factory(session),
+        user=_mock_account(),
+        app_id=None,
+        triggered_from=WorkflowNodeExecutionTriggeredFrom.WORKFLOW_RUN,
+    )
+    repo.get_db_models_by_workflow_run("run", OrderConfig(order_by=["index"], order_direction="asc"))
+
+
+def test_get_by_workflow_run_maps_to_domain(monkeypatch: pytest.MonkeyPatch) -> None:
+    monkeypatch.setattr(
+        "core.repositories.sqlalchemy_workflow_node_execution_repository.FileService",
+        lambda *_: SimpleNamespace(upload_file=Mock()),
+    )
+
+    repo = SQLAlchemyWorkflowNodeExecutionRepository(
+        session_factory=Mock(spec=sessionmaker),
+        user=_mock_account(),
+        app_id=None,
+        triggered_from=WorkflowNodeExecutionTriggeredFrom.WORKFLOW_RUN,
+    )
+
+    db_models = [SimpleNamespace(id="db1"), SimpleNamespace(id="db2")]
+    monkeypatch.setattr(repo, "get_db_models_by_workflow_run", lambda *_args, **_kwargs: db_models)
+    monkeypatch.setattr(repo, "_to_domain_model", lambda m: f"domain:{m.id}")
+
+    class FakeExecutor:
+        def __enter__(self) -> FakeExecutor:
+            return self
+
+        def __exit__(self, exc_type, exc, tb) -> None:
+            return None
+
+        def map(self, func, items, timeout: int):  # type: ignore[no-untyped-def]
+            assert timeout == 30
+            return list(map(func, items))
+
+    monkeypatch.setattr(
+        "core.repositories.sqlalchemy_workflow_node_execution_repository.ThreadPoolExecutor",
+        lambda max_workers: FakeExecutor(),
+    )
+
+    result = repo.get_by_workflow_run("run", order_config=None)
+    assert result == ["domain:db1", "domain:db2"]

api/tests/unit_tests/core/schemas/test_registry.py

@@ -0,0 +1,137 @@
+import json
+from unittest.mock import patch
+
+from core.schemas.registry import SchemaRegistry
+
+
+class TestSchemaRegistry:
+    def test_initialization(self, tmp_path):
+        base_dir = tmp_path / "schemas"
+        base_dir.mkdir()
+        registry = SchemaRegistry(str(base_dir))
+        assert registry.base_dir == base_dir
+        assert registry.versions == {}
+        assert registry.metadata == {}
+
+    def test_default_registry_singleton(self):
+        registry1 = SchemaRegistry.default_registry()
+        registry2 = SchemaRegistry.default_registry()
+        assert registry1 is registry2
+        assert isinstance(registry1, SchemaRegistry)
+
+    def test_load_all_versions_non_existent_dir(self, tmp_path):
+        base_dir = tmp_path / "non_existent"
+        registry = SchemaRegistry(str(base_dir))
+        registry.load_all_versions()
+        assert registry.versions == {}
+
+    def test_load_all_versions_filtering(self, tmp_path):
+        base_dir = tmp_path / "schemas"
+        base_dir.mkdir()
+        (base_dir / "not_a_version_dir").mkdir()
+        (base_dir / "v1").mkdir()
+        (base_dir / "some_file.txt").write_text("content")
+
+        registry = SchemaRegistry(str(base_dir))
+        with patch.object(registry, "_load_version_dir") as mock_load:
+            registry.load_all_versions()
+            mock_load.assert_called_once()
+            assert mock_load.call_args[0][0] == "v1"
+
+    def test_load_version_dir_filtering(self, tmp_path):
+        version_dir = tmp_path / "v1"
+        version_dir.mkdir()
+        (version_dir / "schema1.json").write_text("{}")
+        (version_dir / "not_a_schema.txt").write_text("content")
+
+        registry = SchemaRegistry(str(tmp_path))
+        with patch.object(registry, "_load_schema") as mock_load:
+            registry._load_version_dir("v1", version_dir)
+            mock_load.assert_called_once()
+            assert mock_load.call_args[0][1] == "schema1"
+
+    def test_load_version_dir_non_existent(self, tmp_path):
+        version_dir = tmp_path / "non_existent"
+        registry = SchemaRegistry(str(tmp_path))
+        registry._load_version_dir("v1", version_dir)
+        assert "v1" not in registry.versions
+
+    def test_load_schema_success(self, tmp_path):
+        schema_path = tmp_path / "test.json"
+        schema_content = {"title": "Test Schema", "description": "A test schema"}
+        schema_path.write_text(json.dumps(schema_content))
+
+        registry = SchemaRegistry(str(tmp_path))
+        registry.versions["v1"] = {}
+        registry._load_schema("v1", "test", schema_path)
+
+        assert registry.versions["v1"]["test"] == schema_content
+        uri = "https://dify.ai/schemas/v1/test.json"
+        assert registry.metadata[uri]["title"] == "Test Schema"
+        assert registry.metadata[uri]["version"] == "v1"
+
+    def test_load_schema_invalid_json(self, tmp_path, caplog):
+        schema_path = tmp_path / "invalid.json"
+        schema_path.write_text("invalid json")
+
+        registry = SchemaRegistry(str(tmp_path))
+        registry.versions["v1"] = {}
+        registry._load_schema("v1", "invalid", schema_path)
+
+        assert "Failed to load schema v1/invalid" in caplog.text
+
+    def test_load_schema_os_error(self, tmp_path, caplog):
+        schema_path = tmp_path / "error.json"
+        schema_path.write_text("{}")
+
+        registry = SchemaRegistry(str(tmp_path))
+        registry.versions["v1"] = {}
+
+        with patch("builtins.open", side_effect=OSError("Read error")):
+            registry._load_schema("v1", "error", schema_path)
+
+        assert "Failed to load schema v1/error" in caplog.text
+
+    def test_get_schema(self):
+        registry = SchemaRegistry("/tmp")
+        registry.versions = {"v1": {"test": {"type": "object"}}}
+
+        # Valid URI
+        assert registry.get_schema("https://dify.ai/schemas/v1/test.json") == {"type": "object"}
+
+        # Invalid URI
+        assert registry.get_schema("invalid-uri") is None
+
+        # Missing version
+        assert registry.get_schema("https://dify.ai/schemas/v2/test.json") is None
+
+    def test_list_versions(self):
+        registry = SchemaRegistry("/tmp")
+        registry.versions = {"v2": {}, "v1": {}}
+        assert registry.list_versions() == ["v1", "v2"]
+
+    def test_list_schemas(self):
+        registry = SchemaRegistry("/tmp")
+        registry.versions = {"v1": {"b": {}, "a": {}}}
+
+        assert registry.list_schemas("v1") == ["a", "b"]
+        assert registry.list_schemas("v2") == []
+
+    def test_get_all_schemas_for_version(self):
+        registry = SchemaRegistry("/tmp")
+        registry.versions = {"v1": {"test": {"title": "Test Label"}}}
+
+        results = registry.get_all_schemas_for_version("v1")
+        assert len(results) == 1
+        assert results[0]["name"] == "test"
+        assert results[0]["label"] == "Test Label"
+        assert results[0]["schema"] == {"title": "Test Label"}
+
+        # Default label if title missing
+        registry.versions["v1"]["no_title"] = {}
+        results = registry.get_all_schemas_for_version("v1")
+        item = next(r for r in results if r["name"] == "no_title")
+        assert item["label"] == "no_title"
+
+        # Empty if version missing
+        assert registry.get_all_schemas_for_version("v2") == []

api/tests/unit_tests/core/schemas/test_schema_manager.py

@@ -0,0 +1,80 @@
+from unittest.mock import MagicMock, patch
+
+from core.schemas.registry import SchemaRegistry
+from core.schemas.schema_manager import SchemaManager
+
+
+def test_init_with_provided_registry():
+    mock_registry = MagicMock(spec=SchemaRegistry)
+    manager = SchemaManager(registry=mock_registry)
+    assert manager.registry == mock_registry
+
+
+@patch("core.schemas.schema_manager.SchemaRegistry.default_registry")
+def test_init_with_default_registry(mock_default_registry):
+    mock_registry = MagicMock(spec=SchemaRegistry)
+    mock_default_registry.return_value = mock_registry
+
+    manager = SchemaManager()
+
+    mock_default_registry.assert_called_once()
+    assert manager.registry == mock_registry
+
+
+def test_get_all_schema_definitions():
+    mock_registry = MagicMock(spec=SchemaRegistry)
+    expected_definitions = [{"name": "schema1", "schema": {}}, {"name": "schema2", "schema": {}}]
+    mock_registry.get_all_schemas_for_version.return_value = expected_definitions
+
+    manager = SchemaManager(registry=mock_registry)
+    result = manager.get_all_schema_definitions(version="v2")
+
+    mock_registry.get_all_schemas_for_version.assert_called_once_with("v2")
+    assert result == expected_definitions
+
+
+def test_get_schema_by_name_success():
+    mock_registry = MagicMock(spec=SchemaRegistry)
+    mock_schema = {"type": "object"}
+    mock_registry.get_schema.return_value = mock_schema
+
+    manager = SchemaManager(registry=mock_registry)
+    result = manager.get_schema_by_name("my_schema", version="v1")
+
+    expected_uri = "https://dify.ai/schemas/v1/my_schema.json"
+    mock_registry.get_schema.assert_called_once_with(expected_uri)
+    assert result == {"name": "my_schema", "schema": mock_schema}
+
+
+def test_get_schema_by_name_not_found():
+    mock_registry = MagicMock(spec=SchemaRegistry)
+    mock_registry.get_schema.return_value = None
+
+    manager = SchemaManager(registry=mock_registry)
+    result = manager.get_schema_by_name("non_existent", version="v1")
+
+    assert result is None
+
+
+def test_list_available_schemas():
+    mock_registry = MagicMock(spec=SchemaRegistry)
+    expected_schemas = ["schema1", "schema2"]
+    mock_registry.list_schemas.return_value = expected_schemas
+
+    manager = SchemaManager(registry=mock_registry)
+    result = manager.list_available_schemas(version="v1")
+
+    mock_registry.list_schemas.assert_called_once_with("v1")
+    assert result == expected_schemas
+
+
+def test_list_available_versions():
+    mock_registry = MagicMock(spec=SchemaRegistry)
+    expected_versions = ["v1", "v2"]
+    mock_registry.list_versions.return_value = expected_versions
+
+    manager = SchemaManager(registry=mock_registry)
+    result = manager.list_available_versions()
+
+    mock_registry.list_versions.assert_called_once()
+    assert result == expected_versions

api/tests/unit_tests/core/test_provider_manager.py

@@ -1,32 +1,34 @@
+from unittest.mock import Mock, PropertyMock, patch
+
 import pytest
-from pytest_mock import MockerFixture
 
 from core.entities.provider_entities import ModelSettings
 from core.provider_manager import ProviderManager
+from dify_graph.model_runtime.entities.common_entities import I18nObject
 from dify_graph.model_runtime.entities.model_entities import ModelType
 from models.provider import LoadBalancingModelConfig, ProviderModelSetting
 
 
 @pytest.fixture
-def mock_provider_entity(mocker: MockerFixture):
-    mock_entity = mocker.Mock()
+def mock_provider_entity():
+    mock_entity = Mock()
     mock_entity.provider = "openai"
     mock_entity.configurate_methods = ["predefined-model"]
     mock_entity.supported_model_types = [ModelType.LLM]
 
     # Use PropertyMock to ensure credential_form_schemas is iterable
-    provider_credential_schema = mocker.Mock()
-    type(provider_credential_schema).credential_form_schemas = mocker.PropertyMock(return_value=[])
+    provider_credential_schema = Mock()
+    type(provider_credential_schema).credential_form_schemas = PropertyMock(return_value=[])
     mock_entity.provider_credential_schema = provider_credential_schema
 
-    model_credential_schema = mocker.Mock()
-    type(model_credential_schema).credential_form_schemas = mocker.PropertyMock(return_value=[])
+    model_credential_schema = Mock()
+    type(model_credential_schema).credential_form_schemas = PropertyMock(return_value=[])
     mock_entity.model_credential_schema = model_credential_schema
 
     return mock_entity
 
 
-def test__to_model_settings(mocker: MockerFixture, mock_provider_entity):
+def test__to_model_settings(mock_provider_entity):
     # Mocking the inputs
     ps = ProviderModelSetting(
         tenant_id="tenant_id",
@@ -63,18 +65,18 @@ def test__to_model_settings(mocker: MockerFixture, mock_provider_entity):
     load_balancing_model_configs[0].id = "id1"
     load_balancing_model_configs[1].id = "id2"
 
-    mocker.patch(
-        "core.helper.model_provider_cache.ProviderCredentialsCache.get", return_value={"openai_api_key": "fake_key"}
-    )
+    with patch(
+        "core.helper.model_provider_cache.ProviderCredentialsCache.get",
+        return_value={"openai_api_key": "fake_key"},
+    ):
+        provider_manager = ProviderManager()
 
-    provider_manager = ProviderManager()
-
-    # Running the method
-    result = provider_manager._to_model_settings(
-        provider_entity=mock_provider_entity,
-        provider_model_settings=provider_model_settings,
-        load_balancing_model_configs=load_balancing_model_configs,
-    )
+        # Running the method
+        result = provider_manager._to_model_settings(
+            provider_entity=mock_provider_entity,
+            provider_model_settings=provider_model_settings,
+            load_balancing_model_configs=load_balancing_model_configs,
+        )
 
     # Asserting that the result is as expected
     assert len(result) == 1
@@ -87,7 +89,7 @@ def test__to_model_settings(mocker: MockerFixture, mock_provider_entity):
     assert result[0].load_balancing_configs[1].name == "first"
 
 
-def test__to_model_settings_only_one_lb(mocker: MockerFixture, mock_provider_entity):
+def test__to_model_settings_only_one_lb(mock_provider_entity):
     # Mocking the inputs
 
     ps = ProviderModelSetting(
@@ -113,18 +115,18 @@ def test__to_model_settings_only_one_lb(mocker: MockerFixture, mock_provider_ent
     ]
     load_balancing_model_configs[0].id = "id1"
 
-    mocker.patch(
-        "core.helper.model_provider_cache.ProviderCredentialsCache.get", return_value={"openai_api_key": "fake_key"}
-    )
+    with patch(
+        "core.helper.model_provider_cache.ProviderCredentialsCache.get",
+        return_value={"openai_api_key": "fake_key"},
+    ):
+        provider_manager = ProviderManager()
 
-    provider_manager = ProviderManager()
-
-    # Running the method
-    result = provider_manager._to_model_settings(
-        provider_entity=mock_provider_entity,
-        provider_model_settings=provider_model_settings,
-        load_balancing_model_configs=load_balancing_model_configs,
-    )
+        # Running the method
+        result = provider_manager._to_model_settings(
+            provider_entity=mock_provider_entity,
+            provider_model_settings=provider_model_settings,
+            load_balancing_model_configs=load_balancing_model_configs,
+        )
 
     # Asserting that the result is as expected
     assert len(result) == 1
@@ -135,7 +137,7 @@ def test__to_model_settings_only_one_lb(mocker: MockerFixture, mock_provider_ent
     assert len(result[0].load_balancing_configs) == 0
 
 
-def test__to_model_settings_lb_disabled(mocker: MockerFixture, mock_provider_entity):
+def test__to_model_settings_lb_disabled(mock_provider_entity):
     # Mocking the inputs
     ps = ProviderModelSetting(
         tenant_id="tenant_id",
@@ -170,18 +172,18 @@ def test__to_model_settings_lb_disabled(mocker: MockerFixture, mock_provider_ent
     load_balancing_model_configs[0].id = "id1"
     load_balancing_model_configs[1].id = "id2"
 
-    mocker.patch(
-        "core.helper.model_provider_cache.ProviderCredentialsCache.get", return_value={"openai_api_key": "fake_key"}
-    )
+    with patch(
+        "core.helper.model_provider_cache.ProviderCredentialsCache.get",
+        return_value={"openai_api_key": "fake_key"},
+    ):
+        provider_manager = ProviderManager()
 
-    provider_manager = ProviderManager()
-
-    # Running the method
-    result = provider_manager._to_model_settings(
-        provider_entity=mock_provider_entity,
-        provider_model_settings=provider_model_settings,
-        load_balancing_model_configs=load_balancing_model_configs,
-    )
+        # Running the method
+        result = provider_manager._to_model_settings(
+            provider_entity=mock_provider_entity,
+            provider_model_settings=provider_model_settings,
+            load_balancing_model_configs=load_balancing_model_configs,
+        )
 
     # Asserting that the result is as expected
     assert len(result) == 1
@@ -190,3 +192,39 @@ def test__to_model_settings_lb_disabled(mocker: MockerFixture, mock_provider_ent
     assert result[0].model_type == ModelType.LLM
     assert result[0].enabled is True
     assert len(result[0].load_balancing_configs) == 0
+
+
+def test_get_default_model_uses_first_available_active_model():
+    mock_session = Mock()
+    mock_session.scalar.return_value = None
+
+    provider_configurations = Mock()
+    provider_configurations.get_models.return_value = [
+        Mock(model="gpt-3.5-turbo", provider=Mock(provider="openai")),
+        Mock(model="gpt-4", provider=Mock(provider="openai")),
+    ]
+
+    manager = ProviderManager()
+    with (
+        patch("core.provider_manager.db.session", mock_session),
+        patch.object(manager, "get_configurations", return_value=provider_configurations),
+        patch("core.provider_manager.ModelProviderFactory") as mock_factory_cls,
+    ):
+        mock_factory_cls.return_value.get_provider_schema.return_value = Mock(
+            provider="openai",
+            label=I18nObject(en_US="OpenAI", zh_Hans="OpenAI"),
+            icon_small=I18nObject(en_US="icon_small.png", zh_Hans="icon_small.png"),
+            supported_model_types=[ModelType.LLM],
+        )
+
+        result = manager.get_default_model("tenant-id", ModelType.LLM)
+
+        assert result is not None
+        assert result.model == "gpt-3.5-turbo"
+        assert result.provider.provider == "openai"
+        provider_configurations.get_models.assert_called_once_with(model_type=ModelType.LLM, only_active=True)
+        mock_session.add.assert_called_once()
+        saved_default_model = mock_session.add.call_args.args[0]
+        assert saved_default_model.model_name == "gpt-3.5-turbo"
+        assert saved_default_model.provider_name == "openai"
+        mock_session.commit.assert_called_once()

api/tests/unit_tests/core/workflow/graph_engine/test_mock_nodes.py

@@ -20,7 +20,7 @@ from dify_graph.nodes.code import CodeNode
 from dify_graph.nodes.document_extractor import DocumentExtractorNode
 from dify_graph.nodes.http_request import HttpRequestNode
 from dify_graph.nodes.llm import LLMNode
-from dify_graph.nodes.llm.protocols import CredentialsProvider, ModelFactory
+from dify_graph.nodes.llm.protocols import CredentialsProvider, ModelFactory, TemplateRenderer
 from dify_graph.nodes.parameter_extractor import ParameterExtractorNode
 from dify_graph.nodes.protocols import HttpClientProtocol, ToolFileManagerProtocol
 from dify_graph.nodes.question_classifier import QuestionClassifierNode
@@ -68,6 +68,8 @@ class MockNodeMixin:
             kwargs.setdefault("model_instance", MagicMock(spec=ModelInstance))
             # LLM-like nodes now require an http_client; provide a mock by default for tests.
             kwargs.setdefault("http_client", MagicMock(spec=HttpClientProtocol))
+            if isinstance(self, (LLMNode, QuestionClassifierNode)):
+                kwargs.setdefault("template_renderer", MagicMock(spec=TemplateRenderer))
 
         # Ensure TemplateTransformNode receives a renderer now required by constructor
         if isinstance(self, TemplateTransformNode):

api/tests/unit_tests/core/workflow/graph_engine/test_response_session.py

@@ -4,9 +4,7 @@ from __future__ import annotations
 
 import pytest
 
-import dify_graph.graph_engine.response_coordinator.session as response_session_module
 from dify_graph.enums import BuiltinNodeTypes, NodeExecutionType, NodeState, NodeType
-from dify_graph.graph_engine.response_coordinator import RESPONSE_SESSION_NODE_TYPES
 from dify_graph.graph_engine.response_coordinator.session import ResponseSession
 from dify_graph.nodes.base.template import Template, TextSegment
 
@@ -35,28 +33,14 @@ class DummyNodeWithoutStreamingTemplate:
         self.state = NodeState.UNKNOWN
 
 
-def test_response_session_from_node_rejects_node_types_outside_allowlist() -> None:
-    """Unsupported node types are rejected even if they expose a template."""
+def test_response_session_from_node_accepts_nodes_outside_previous_allowlist() -> None:
+    """Session creation depends on the streaming-template contract rather than node type."""
     node = DummyResponseNode(
         node_id="llm-node",
         node_type=BuiltinNodeTypes.LLM,
         template=Template(segments=[TextSegment(text="hello")]),
     )
 
-    with pytest.raises(TypeError, match="RESPONSE_SESSION_NODE_TYPES"):
-        ResponseSession.from_node(node)
-
-
-def test_response_session_from_node_supports_downstream_allowlist_extension(monkeypatch) -> None:
-    """Downstream applications can extend the supported node-type list."""
-    node = DummyResponseNode(
-        node_id="llm-node",
-        node_type=BuiltinNodeTypes.LLM,
-        template=Template(segments=[TextSegment(text="hello")]),
-    )
-    extended_node_types = [*RESPONSE_SESSION_NODE_TYPES, BuiltinNodeTypes.LLM]
-    monkeypatch.setattr(response_session_module, "RESPONSE_SESSION_NODE_TYPES", extended_node_types)
-
     session = ResponseSession.from_node(node)
 
     assert session.node_id == "llm-node"

api/tests/unit_tests/core/workflow/graph_engine/test_worker.py

@@ -0,0 +1,145 @@
+import queue
+from collections.abc import Generator
+from datetime import UTC, datetime, timedelta
+from types import SimpleNamespace
+from unittest.mock import MagicMock, patch
+
+from dify_graph.enums import BuiltinNodeTypes, WorkflowNodeExecutionStatus
+from dify_graph.graph_engine.ready_queue import InMemoryReadyQueue
+from dify_graph.graph_engine.worker import Worker
+from dify_graph.graph_events import NodeRunFailedEvent, NodeRunStartedEvent
+
+
+def test_build_fallback_failure_event_uses_naive_utc_and_failed_node_run_result(mocker) -> None:
+    fixed_time = datetime(2024, 1, 1, 12, 0, 0, tzinfo=UTC).replace(tzinfo=None)
+    mocker.patch("dify_graph.graph_engine.worker.naive_utc_now", return_value=fixed_time)
+
+    worker = Worker(
+        ready_queue=InMemoryReadyQueue(),
+        event_queue=queue.Queue(),
+        graph=MagicMock(),
+        layers=[],
+    )
+    node = SimpleNamespace(
+        execution_id="exec-1",
+        id="node-1",
+        node_type=BuiltinNodeTypes.LLM,
+    )
+
+    event = worker._build_fallback_failure_event(node, RuntimeError("boom"))
+
+    assert event.start_at == fixed_time
+    assert event.finished_at == fixed_time
+    assert event.error == "boom"
+    assert event.node_run_result.status == WorkflowNodeExecutionStatus.FAILED
+    assert event.node_run_result.error == "boom"
+    assert event.node_run_result.error_type == "RuntimeError"
+
+
+def test_worker_fallback_failure_event_reuses_observed_start_time() -> None:
+    start_at = datetime(2024, 1, 1, 12, 0, 0, tzinfo=UTC).replace(tzinfo=None)
+    failure_time = start_at + timedelta(seconds=5)
+    captured_events: list[NodeRunFailedEvent | NodeRunStartedEvent] = []
+
+    class FakeNode:
+        execution_id = "exec-1"
+        id = "node-1"
+        node_type = BuiltinNodeTypes.LLM
+
+        def ensure_execution_id(self) -> str:
+            return self.execution_id
+
+        def run(self) -> Generator[NodeRunStartedEvent, None, None]:
+            yield NodeRunStartedEvent(
+                id=self.execution_id,
+                node_id=self.id,
+                node_type=self.node_type,
+                node_title="LLM",
+                start_at=start_at,
+            )
+
+    worker = Worker(
+        ready_queue=MagicMock(),
+        event_queue=MagicMock(),
+        graph=MagicMock(nodes={"node-1": FakeNode()}),
+        layers=[],
+    )
+
+    worker._ready_queue.get.side_effect = ["node-1"]
+
+    def put_side_effect(event: NodeRunFailedEvent | NodeRunStartedEvent) -> None:
+        captured_events.append(event)
+        if len(captured_events) == 1:
+            raise RuntimeError("queue boom")
+        worker.stop()
+
+    worker._event_queue.put.side_effect = put_side_effect
+
+    with patch("dify_graph.graph_engine.worker.naive_utc_now", return_value=failure_time):
+        worker.run()
+
+    fallback_event = captured_events[-1]
+
+    assert isinstance(fallback_event, NodeRunFailedEvent)
+    assert fallback_event.start_at == start_at
+    assert fallback_event.finished_at == failure_time
+    assert fallback_event.error == "queue boom"
+    assert fallback_event.node_run_result.status == WorkflowNodeExecutionStatus.FAILED
+
+
+def test_worker_fallback_failure_event_ignores_nested_iteration_child_start_times() -> None:
+    parent_start = datetime(2024, 1, 1, 12, 0, 0, tzinfo=UTC).replace(tzinfo=None)
+    child_start = parent_start + timedelta(seconds=3)
+    failure_time = parent_start + timedelta(seconds=5)
+    captured_events: list[NodeRunFailedEvent | NodeRunStartedEvent] = []
+
+    class FakeIterationNode:
+        execution_id = "iteration-exec"
+        id = "iteration-node"
+        node_type = BuiltinNodeTypes.ITERATION
+
+        def ensure_execution_id(self) -> str:
+            return self.execution_id
+
+        def run(self) -> Generator[NodeRunStartedEvent, None, None]:
+            yield NodeRunStartedEvent(
+                id=self.execution_id,
+                node_id=self.id,
+                node_type=self.node_type,
+                node_title="Iteration",
+                start_at=parent_start,
+            )
+            yield NodeRunStartedEvent(
+                id="child-exec",
+                node_id="child-node",
+                node_type=BuiltinNodeTypes.LLM,
+                node_title="LLM",
+                start_at=child_start,
+                in_iteration_id=self.id,
+            )
+
+    worker = Worker(
+        ready_queue=MagicMock(),
+        event_queue=MagicMock(),
+        graph=MagicMock(nodes={"iteration-node": FakeIterationNode()}),
+        layers=[],
+    )
+
+    worker._ready_queue.get.side_effect = ["iteration-node"]
+
+    def put_side_effect(event: NodeRunFailedEvent | NodeRunStartedEvent) -> None:
+        captured_events.append(event)
+        if len(captured_events) == 2:
+            raise RuntimeError("queue boom")
+        worker.stop()
+
+    worker._event_queue.put.side_effect = put_side_effect
+
+    with patch("dify_graph.graph_engine.worker.naive_utc_now", return_value=failure_time):
+        worker.run()
+
+    fallback_event = captured_events[-1]
+
+    assert isinstance(fallback_event, NodeRunFailedEvent)
+    assert fallback_event.start_at == parent_start
+    assert fallback_event.finished_at == failure_time

api/tests/unit_tests/core/workflow/nodes/human_input/test_email_delivery_config.py

@@ -14,3 +14,64 @@ def test_render_body_template_replaces_variable_values():
     result = config.render_body_template(body=config.body, url="https://example.com", variable_pool=variable_pool)
 
     assert result == "Hello World https://example.com"
+
+
+def test_render_markdown_body_renders_markdown_to_html():
+    rendered = EmailDeliveryConfig.render_markdown_body("**Bold** and [link](https://example.com)")
+
+    assert "<strong>Bold</strong>" in rendered
+    assert '<a href="https://example.com">link</a>' in rendered
+
+
+def test_render_markdown_body_sanitizes_unsafe_html():
+    rendered = EmailDeliveryConfig.render_markdown_body(
+        '<script>alert("xss")</script><a href="javascript:alert(1)" onclick="alert(2)">Click</a>'
+    )
+
+    assert "<script" not in rendered
+    assert "<a" not in rendered
+    assert "onclick" not in rendered
+    assert "javascript:" not in rendered
+    assert "Click" in rendered
+
+
+def test_render_markdown_body_sanitizes_markdown_link_with_javascript_href():
+    rendered = EmailDeliveryConfig.render_markdown_body("[bad](javascript:alert(1)) and [ok](https://example.com)")
+
+    assert "javascript:" not in rendered
+    assert "<a>bad</a>" in rendered
+    assert '<a href="https://example.com">ok</a>' in rendered
+
+
+def test_render_markdown_body_does_not_allow_raw_html_tags():
+    rendered = EmailDeliveryConfig.render_markdown_body("<b>raw html</b> and **markdown**")
+
+    assert "<b>" not in rendered
+    assert "raw html" in rendered
+    assert "<strong>markdown</strong>" in rendered
+
+
+def test_render_markdown_body_supports_table_syntax():
+    rendered = EmailDeliveryConfig.render_markdown_body("| h1 | h2 |\n| --- | ---: |\n| v1 | v2 |")
+
+    assert "<table>" in rendered
+    assert "<thead>" in rendered
+    assert "<tbody>" in rendered
+    assert 'align="right"' in rendered
+    assert "style=" not in rendered
+
+
+def test_sanitize_subject_removes_crlf():
+    sanitized = EmailDeliveryConfig.sanitize_subject("Notice\r\nBCC:attacker@example.com")
+
+    assert "\r" not in sanitized
+    assert "\n" not in sanitized
+    assert sanitized == "Notice BCC:attacker@example.com"
+
+
+def test_sanitize_subject_removes_html_tags():
+    sanitized = EmailDeliveryConfig.sanitize_subject("<b>Alert</b><img src=x onerror=1>")
+
+    assert "<" not in sanitized
+    assert ">" not in sanitized
+    assert sanitized == "Alert"

api/tests/unit_tests/core/workflow/nodes/iteration/test_parallel_iteration_duration.py

@@ -0,0 +1,63 @@
+import time
+from contextlib import nullcontext
+from datetime import UTC, datetime
+
+import pytest
+
+from dify_graph.enums import BuiltinNodeTypes
+from dify_graph.graph_events import NodeRunSucceededEvent
+from dify_graph.model_runtime.entities.llm_entities import LLMUsage
+from dify_graph.nodes.iteration.entities import ErrorHandleMode, IterationNodeData
+from dify_graph.nodes.iteration.iteration_node import IterationNode
+
+
+def test_parallel_iteration_duration_map_uses_worker_measured_time() -> None:
+    node = IterationNode.__new__(IterationNode)
+    node._node_data = IterationNodeData(
+        title="Parallel Iteration",
+        iterator_selector=["start", "items"],
+        output_selector=["iteration", "output"],
+        is_parallel=True,
+        parallel_nums=2,
+        error_handle_mode=ErrorHandleMode.TERMINATED,
+    )
+    node._capture_execution_context = lambda: nullcontext()
+    node._sync_conversation_variables_from_snapshot = lambda snapshot: None
+    node._merge_usage = lambda current, new: new if current.total_tokens == 0 else current.plus(new)
+
+    def fake_execute_single_iteration_parallel(*, index: int, item: object, execution_context: object):
+        return (
+            0.1 + (index * 0.1),
+            [
+                NodeRunSucceededEvent(
+                    id=f"exec-{index}",
+                    node_id=f"llm-{index}",
+                    node_type=BuiltinNodeTypes.LLM,
+                    start_at=datetime.now(UTC).replace(tzinfo=None),
+                ),
+            ],
+            f"output-{item}",
+            {},
+            LLMUsage.empty_usage(),
+        )
+
+    node._execute_single_iteration_parallel = fake_execute_single_iteration_parallel
+
+    outputs: list[object] = []
+    iter_run_map: dict[str, float] = {}
+    usage_accumulator = [LLMUsage.empty_usage()]
+
+    generator = node._execute_parallel_iterations(
+        iterator_list_value=["a", "b"],
+        outputs=outputs,
+        iter_run_map=iter_run_map,
+        usage_accumulator=usage_accumulator,
+    )
+
+    for _ in generator:
+        # Simulate a slow consumer replaying buffered events.
+        time.sleep(0.02)
+
+    assert outputs == ["output-a", "output-b"]
+    assert iter_run_map["0"] == pytest.approx(0.1)
+    assert iter_run_map["1"] == pytest.approx(0.2)

api/tests/unit_tests/core/workflow/nodes/llm/test_llm_utils.py

@@ -1,18 +1,26 @@
-"""Tests for llm_utils module, specifically multimodal content handling."""
+"""Tests for llm_utils module, specifically multimodal content handling and prompt message construction."""
 
 import string
+from unittest import mock
 from unittest.mock import patch
 
+import pytest
+
+from core.model_manager import ModelInstance
+from dify_graph.model_runtime.entities import ImagePromptMessageContent, PromptMessageRole, TextPromptMessageContent
 from dify_graph.model_runtime.entities.message_entities import (
-    ImagePromptMessageContent,
-    TextPromptMessageContent,
+    SystemPromptMessage,
     UserPromptMessage,
 )
+from dify_graph.nodes.llm import llm_utils
+from dify_graph.nodes.llm.entities import LLMNodeChatModelMessage
+from dify_graph.nodes.llm.exc import NoPromptFoundError
 from dify_graph.nodes.llm.llm_utils import (
     _truncate_multimodal_content,
     build_context,
     restore_multimodal_content_in_messages,
 )
+from dify_graph.runtime import VariablePool
 
 
 class TestTruncateMultimodalContent:
@@ -50,7 +58,6 @@ class TestTruncateMultimodalContent:
         assert isinstance(result_content, ImagePromptMessageContent)
         assert result_content.base64_data == ""
         assert result_content.url == ""
-        # file_ref should be preserved
         assert result_content.file_ref == "local:test-file-id"
 
     def test_truncates_base64_when_no_file_ref(self):
@@ -70,7 +77,6 @@ class TestTruncateMultimodalContent:
         assert isinstance(result.content, list)
         result_content = result.content[0]
         assert isinstance(result_content, ImagePromptMessageContent)
-        # Should be truncated with marker
         assert "...[TRUNCATED]..." in result_content.base64_data
         assert len(result_content.base64_data) < len(long_base64)
 
@@ -89,9 +95,7 @@ class TestTruncateMultimodalContent:
 
         assert isinstance(result.content, list)
         assert len(result.content) == 2
-        # Text content unchanged
         assert result.content[0].data == "Hello!"
-        # Image content base64 cleared
         assert result.content[1].base64_data == ""
 
 
@@ -100,8 +104,6 @@ class TestBuildContext:
 
     def test_excludes_system_messages(self):
         """System messages should be excluded from context."""
-        from dify_graph.model_runtime.entities.message_entities import SystemPromptMessage
-
         messages = [
             SystemPromptMessage(content="You are a helpful assistant."),
             UserPromptMessage(content="Hello!"),
@@ -109,7 +111,6 @@ class TestBuildContext:
 
         context = build_context(messages, "Hi there!")
 
-        # Should have user message + assistant response, no system message
         assert len(context) == 2
         assert context[0].content == "Hello!"
         assert context[1].content == "Hi there!"
@@ -140,7 +141,6 @@ class TestBuildContext:
 
         messages = [UserPromptMessage(content="What's the weather in Beijing?")]
 
-        # Create trace with tool call and result
         generation_data = LLMGenerationData(
             text="The weather in Beijing is sunny, 25°C.",
             reasoning_contents=[],
@@ -183,7 +183,6 @@ class TestBuildContext:
         accumulated_response = "Let me check the weather.The weather in Beijing is sunny, 25°C."
         context = build_context(messages, accumulated_response, generation_data)
 
-        # Should have: user message + assistant with tool_call + tool result + final assistant
         assert len(context) == 4
         assert context[0].content == "What's the weather in Beijing?"
         assert isinstance(context[1], AssistantPromptMessage)
@@ -223,7 +222,6 @@ class TestBuildContext:
             finish_reason="stop",
             files=[],
             trace=[
-                # First model call with two tool calls
                 LLMTraceSegment(
                     type="model",
                     duration=0.5,
@@ -237,7 +235,6 @@ class TestBuildContext:
                         ],
                     ),
                 ),
-                # First tool result
                 LLMTraceSegment(
                     type="tool",
                     duration=0.2,
@@ -249,7 +246,6 @@ class TestBuildContext:
                         output="Sunny, 25°C",
                     ),
                 ),
-                # Second tool result
                 LLMTraceSegment(
                     type="tool",
                     duration=0.2,
@@ -267,7 +263,6 @@ class TestBuildContext:
         accumulated_response = "I'll check both cities.Beijing is sunny at 25°C, Shanghai is cloudy at 22°C."
         context = build_context(messages, accumulated_response, generation_data)
 
-        # Should have: user + assistant with 2 tool_calls + 2 tool results + final assistant
         assert len(context) == 5
         assert context[0].content == "Compare weather in Beijing and Shanghai"
         assert isinstance(context[1], AssistantPromptMessage)
@@ -304,12 +299,11 @@ class TestBuildContext:
             usage=LLMUsage.empty_usage(),
             finish_reason="stop",
             files=[],
-            trace=[],  # Empty trace
+            trace=[],
         )
 
         context = build_context(messages, "Hi there!", generation_data)
 
-        # Should fallback to simple context
         assert len(context) == 2
         assert context[0].content == "Hello!"
         assert context[1].content == "Hi there!"
@@ -321,7 +315,6 @@ class TestRestoreMultimodalContentInMessages:
     @patch("dify_graph.file.file_manager.restore_multimodal_content")
     def test_restores_multimodal_content(self, mock_restore):
         """Should restore multimodal content in messages."""
-        # Setup mock
         restored_content = ImagePromptMessageContent(
             format="png",
             base64_data="restored-base64",
@@ -330,7 +323,6 @@ class TestRestoreMultimodalContentInMessages:
         )
         mock_restore.return_value = restored_content
 
-        # Create message with truncated content
         truncated_content = ImagePromptMessageContent(
             format="png",
             base64_data="",
@@ -363,3 +355,98 @@ class TestRestoreMultimodalContentInMessages:
 
         assert len(result) == 1
         assert result[0].content[0].data == "Hello!"
+
+
+def _fetch_prompt_messages_with_mocked_content(content):
+    variable_pool = VariablePool.empty()
+    model_instance = mock.MagicMock(spec=ModelInstance)
+    prompt_template = [
+        LLMNodeChatModelMessage(
+            text="You are a classifier.",
+            role=PromptMessageRole.SYSTEM,
+            edition_type="basic",
+        )
+    ]
+
+    with (
+        mock.patch(
+            "dify_graph.nodes.llm.llm_utils.fetch_model_schema",
+            return_value=mock.MagicMock(features=[]),
+        ),
+        mock.patch(
+            "dify_graph.nodes.llm.llm_utils.handle_list_messages",
+            return_value=[SystemPromptMessage(content=content)],
+        ),
+        mock.patch(
+            "dify_graph.nodes.llm.llm_utils.handle_memory_chat_mode",
+            return_value=[],
+        ),
+    ):
+        return llm_utils.fetch_prompt_messages(
+            sys_query=None,
+            sys_files=[],
+            context=None,
+            memory=None,
+            model_instance=model_instance,
+            prompt_template=prompt_template,
+            stop=["END"],
+            memory_config=None,
+            vision_enabled=False,
+            vision_detail=ImagePromptMessageContent.DETAIL.HIGH,
+            variable_pool=variable_pool,
+            jinja2_variables=[],
+            template_renderer=None,
+        )
+
+
+def test_fetch_prompt_messages_skips_messages_when_all_contents_are_filtered_out():
+    with pytest.raises(NoPromptFoundError):
+        _fetch_prompt_messages_with_mocked_content(
+            [
+                ImagePromptMessageContent(
+                    format="url",
+                    url="https://example.com/image.png",
+                    mime_type="image/png",
+                ),
+            ]
+        )
+
+
+def test_fetch_prompt_messages_flattens_single_text_content_after_filtering_unsupported_multimodal_items():
+    prompt_messages, stop = _fetch_prompt_messages_with_mocked_content(
+        [
+            TextPromptMessageContent(data="You are a classifier."),
+            ImagePromptMessageContent(
+                format="url",
+                url="https://example.com/image.png",
+                mime_type="image/png",
+            ),
+        ]
+    )
+
+    assert stop == ["END"]
+    assert prompt_messages == [SystemPromptMessage(content="You are a classifier.")]
+
+
+def test_fetch_prompt_messages_keeps_list_content_when_multiple_supported_items_remain():
+    prompt_messages, stop = _fetch_prompt_messages_with_mocked_content(
+        [
+            TextPromptMessageContent(data="You are"),
+            TextPromptMessageContent(data=" a classifier."),
+            ImagePromptMessageContent(
+                format="url",
+                url="https://example.com/image.png",
+                mime_type="image/png",
+            ),
+        ]
+    )
+
+    assert stop == ["END"]
+    assert prompt_messages == [
+        SystemPromptMessage(
+            content=[
+                TextPromptMessageContent(data="You are"),
+                TextPromptMessageContent(data=" a classifier."),
+            ]
+        )
+    ]

api/tests/unit_tests/core/workflow/nodes/llm/test_node.py

@@ -34,8 +34,8 @@ from dify_graph.nodes.llm.entities import (
     VisionConfigOptions,
 )
 from dify_graph.nodes.llm.file_saver import LLMFileSaver
-from dify_graph.nodes.llm.node import LLMNode, _handle_memory_completion_mode
-from dify_graph.nodes.llm.protocols import CredentialsProvider, ModelFactory
+from dify_graph.nodes.llm.node import LLMNode
+from dify_graph.nodes.llm.protocols import CredentialsProvider, ModelFactory, TemplateRenderer
 from dify_graph.runtime import GraphRuntimeState, VariablePool
 from dify_graph.system_variable import SystemVariable
 from dify_graph.variables import ArrayAnySegment, ArrayFileSegment, NoneSegment
@@ -107,6 +107,7 @@ def llm_node(
     mock_file_saver = mock.MagicMock(spec=LLMFileSaver)
     mock_credentials_provider = mock.MagicMock(spec=CredentialsProvider)
     mock_model_factory = mock.MagicMock(spec=ModelFactory)
+    mock_template_renderer = mock.MagicMock(spec=TemplateRenderer)
     node_config = {
         "id": "1",
         "data": llm_node_data.model_dump(),
@@ -121,6 +122,7 @@ def llm_node(
         model_factory=mock_model_factory,
         model_instance=mock.MagicMock(spec=ModelInstance),
         llm_file_saver=mock_file_saver,
+        template_renderer=mock_template_renderer,
         http_client=http_client,
     )
     return node
@@ -590,6 +592,33 @@ def test_handle_list_messages_basic(llm_node):
     assert result[0].content == [TextPromptMessageContent(data="Hello, world")]
 
 
+def test_handle_list_messages_jinja2_uses_template_renderer(llm_node):
+    llm_node._template_renderer.render_jinja2.return_value = "Hello, world"
+    messages = [
+        LLMNodeChatModelMessage(
+            text="",
+            jinja2_text="Hello, {{ name }}",
+            role=PromptMessageRole.USER,
+            edition_type="jinja2",
+        )
+    ]
+
+    result = llm_node.handle_list_messages(
+        messages=messages,
+        context=None,
+        jinja2_variables=[],
+        variable_pool=llm_node.graph_runtime_state.variable_pool,
+        vision_detail_config=ImagePromptMessageContent.DETAIL.HIGH,
+        template_renderer=llm_node._template_renderer,
+    )
+
+    assert result == [UserPromptMessage(content=[TextPromptMessageContent(data="Hello, world")])]
+    llm_node._template_renderer.render_jinja2.assert_called_once_with(
+        template="Hello, {{ name }}",
+        inputs={},
+    )
+
+
 def test_handle_memory_completion_mode_uses_prompt_message_interface():
     memory = mock.MagicMock(spec=MockTokenBufferMemory)
     memory.get_history_prompt_messages.return_value = [
@@ -613,8 +642,8 @@ def test_handle_memory_completion_mode_uses_prompt_message_interface():
         window=MemoryConfig.WindowConfig(enabled=True, size=3),
     )
 
-    with mock.patch("dify_graph.nodes.llm.node._calculate_rest_token", return_value=2000) as mock_rest_token:
-        memory_text = _handle_memory_completion_mode(
+    with mock.patch("dify_graph.nodes.llm.llm_utils.calculate_rest_token", return_value=2000) as mock_rest_token:
+        memory_text = llm_utils.handle_memory_completion_mode(
             memory=memory,
             memory_config=memory_config,
             model_instance=model_instance,
@@ -630,6 +659,7 @@ def llm_node_for_multimodal(llm_node_data, graph_init_params, graph_runtime_stat
     mock_file_saver: LLMFileSaver = mock.MagicMock(spec=LLMFileSaver)
     mock_credentials_provider = mock.MagicMock(spec=CredentialsProvider)
     mock_model_factory = mock.MagicMock(spec=ModelFactory)
+    mock_template_renderer = mock.MagicMock(spec=TemplateRenderer)
     node_config = {
         "id": "1",
         "data": llm_node_data.model_dump(),
@@ -644,6 +674,7 @@ def llm_node_for_multimodal(llm_node_data, graph_init_params, graph_runtime_stat
         model_factory=mock_model_factory,
         model_instance=mock.MagicMock(spec=ModelInstance),
         llm_file_saver=mock_file_saver,
+        template_renderer=mock_template_renderer,
         http_client=http_client,
     )
     return node, mock_file_saver

api/tests/unit_tests/core/workflow/nodes/test_question_classifier_node.py

@@ -1,5 +1,14 @@
+from types import SimpleNamespace
+from unittest.mock import MagicMock
+
 from dify_graph.model_runtime.entities import ImagePromptMessageContent
-from dify_graph.nodes.question_classifier import QuestionClassifierNodeData
+from dify_graph.nodes.llm.protocols import CredentialsProvider, ModelFactory, TemplateRenderer
+from dify_graph.nodes.protocols import HttpClientProtocol
+from dify_graph.nodes.question_classifier import (
+    QuestionClassifierNode,
+    QuestionClassifierNodeData,
+)
+from tests.workflow_test_utils import build_test_graph_init_params
 
 
 def test_init_question_classifier_node_data():
@@ -65,3 +74,52 @@ def test_init_question_classifier_node_data_without_vision_config():
     assert node_data.vision.enabled == False
     assert node_data.vision.configs.variable_selector == ["sys", "files"]
     assert node_data.vision.configs.detail == ImagePromptMessageContent.DETAIL.HIGH
+
+
+def test_question_classifier_calculate_rest_token_uses_shared_prompt_builder(monkeypatch):
+    node_data = QuestionClassifierNodeData.model_validate(
+        {
+            "title": "test classifier node",
+            "query_variable_selector": ["id", "name"],
+            "model": {"provider": "openai", "name": "gpt-3.5-turbo", "mode": "completion", "completion_params": {}},
+            "classes": [{"id": "1", "name": "class 1"}],
+            "instruction": "This is a test instruction",
+        }
+    )
+    template_renderer = MagicMock(spec=TemplateRenderer)
+    node = QuestionClassifierNode(
+        id="node-id",
+        config={"id": "node-id", "data": node_data.model_dump(mode="json")},
+        graph_init_params=build_test_graph_init_params(
+            workflow_id="workflow-id",
+            graph_config={},
+            tenant_id="tenant-id",
+            app_id="app-id",
+            user_id="user-id",
+        ),
+        graph_runtime_state=SimpleNamespace(variable_pool=MagicMock()),
+        credentials_provider=MagicMock(spec=CredentialsProvider),
+        model_factory=MagicMock(spec=ModelFactory),
+        model_instance=MagicMock(),
+        http_client=MagicMock(spec=HttpClientProtocol),
+        llm_file_saver=MagicMock(),
+        template_renderer=template_renderer,
+    )
+    fetch_prompt_messages = MagicMock(return_value=([], None))
+    monkeypatch.setattr(
+        "dify_graph.nodes.question_classifier.question_classifier_node.llm_utils.fetch_prompt_messages",
+        fetch_prompt_messages,
+    )
+    monkeypatch.setattr(
+        "dify_graph.nodes.question_classifier.question_classifier_node.llm_utils.fetch_model_schema",
+        MagicMock(return_value=SimpleNamespace(model_properties={}, parameter_rules=[])),
+    )
+
+    node._calculate_rest_token(
+        node_data=node_data,
+        query="hello",
+        model_instance=MagicMock(stop=(), parameters={}),
+        context="",
+    )
+
+    assert fetch_prompt_messages.call_args.kwargs["template_renderer"] is template_renderer

api/tests/unit_tests/core/workflow/nodes/trigger_plugin/test_trigger_event_node.py

@@ -0,0 +1,63 @@
+from collections.abc import Mapping
+
+from core.trigger.constants import TRIGGER_PLUGIN_NODE_TYPE
+from core.workflow.nodes.trigger_plugin.trigger_event_node import TriggerEventNode
+from dify_graph.entities import GraphInitParams
+from dify_graph.entities.graph_config import NodeConfigDict, NodeConfigDictAdapter
+from dify_graph.enums import WorkflowNodeExecutionMetadataKey, WorkflowNodeExecutionStatus
+from dify_graph.runtime import GraphRuntimeState, VariablePool
+from dify_graph.system_variable import SystemVariable
+from tests.workflow_test_utils import build_test_graph_init_params
+
+
+def _build_context(graph_config: Mapping[str, object]) -> tuple[GraphInitParams, GraphRuntimeState]:
+    init_params = build_test_graph_init_params(
+        graph_config=graph_config,
+        user_from="account",
+        invoke_from="debugger",
+    )
+    runtime_state = GraphRuntimeState(
+        variable_pool=VariablePool(
+            system_variables=SystemVariable(user_id="user", files=[]),
+            user_inputs={"payload": "value"},
+        ),
+        start_at=0.0,
+    )
+    return init_params, runtime_state
+
+
+def _build_node_config() -> NodeConfigDict:
+    return NodeConfigDictAdapter.validate_python(
+        {
+            "id": "node-1",
+            "data": {
+                "type": TRIGGER_PLUGIN_NODE_TYPE,
+                "title": "Trigger Event",
+                "plugin_id": "plugin-id",
+                "provider_id": "provider-id",
+                "event_name": "event-name",
+                "subscription_id": "subscription-id",
+                "plugin_unique_identifier": "plugin-unique-identifier",
+                "event_parameters": {},
+            },
+        }
+    )
+
+
+def test_trigger_event_node_run_populates_trigger_info_metadata() -> None:
+    init_params, runtime_state = _build_context(graph_config={})
+    node = TriggerEventNode(
+        id="node-1",
+        config=_build_node_config(),
+        graph_init_params=init_params,
+        graph_runtime_state=runtime_state,
+    )
+
+    result = node._run()
+
+    assert result.status == WorkflowNodeExecutionStatus.SUCCEEDED
+    assert result.metadata[WorkflowNodeExecutionMetadataKey.TRIGGER_INFO] == {
+        "provider_id": "provider-id",
+        "event_name": "event-name",
+        "plugin_unique_identifier": "plugin-unique-identifier",
+    }

api/tests/unit_tests/core/workflow/test_node_factory.py

@@ -140,6 +140,29 @@ class TestDefaultWorkflowCodeExecutor:
         assert executor.is_execution_error(RuntimeError("boom")) is False
 
 
+class TestDefaultLLMTemplateRenderer:
+    def test_render_jinja2_delegates_to_code_executor(self, monkeypatch):
+        renderer = node_factory.DefaultLLMTemplateRenderer()
+        execute_workflow_code_template = MagicMock(return_value={"result": "hello world"})
+        monkeypatch.setattr(
+            node_factory.CodeExecutor,
+            "execute_workflow_code_template",
+            execute_workflow_code_template,
+        )
+
+        result = renderer.render_jinja2(
+            template="Hello {{ name }}",
+            inputs={"name": "world"},
+        )
+
+        assert result == "hello world"
+        execute_workflow_code_template.assert_called_once_with(
+            language=CodeLanguage.JINJA2,
+            code="Hello {{ name }}",
+            inputs={"name": "world"},
+        )
+
+
 class TestDifyNodeFactoryInit:
     def test_init_builds_default_dependencies(self):
         graph_init_params = SimpleNamespace(run_context={"context": "value"})
@@ -150,6 +173,7 @@ class TestDifyNodeFactoryInit:
         http_request_config = sentinel.http_request_config
         credentials_provider = sentinel.credentials_provider
         model_factory = sentinel.model_factory
+        llm_template_renderer = sentinel.llm_template_renderer
 
         with (
             patch.object(
@@ -172,6 +196,11 @@ class TestDifyNodeFactoryInit:
                 "build_http_request_config",
                 return_value=http_request_config,
             ),
+            patch.object(
+                node_factory,
+                "DefaultLLMTemplateRenderer",
+                return_value=llm_template_renderer,
+            ) as llm_renderer_factory,
             patch.object(
                 node_factory,
                 "build_dify_model_access",
@@ -186,11 +215,14 @@ class TestDifyNodeFactoryInit:
         resolve_dify_context.assert_called_once_with(graph_init_params.run_context)
         build_dify_model_access.assert_called_once_with("tenant-id")
         renderer_factory.assert_called_once()
+        llm_renderer_factory.assert_called_once()
         assert renderer_factory.call_args.kwargs["code_executor"] is factory._code_executor
         assert factory.graph_init_params is graph_init_params
         assert factory.graph_runtime_state is graph_runtime_state
         assert factory._dify_context is dify_context
         assert factory._template_renderer is template_renderer
+
+        assert factory._llm_template_renderer is llm_template_renderer
         assert factory._document_extractor_unstructured_api_config is unstructured_api_config
         assert factory._http_request_config is http_request_config
         assert factory._llm_credentials_provider is credentials_provider
@@ -242,6 +274,7 @@ class TestDifyNodeFactoryCreateNode:
         factory._code_executor = sentinel.code_executor
         factory._code_limits = sentinel.code_limits
         factory._template_renderer = sentinel.template_renderer
+        factory._llm_template_renderer = sentinel.llm_template_renderer
         factory._template_transform_max_output_length = 2048
         factory._http_request_http_client = sentinel.http_client
         factory._http_request_tool_file_manager_factory = sentinel.tool_file_manager_factory
@@ -378,8 +411,22 @@ class TestDifyNodeFactoryCreateNode:
     @pytest.mark.parametrize(
         ("node_type", "constructor_name", "expected_extra_kwargs"),
         [
-            (BuiltinNodeTypes.LLM, "LLMNode", {"http_client": sentinel.http_client}),
-            (BuiltinNodeTypes.QUESTION_CLASSIFIER, "QuestionClassifierNode", {"http_client": sentinel.http_client}),
+            (
+                BuiltinNodeTypes.LLM,
+                "LLMNode",
+                {
+                    "http_client": sentinel.http_client,
+                    "template_renderer": sentinel.llm_template_renderer,
+                },
+            ),
+            (
+                BuiltinNodeTypes.QUESTION_CLASSIFIER,
+                "QuestionClassifierNode",
+                {
+                    "http_client": sentinel.http_client,
+                    "template_renderer": sentinel.llm_template_renderer,
+                },
+            ),
             (BuiltinNodeTypes.PARAMETER_EXTRACTOR, "ParameterExtractorNode", {}),
         ],
     )