Merge remote-tracking branch 'origin/main' into fix/onboarding_store

.github/workflows/codeql.yml

@@ -0,0 +1,66 @@
+name: "CodeQL"
+
+on:
+  push:
+    branches: [ 'main' ]
+  pull_request:
+    # The branches below must be a subset of the branches above
+    branches: [ 'main' ]
+  schedule:
+    - cron: '17 2 * * 1'
+
+jobs:
+  analyze:
+    name: Analyze
+    runs-on: ${{ (matrix.language == 'swift' && 'macos-latest') || 'ubuntu-latest' }}
+    timeout-minutes: ${{ (matrix.language == 'swift' && 120) || 360 }}
+    permissions:
+      actions: read
+      contents: read
+      security-events: write
+
+    strategy:
+      fail-fast: false
+      matrix:
+        language: [ 'python', 'javascript' ]
+        # CodeQL supports [ 'cpp', 'csharp', 'go', 'java', 'javascript', 'python', 'ruby' ]
+        # Use only 'java' to analyze code written in Java, Kotlin or both
+        # Use only 'javascript' to analyze code written in JavaScript, TypeScript or both
+        # Learn more about CodeQL language support at https://aka.ms/codeql-docs/language-support
+
+    steps:
+    - name: Checkout repository
+      uses: actions/checkout@v6
+
+    # Initializes the CodeQL tools for scanning.
+    - name: Initialize CodeQL
+      uses: github/codeql-action/init@v3
+      with:
+        languages: ${{ matrix.language }}
+        # If you wish to specify custom queries, you can do so here or in a config file.
+        # By default, queries listed here will override any specified in a config file.
+        # Prefix the list here with "+" to use these queries and those in the config file.
+
+        # For more details on CodeQL's query packs, refer to: https://docs.github.com/en/code-security/code-scanning/automatically-scanning-your-code-for-vulnerabilities-and-errors/configuring-code-scanning#using-queries-in-ql-packs
+        # queries: security-extended,security-and-quality
+
+    # Autobuild attempts to build any compiled languages (C/C++, C#, Go, Java, or Swift).
+    # If this step fails, then you should remove it and run the build manually (see below)
+    - name: Autobuild
+      uses: github/codeql-action/autobuild@v3
+
+    # ℹ️ Command-line programs to run using the OS shell.
+    # 📚 See https://docs.github.com/en/actions/using-workflows/workflow-syntax-for-github-actions#jobsjob_idstepsrun
+
+    #   If the Autobuild fails above, remove it and uncomment the following three lines.
+    #   modify them (or add more) to build your code if your project, please refer to the EXAMPLE below for guidance.
+
+    # - run: |
+    #     echo "Run, Build Application using script"
+    #     ./location_of_script_within_repo/buildscript.sh
+
+    - name: Perform CodeQL Analysis
+      uses: github/codeql-action/analyze@v3
+      with:
+        category: "/language:${{matrix.language}}"
+

.github/workflows/publish-sdk-python.yml

@@ -2,8 +2,11 @@ name: Publish Python SDK
 
 on:
   push:
-    tags:
+    branches:
-      - 'sdk-py-v*'
+      - main
+    paths:
+      - 'sdks/python/pyproject.toml'
+  workflow_dispatch:
 
 jobs:
   publish:
@@ -22,24 +25,35 @@ jobs:
         with:
           python-version: '3.12'
 
-      - name: Install build tools
+      - name: Install uv
-        run: pip install build twine
+        uses: astral-sh/setup-uv@v4
 
-      - name: Extract version from tag
+      - name: Extract version from pyproject.toml
         id: version
         run: |
-          VERSION=${GITHUB_REF_NAME#sdk-py-v}
+          VERSION=$(grep -Po '(?<=^version = ")[^"]*' pyproject.toml)
           echo "version=$VERSION" >> $GITHUB_OUTPUT
 
-      - name: Update version in pyproject.toml
+      - name: Check if version already published
+        id: check
         run: |
-          sed -i "s/^version = .*/version = \"${{ steps.version.outputs.version }}\"/" pyproject.toml
+          HTTP_STATUS=$(curl -s -o /dev/null -w "%{http_code}" https://pypi.org/pypi/openrag-sdk/${{ steps.version.outputs.version }}/json)
+          if [ "$HTTP_STATUS" = "200" ]; then
+            echo "exists=true" >> $GITHUB_OUTPUT
+          else
+            echo "exists=false" >> $GITHUB_OUTPUT
+          fi
 
       - name: Build package
-        run: python -m build
+        if: steps.check.outputs.exists == 'false'
+        run: uv build
 
       - name: Publish to PyPI
-        run: twine upload dist/*
+        if: steps.check.outputs.exists == 'false'
+        run: uv publish
         env:
-          TWINE_USERNAME: __token__
+          UV_PUBLISH_TOKEN: ${{ secrets.UV_PUBLISH_TOKEN }}
-          TWINE_PASSWORD: ${{ secrets.PYPI_TOKEN }}
+
+      - name: Skip publish (version exists)
+        if: steps.check.outputs.exists == 'true'
+        run: echo "Version ${{ steps.version.outputs.version }} already exists on PyPI, skipping publish"

.github/workflows/publish-sdk-typescript.yml

@@ -2,13 +2,19 @@ name: Publish TypeScript SDK
 
 on:
   push:
-    tags:
+    branches:
-      - 'sdk-ts-v*'
+      - main
+    paths:
+      - 'sdks/typescript/package.json'
+  workflow_dispatch:
 
 jobs:
   publish:
     name: Publish to npm
     runs-on: ubuntu-latest
+    permissions:
+      contents: read
+      id-token: write
     defaults:
       run:
         working-directory: sdks/typescript
@@ -23,22 +29,36 @@ jobs:
           node-version: '20'
           registry-url: 'https://registry.npmjs.org'
 
+      - name: Update npm to latest
+        run: npm install -g npm@latest
+
+      - name: Extract version from package.json
+        id: version
+        run: |
+          VERSION=$(node -p "require('./package.json').version")
+          echo "version=$VERSION" >> $GITHUB_OUTPUT
+
+      - name: Check if version already published
+        id: check
+        run: |
+          if npm view openrag-sdk@${{ steps.version.outputs.version }} version 2>/dev/null; then
+            echo "exists=true" >> $GITHUB_OUTPUT
+          else
+            echo "exists=false" >> $GITHUB_OUTPUT
+          fi
+
       - name: Install dependencies
+        if: steps.check.outputs.exists == 'false'
         run: npm ci
 
       - name: Build
+        if: steps.check.outputs.exists == 'false'
         run: npm run build
 
-      - name: Extract version from tag
-        id: version
-        run: |
-          VERSION=${GITHUB_REF_NAME#sdk-ts-v}
-          echo "version=$VERSION" >> $GITHUB_OUTPUT
-
-      - name: Update package version
-        run: npm version ${{ steps.version.outputs.version }} --no-git-tag-version
-
       - name: Publish to npm
-        run: npm publish --access public
+        if: steps.check.outputs.exists == 'false'
-        env:
+        run: npm publish --access public --provenance
-          NODE_AUTH_TOKEN: ${{ secrets.NPM_TOKEN }}
+
+      - name: Skip publish (version exists)
+        if: steps.check.outputs.exists == 'true'
+        run: echo "Version ${{ steps.version.outputs.version }} already exists on npm, skipping publish"

Makefile

@@ -53,9 +53,12 @@ help:
 	@echo ""
 
 # Development environments
+# Use centralized env file from TUI if it exists, otherwise fall back to local .env
+OPENRAG_ENV_FILE := $(shell if [ -f ~/.openrag/tui/.env ]; then echo "--env-file ~/.openrag/tui/.env"; fi)
+
 dev:
 	@echo "🚀 Starting OpenRAG with GPU support..."
-	docker compose -f docker-compose.yml -f docker-compose.gpu.yml up -d
+	docker compose $(OPENRAG_ENV_FILE) -f docker-compose.yml -f docker-compose.gpu.yml up -d
 	@echo "✅ Services started!"
 	@echo "   Backend: http://localhost:8000"
 	@echo "   Frontend: http://localhost:3000"
@@ -65,7 +68,7 @@ dev:
 
 dev-cpu:
 	@echo "🚀 Starting OpenRAG with CPU only..."
-	docker compose up -d
+	docker compose $(OPENRAG_ENV_FILE) up -d
 	@echo "✅ Services started!"
 	@echo "   Backend: http://localhost:8000"
 	@echo "   Frontend: http://localhost:3000"
@@ -75,7 +78,7 @@ dev-cpu:
 
 dev-local:
 	@echo "🔧 Starting infrastructure only (for local development)..."
-	docker compose up -d opensearch dashboards langflow
+	docker compose $(OPENRAG_ENV_FILE) up -d opensearch dashboards langflow
 	@echo "✅ Infrastructure started!"
 	@echo "   Langflow: http://localhost:7860"
 	@echo "   OpenSearch: http://localhost:9200"
@@ -85,7 +88,7 @@ dev-local:
 
 infra:
 	@echo "🔧 Starting infrastructure services only..."
-	docker compose up -d opensearch dashboards langflow
+	docker compose $(OPENRAG_ENV_FILE) up -d opensearch dashboards langflow
 	@echo "✅ Infrastructure services started!"
 	@echo "   Langflow: http://localhost:7860"
 	@echo "   OpenSearch: http://localhost:9200"
@@ -93,7 +96,7 @@ infra:
 
 infra-cpu:
 	@echo "🔧 Starting infrastructure services only..."
-	docker compose up -d opensearch dashboards langflow
+	docker compose $(OPENRAG_ENV_FILE) up -d opensearch dashboards langflow
 	@echo "✅ Infrastructure services started!"
 	@echo "   Langflow: http://localhost:7860"
 	@echo "   OpenSearch: http://localhost:9200"
@@ -102,13 +105,13 @@ infra-cpu:
 # Container management
 stop:
 	@echo "🛑 Stopping all containers..."
-	docker compose down
+	docker compose $(OPENRAG_ENV_FILE) down
 
 restart: stop dev
 
 clean: stop
 	@echo "🧹 Cleaning up containers and volumes..."
-	docker compose down -v --remove-orphans
+	docker compose $(OPENRAG_ENV_FILE) down -v --remove-orphans
 	docker system prune -f
 
 # Local development
@@ -153,36 +156,36 @@ build-fe:
 # Logging and debugging
 logs:
 	@echo "📋 Showing all container logs..."
-	docker compose logs -f
+	docker compose $(OPENRAG_ENV_FILE) logs -f
 
 logs-be:
 	@echo "📋 Showing backend logs..."
-	docker compose logs -f openrag-backend
+	docker compose $(OPENRAG_ENV_FILE) logs -f openrag-backend
 
 logs-fe:
 	@echo "📋 Showing frontend logs..."
-	docker compose logs -f openrag-frontend
+	docker compose $(OPENRAG_ENV_FILE) logs -f openrag-frontend
 
 logs-lf:
 	@echo "📋 Showing langflow logs..."
-	docker compose logs -f langflow
+	docker compose $(OPENRAG_ENV_FILE) logs -f langflow
 
 logs-os:
 	@echo "📋 Showing opensearch logs..."
-	docker compose logs -f opensearch
+	docker compose $(OPENRAG_ENV_FILE) logs -f opensearch
 
 # Shell access
 shell-be:
 	@echo "🐚 Opening shell in backend container..."
-	docker compose exec openrag-backend /bin/bash
+	docker compose $(OPENRAG_ENV_FILE) exec openrag-backend /bin/bash
 
 shell-lf:
 	@echo "🐚 Opening shell in langflow container..."
-	docker compose exec langflow /bin/bash
+	docker compose $(OPENRAG_ENV_FILE) exec langflow /bin/bash
 
 shell-os:
 	@echo "🐚 Opening shell in opensearch container..."
-	docker compose exec opensearch /bin/bash
+	docker compose $(OPENRAG_ENV_FILE) exec opensearch /bin/bash
 
 # Testing and quality
 test:
@@ -414,7 +417,7 @@ lint:
 # Service status
 status:
 	@echo "📊 Container status:"
-	@docker compose ps 2>/dev/null || echo "No containers running"
+	@docker compose $(OPENRAG_ENV_FILE) ps 2>/dev/null || echo "No containers running"
 
 health:
 	@echo "🏥 Health check:"

docs/docs/_partial-export-flows.mdx

@@ -1 +1 @@
-1. If you modified the built-in flows or created custom flows in your [OpenRAG Langflow instance](/agents), and you want to preserve those changes, then you must [export your flows](https://docs.langflow.org/concepts-flows-import) before starting this process. Afterwards, you can import your flows or reference the exported flow JSON as needed.
+1. If you modified the built-in flows or created custom flows in your [OpenRAG Langflow instance](/agents), [export your flows](https://docs.langflow.org/concepts-flows-import) before starting this process. Although OpenRAG can preserve changes to the built-in flows, it doesn't preserve user-created flows. As a general best practice, exporting your flows is recommended to create backups of your customizations. Afterwards, you can reimport your flows or reference the exported flow JSON as needed.

docs/docs/core-components/ingestion.mdx

@@ -157,13 +157,18 @@ You can [monitor ingestion](#monitor-ingestion) to see the progress of the uploa
 
 ## Ingest knowledge from URLs {#url-flow}
 
+When using the OpenRAG chat, you can enter URLs into the chat to be ingested in real-time during your conversation.
+
+:::tip
+Use [UTF-8 encoding](https://www.w3schools.com/tags/ref_urlencode.ASP) for URLs with special characters other than the standard slash, period, and colon characters.
+For example, use `https://en.wikipedia.org/wiki/Caf%C3%A9` instead of `https://en.wikipedia.org/wiki/Café` or `https://en.wikipedia.org/wiki/Coffee%5Fculture` instead of  `https://en.wikipedia.org/wiki/Coffee_culture`.
+:::
+
 The **OpenSearch URL Ingestion** flow is used to ingest web content from URLs.
 This flow isn't directly accessible from the OpenRAG user interface.
 Instead, this flow is called by the [**OpenRAG OpenSearch Agent** flow](/chat#flow) as a Model Context Protocol (MCP) tool.
 The agent can call this component to fetch web content from a given URL, and then ingest that content into your OpenSearch knowledge base.
-
 Like all OpenRAG flows, you can [inspect the flow in Langflow](/agents#inspect-and-modify-flows), and you can customize it.
-
 For more information about MCP in Langflow, see the Langflow documentation on [MCP clients](https://docs.langflow.org/mcp-client) and [MCP servers](https://docs.langflow.org/mcp-tutorial).
 
 ## Monitor ingestion

docs/docs/get-started/manage-services.mdx

@@ -139,6 +139,8 @@ To reset your OpenRAG deployment _and_ delete all OpenRAG data, see [Reinstall O
 
 3. Repeat the [setup process](/install#setup) to restart the services and launch the OpenRAG app. Your OpenRAG passwords, OAuth credentials (if previously set), and onboarding configuration are restored from the `.env` file.
 
+4. If you exported customized flows, [import your flows](https://docs.langflow.org/concepts-flows-import) into Langflow after completing the onboarding process.
+
 </TabItem>
 <TabItem value="env" label="Self-managed services">
 

frontend/components/knowledge-filter-list.tsx

@@ -53,7 +53,20 @@ export function KnowledgeFilterList({
 	};
 
 	const parseQueryData = (queryData: string): ParsedQueryData => {
-		return JSON.parse(queryData) as ParsedQueryData;
+		const parsed = JSON.parse(queryData);
+		// Provide defaults for missing fields to handle API-created filters
+		return {
+			query: parsed.query ?? "",
+			filters: {
+				data_sources: parsed.filters?.data_sources ?? ["*"],
+				document_types: parsed.filters?.document_types ?? ["*"],
+				owners: parsed.filters?.owners ?? ["*"],
+			},
+			limit: parsed.limit ?? 10,
+			scoreThreshold: parsed.scoreThreshold ?? 0,
+			color: parsed.color ?? "zinc",
+			icon: parsed.icon ?? "filter",
+		};
 	};
 
 	return (

frontend/components/knowledge-filter-panel.tsx

@@ -96,15 +96,16 @@ export function KnowledgeFilterPanel() {
       setQuery(parsedFilterData.query || "");
 
       // Set the actual filter selections from the saved knowledge filter
-      const filters = parsedFilterData.filters;
+      const filters = parsedFilterData.filters || {};
 
       // Use the exact selections from the saved filter
       // Empty arrays mean "none selected" not "all selected"
+      // Provide defaults for missing fields to handle API-created filters
       const processedFilters = {
-        data_sources: filters.data_sources,
+        data_sources: filters.data_sources ?? ["*"],
-        document_types: filters.document_types,
+        document_types: filters.document_types ?? ["*"],
-        owners: filters.owners,
+        owners: filters.owners ?? ["*"],
-        connector_types: filters.connector_types || ["*"],
+        connector_types: filters.connector_types ?? ["*"],
       };
 
       console.log("[DEBUG] Loading filter selections:", processedFilters);
@@ -114,8 +115,8 @@ export function KnowledgeFilterPanel() {
       setScoreThreshold(parsedFilterData.scoreThreshold || 0);
       setName(selectedFilter.name);
       setDescription(selectedFilter.description || "");
-      setColor(parsedFilterData.color);
+      setColor(parsedFilterData.color ?? "zinc");
-      setIconKey(parsedFilterData.icon);
+      setIconKey(parsedFilterData.icon ?? "filter");
     }
   }, [selectedFilter, parsedFilterData]);
 
@@ -123,13 +124,20 @@ export function KnowledgeFilterPanel() {
   useEffect(() => {
     if (createMode && parsedFilterData) {
       setQuery(parsedFilterData.query || "");
-      setSelectedFilters(parsedFilterData.filters);
+      // Provide defaults for missing filter fields
+      const filters = parsedFilterData.filters || {};
+      setSelectedFilters({
+        data_sources: filters.data_sources ?? ["*"],
+        document_types: filters.document_types ?? ["*"],
+        owners: filters.owners ?? ["*"],
+        connector_types: filters.connector_types ?? ["*"],
+      });
       setResultLimit(parsedFilterData.limit || 10);
       setScoreThreshold(parsedFilterData.scoreThreshold || 0);
       setName("");
       setDescription("");
-      setColor(parsedFilterData.color);
+      setColor(parsedFilterData.color ?? "zinc");
-      setIconKey(parsedFilterData.icon);
+      setIconKey(parsedFilterData.icon ?? "filter");
     }
   }, [createMode, parsedFilterData]);
 

frontend/components/ui/multi-select.tsx

@@ -50,7 +50,10 @@ export function MultiSelect({
   const [open, setOpen] = React.useState(false);
   const [searchValue, setSearchValue] = React.useState("");
 
-  const isAllSelected = value.includes("*");
+  // Normalize value to empty array if undefined/null to prevent crashes
+  const safeValue = value ?? [];
+
+  const isAllSelected = safeValue.includes("*");
 
   const filteredOptions = options.filter((option) =>
     option.label.toLowerCase().includes(searchValue.toLowerCase()),
@@ -66,12 +69,12 @@ export function MultiSelect({
       }
     } else {
       let newValue: string[];
-      if (value.includes(optionValue)) {
+      if (safeValue.includes(optionValue)) {
         // Remove the item
-        newValue = value.filter((v) => v !== optionValue && v !== "*");
+        newValue = safeValue.filter((v) => v !== optionValue && v !== "*");
       } else {
         // Add the item and remove "All" if present
-        newValue = [...value.filter((v) => v !== "*"), optionValue];
+        newValue = [...safeValue.filter((v) => v !== "*"), optionValue];
 
         // Check max selection limit
         if (maxSelection && newValue.length > maxSelection) {
@@ -87,7 +90,7 @@ export function MultiSelect({
       return allOptionLabel;
     }
 
-    if (value.length === 0) {
+    if (safeValue.length === 0) {
       return placeholder;
     }
 
@@ -96,7 +99,7 @@ export function MultiSelect({
       .toLowerCase()
       .replace("select ", "")
       .replace("...", "");
-    return `${value.length} ${noun}`;
+    return `${safeValue.length} ${noun}`;
   };
 
   return (
@@ -152,7 +155,7 @@ export function MultiSelect({
                   <Check
                     className={cn(
                       "mr-2 h-4 w-4",
-                      value.includes(option.value)
+                      safeValue.includes(option.value)
                         ? "opacity-100"
                         : "opacity-0",
                     )}

frontend/contexts/knowledge-filter-context.tsx

@@ -84,7 +84,22 @@ export function KnowledgeFilterProvider({
     if (filter) {
       setCreateMode(false);
       try {
-        const parsed = JSON.parse(filter.query_data) as ParsedQueryData;
+        const raw = JSON.parse(filter.query_data);
+        // Normalize parsed data with defaults for missing fields
+        // This handles filters created via API with incomplete queryData
+        const parsed: ParsedQueryData = {
+          query: raw.query ?? "",
+          filters: {
+            data_sources: raw.filters?.data_sources ?? ["*"],
+            document_types: raw.filters?.document_types ?? ["*"],
+            owners: raw.filters?.owners ?? ["*"],
+            connector_types: raw.filters?.connector_types ?? ["*"],
+          },
+          limit: raw.limit ?? 10,
+          scoreThreshold: raw.scoreThreshold ?? 0,
+          color: raw.color ?? "zinc",
+          icon: raw.icon ?? "filter",
+        };
         setParsedFilterData(parsed);
 
         // Auto-open panel when filter is selected

sdks/python/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "openrag-sdk"
-version = "0.1.1"
+version = "0.1.2"
 description = "Official Python SDK for OpenRAG API"
 readme = "README.md"
 license = "MIT"
@@ -59,3 +59,4 @@ select = ["E", "F", "I", "UP"]
 [tool.mypy]
 python_version = "3.10"
 strict = true
+

sdks/typescript/package.json

@@ -1,6 +1,6 @@
 {
   "name": "openrag-sdk",
-  "version": "0.1.1",
+  "version": "0.1.2",
   "description": "Official TypeScript/JavaScript SDK for OpenRAG API",
   "main": "./dist/index.js",
   "module": "./dist/index.mjs",
@@ -56,3 +56,4 @@
     "node": ">=18.0.0"
   }
 }
+

src/api/knowledge_filter.py

@@ -8,6 +8,42 @@ from utils.logging_config import get_logger
 logger = get_logger(__name__)
 
 
+def normalize_query_data(query_data: str | dict) -> str:
+    """
+    Normalize query_data to ensure all required fields exist with defaults.
+    This prevents frontend crashes when API-created filters have incomplete data.
+    """
+    # Parse if string
+    if isinstance(query_data, str):
+        try:
+            data = json.loads(query_data)
+        except json.JSONDecodeError:
+            data = {}
+    else:
+        data = query_data or {}
+
+    # Ensure filters object exists with all required fields
+    filters = data.get("filters") or {}
+    normalized_filters = {
+        "data_sources": filters.get("data_sources", ["*"]),
+        "document_types": filters.get("document_types", ["*"]),
+        "owners": filters.get("owners", ["*"]),
+        "connector_types": filters.get("connector_types", ["*"]),
+    }
+
+    # Build normalized query_data with defaults
+    normalized = {
+        "query": data.get("query", ""),
+        "filters": normalized_filters,
+        "limit": data.get("limit", 10),
+        "scoreThreshold": data.get("scoreThreshold", 0),
+        "color": data.get("color", "zinc"),
+        "icon": data.get("icon", "filter"),
+    }
+
+    return json.dumps(normalized)
+
+
 async def create_knowledge_filter(
     request: Request, knowledge_filter_service, session_manager
 ):
@@ -25,6 +61,15 @@ async def create_knowledge_filter(
     if not query_data:
         return JSONResponse({"error": "Query data is required"}, status_code=400)
 
+    # Normalize query_data to ensure all required fields exist
+    try:
+        normalized_query_data = normalize_query_data(query_data)
+    except Exception as e:
+        logger.error(f"Failed to normalize query_data: {e}")
+        return JSONResponse(
+            {"error": f"Invalid queryData format: {str(e)}"}, status_code=400
+        )
+
     user = request.state.user
     jwt_token = session_manager.get_effective_jwt_token(user.user_id, request.state.jwt_token)
 
@@ -34,7 +79,7 @@ async def create_knowledge_filter(
         "id": filter_id,
         "name": name,
         "description": description,
-        "query_data": query_data,  # Store the full search query JSON
+        "query_data": normalized_query_data,  # Store normalized query JSON with defaults
         "owner": user.user_id,
         "allowed_users": payload.get("allowedUsers", []),  # ACL field for future use
         "allowed_groups": payload.get("allowedGroups", []),  # ACL field for future use
@@ -158,12 +203,22 @@ async def update_knowledge_filter(
             {"error": "Failed to delete existing knowledge filter"}, status_code=500
         )
 
+    # Normalize query_data if provided, otherwise use existing
+    query_data = payload.get("queryData", existing_filter["query_data"])
+    try:
+        normalized_query_data = normalize_query_data(query_data)
+    except Exception as e:
+        logger.error(f"Failed to normalize query_data: {e}")
+        return JSONResponse(
+            {"error": f"Invalid queryData format: {str(e)}"}, status_code=400
+        )
+
     # Create updated knowledge filter document with same ID
     updated_filter = {
         "id": filter_id,
         "name": payload.get("name", existing_filter["name"]),
         "description": payload.get("description", existing_filter["description"]),
-        "query_data": payload.get("queryData", existing_filter["query_data"]),
+        "query_data": normalized_query_data,
         "owner": existing_filter["owner"],
         "allowed_users": payload.get(
             "allowedUsers", existing_filter.get("allowed_users", [])