feat: add filter presets and update version to v0.9.15

- Added filter presets feature allowing users to save and quickly switch between filter combinations
- Fixed various bugs to improve overall stability
- Updated project version from 0.9.14 to 0.9.15 in pyproject.toml

.agents/skills/lora-manager-e2e/SKILL.md

@@ -0,0 +1,201 @@
+---
+name: lora-manager-e2e
+description: End-to-end testing and validation for LoRa Manager features. Use when performing automated E2E validation of LoRa Manager standalone mode, including starting/restarting the server, using Chrome DevTools MCP to interact with the web UI at http://127.0.0.1:8188/loras, and verifying frontend-to-backend functionality. Covers workflow validation, UI interaction testing, and integration testing between the standalone Python backend and the browser frontend.
+---
+
+# LoRa Manager E2E Testing
+
+This skill provides workflows and utilities for end-to-end testing of LoRa Manager using Chrome DevTools MCP.
+
+## Prerequisites
+
+- LoRa Manager project cloned and dependencies installed (`pip install -r requirements.txt`)
+- Chrome browser available for debugging
+- Chrome DevTools MCP connected
+
+## Quick Start Workflow
+
+### 1. Start LoRa Manager Standalone
+
+```python
+# Use the provided script to start the server
+python .agents/skills/lora-manager-e2e/scripts/start_server.py --port 8188
+```
+
+Or manually:
+```bash
+cd /home/miao/workspace/ComfyUI/custom_nodes/ComfyUI-Lora-Manager
+python standalone.py --port 8188
+```
+
+Wait for server ready message before proceeding.
+
+### 2. Open Chrome Debug Mode
+
+```bash
+# Chrome with remote debugging on port 9222
+google-chrome --remote-debugging-port=9222 --user-data-dir=/tmp/chrome-lora-manager http://127.0.0.1:8188/loras
+```
+
+### 3. Connect Chrome DevTools MCP
+
+Ensure the MCP server is connected to Chrome at `http://localhost:9222`.
+
+### 4. Navigate and Interact
+
+Use Chrome DevTools MCP tools to:
+- Take snapshots: `take_snapshot`
+- Click elements: `click`
+- Fill forms: `fill` or `fill_form`
+- Evaluate scripts: `evaluate_script`
+- Wait for elements: `wait_for`
+
+## Common E2E Test Patterns
+
+### Pattern: Full Page Load Verification
+
+```python
+# Navigate to LoRA list page
+navigate_page(type="url", url="http://127.0.0.1:8188/loras")
+
+# Wait for page to load
+wait_for(text="LoRAs", timeout=10000)
+
+# Take snapshot to verify UI state
+snapshot = take_snapshot()
+```
+
+### Pattern: Restart Server for Configuration Changes
+
+```python
+# Stop current server (if running)
+# Start with new configuration
+python .agents/skills/lora-manager-e2e/scripts/start_server.py --port 8188 --restart
+
+# Wait and refresh browser
+navigate_page(type="reload", ignoreCache=True)
+wait_for(text="LoRAs", timeout=15000)
+```
+
+### Pattern: Verify Backend API via Frontend
+
+```python
+# Execute script in browser to call backend API
+result = evaluate_script(function="""
+async () => {
+  const response = await fetch('/loras/api/list');
+  const data = await response.json();
+  return { count: data.length, firstItem: data[0]?.name };
+}
+""")
+```
+
+### Pattern: Form Submission Flow
+
+```python
+# Fill a form (e.g., search or filter)
+fill_form(elements=[
+    {"uid": "search-input", "value": "character"},
+])
+
+# Click submit button
+click(uid="search-button")
+
+# Wait for results
+wait_for(text="Results", timeout=5000)
+
+# Verify results via snapshot
+snapshot = take_snapshot()
+```
+
+### Pattern: Modal Dialog Interaction
+
+```python
+# Open modal (e.g., add LoRA)
+click(uid="add-lora-button")
+
+# Wait for modal to appear
+wait_for(text="Add LoRA", timeout=3000)
+
+# Fill modal form
+fill_form(elements=[
+    {"uid": "lora-name", "value": "Test LoRA"},
+    {"uid": "lora-path", "value": "/path/to/lora.safetensors"},
+])
+
+# Submit
+click(uid="modal-submit-button")
+
+# Wait for success message or close
+wait_for(text="Success", timeout=5000)
+```
+
+## Available Scripts
+
+### scripts/start_server.py
+
+Starts or restarts the LoRa Manager standalone server.
+
+```bash
+python scripts/start_server.py [--port PORT] [--restart] [--wait]
+```
+
+Options:
+- `--port`: Server port (default: 8188)
+- `--restart`: Kill existing server before starting
+- `--wait`: Wait for server to be ready before exiting
+
+### scripts/wait_for_server.py
+
+Polls server until ready or timeout.
+
+```bash
+python scripts/wait_for_server.py [--port PORT] [--timeout SECONDS]
+```
+
+## Test Scenarios Reference
+
+See [references/test-scenarios.md](references/test-scenarios.md) for detailed test scenarios including:
+- LoRA list display and filtering
+- Model metadata editing
+- Recipe creation and management
+- Settings configuration
+- Import/export functionality
+
+## Network Request Verification
+
+Use `list_network_requests` and `get_network_request` to verify API calls:
+
+```python
+# List recent XHR/fetch requests
+requests = list_network_requests(resourceTypes=["xhr", "fetch"])
+
+# Get details of specific request
+details = get_network_request(reqid=123)
+```
+
+## Console Message Monitoring
+
+```python
+# Check for errors or warnings
+messages = list_console_messages(types=["error", "warn"])
+```
+
+## Performance Testing
+
+```python
+# Start performance trace
+performance_start_trace(reload=True, autoStop=False)
+
+# Perform actions...
+
+# Stop and analyze
+results = performance_stop_trace()
+```
+
+## Cleanup
+
+Always ensure proper cleanup after tests:
+1. Stop the standalone server
+2. Close browser pages (keep at least one open)
+3. Clear temporary data if needed

.agents/skills/lora-manager-e2e/references/mcp-cheatsheet.md

@@ -0,0 +1,324 @@
+# Chrome DevTools MCP Cheatsheet for LoRa Manager
+
+Quick reference for common MCP commands used in LoRa Manager E2E testing.
+
+## Navigation
+
+```python
+# Navigate to LoRA list page
+navigate_page(type="url", url="http://127.0.0.1:8188/loras")
+
+# Reload page with cache clear
+navigate_page(type="reload", ignoreCache=True)
+
+# Go back/forward
+navigate_page(type="back")
+navigate_page(type="forward")
+```
+
+## Waiting
+
+```python
+# Wait for text to appear
+wait_for(text="LoRAs", timeout=10000)
+
+# Wait for specific element (via evaluate_script)
+evaluate_script(function="""
+() => {
+    return new Promise((resolve) => {
+        const check = () => {
+            if (document.querySelector('.lora-card')) {
+                resolve(true);
+            } else {
+                setTimeout(check, 100);
+            }
+        };
+        check();
+    });
+}
+""")
+```
+
+## Taking Snapshots
+
+```python
+# Full page snapshot
+snapshot = take_snapshot()
+
+# Verbose snapshot (more details)
+snapshot = take_snapshot(verbose=True)
+
+# Save to file
+take_snapshot(filePath="test-snapshots/page-load.json")
+```
+
+## Element Interaction
+
+```python
+# Click element
+click(uid="element-uid-from-snapshot")
+
+# Double click
+click(uid="element-uid", dblClick=True)
+
+# Fill input
+fill(uid="search-input", value="test query")
+
+# Fill multiple inputs
+fill_form(elements=[
+    {"uid": "input-1", "value": "value 1"},
+    {"uid": "input-2", "value": "value 2"},
+])
+
+# Hover
+hover(uid="lora-card-1")
+
+# Upload file
+upload_file(uid="file-input", filePath="/path/to/file.safetensors")
+```
+
+## Keyboard Input
+
+```python
+# Press key
+press_key(key="Enter")
+press_key(key="Escape")
+press_key(key="Tab")
+
+# Keyboard shortcuts
+press_key(key="Control+A")  # Select all
+press_key(key="Control+F")  # Find
+```
+
+## JavaScript Evaluation
+
+```python
+# Simple evaluation
+result = evaluate_script(function="() => document.title")
+
+# Async evaluation
+result = evaluate_script(function="""
+async () => {
+    const response = await fetch('/loras/api/list');
+    return await response.json();
+}
+""")
+
+# Check element existence
+exists = evaluate_script(function="""
+() => document.querySelector('.lora-card') !== null
+""")
+
+# Get element count
+count = evaluate_script(function="""
+() => document.querySelectorAll('.lora-card').length
+""")
+```
+
+## Network Monitoring
+
+```python
+# List all network requests
+requests = list_network_requests()
+
+# Filter by resource type
+xhr_requests = list_network_requests(resourceTypes=["xhr", "fetch"])
+
+# Get specific request details
+details = get_network_request(reqid=123)
+
+# Include preserved requests from previous navigations
+all_requests = list_network_requests(includePreservedRequests=True)
+```
+
+## Console Monitoring
+
+```python
+# List all console messages
+messages = list_console_messages()
+
+# Filter by type
+errors = list_console_messages(types=["error", "warn"])
+
+# Include preserved messages
+all_messages = list_console_messages(includePreservedMessages=True)
+
+# Get specific message
+details = get_console_message(msgid=1)
+```
+
+## Performance Testing
+
+```python
+# Start trace with page reload
+performance_start_trace(reload=True, autoStop=False)
+
+# Start trace without reload
+performance_start_trace(reload=False, autoStop=True, filePath="trace.json.gz")
+
+# Stop trace
+results = performance_stop_trace()
+
+# Stop and save
+performance_stop_trace(filePath="trace-results.json.gz")
+
+# Analyze specific insight
+insight = performance_analyze_insight(
+    insightSetId="results.insightSets[0].id",
+    insightName="LCPBreakdown"
+)
+```
+
+## Page Management
+
+```python
+# List open pages
+pages = list_pages()
+
+# Select a page
+select_page(pageId=0, bringToFront=True)
+
+# Create new page
+new_page(url="http://127.0.0.1:8188/loras")
+
+# Close page (keep at least one open!)
+close_page(pageId=1)
+
+# Resize page
+resize_page(width=1920, height=1080)
+```
+
+## Screenshots
+
+```python
+# Full page screenshot
+take_screenshot(fullPage=True)
+
+# Viewport screenshot
+take_screenshot()
+
+# Element screenshot
+take_screenshot(uid="lora-card-1")
+
+# Save to file
+take_screenshot(filePath="screenshots/page.png", format="png")
+
+# JPEG with quality
+take_screenshot(filePath="screenshots/page.jpg", format="jpeg", quality=90)
+```
+
+## Dialog Handling
+
+```python
+# Accept dialog
+handle_dialog(action="accept")
+
+# Accept with text input
+handle_dialog(action="accept", promptText="user input")
+
+# Dismiss dialog
+handle_dialog(action="dismiss")
+```
+
+## Device Emulation
+
+```python
+# Mobile viewport
+emulate(viewport={"width": 375, "height": 667, "isMobile": True, "hasTouch": True})
+
+# Tablet viewport
+emulate(viewport={"width": 768, "height": 1024, "isMobile": True, "hasTouch": True})
+
+# Desktop viewport
+emulate(viewport={"width": 1920, "height": 1080})
+
+# Network throttling
+emulate(networkConditions="Slow 3G")
+emulate(networkConditions="Fast 4G")
+
+# CPU throttling
+emulate(cpuThrottlingRate=4)  # 4x slowdown
+
+# Geolocation
+emulate(geolocation={"latitude": 37.7749, "longitude": -122.4194})
+
+# User agent
+emulate(userAgent="Mozilla/5.0 (Custom)")
+
+# Reset emulation
+emulate(viewport=None, networkConditions="No emulation", userAgent=None)
+```
+
+## Drag and Drop
+
+```python
+# Drag element to another
+drag(from_uid="draggable-item", to_uid="drop-zone")
+```
+
+## Common LoRa Manager Test Patterns
+
+### Verify LoRA Cards Loaded
+
+```python
+navigate_page(type="url", url="http://127.0.0.1:8188/loras")
+wait_for(text="LoRAs", timeout=10000)
+
+# Check if cards loaded
+result = evaluate_script(function="""
+() => {
+    const cards = document.querySelectorAll('.lora-card');
+    return {
+        count: cards.length,
+        hasData: cards.length > 0
+    };
+}
+""")
+```
+
+### Search and Verify Results
+
+```python
+fill(uid="search-input", value="character")
+press_key(key="Enter")
+wait_for(timeout=2000)  # Wait for debounce
+
+# Check results
+result = evaluate_script(function="""
+() => {
+    const cards = document.querySelectorAll('.lora-card');
+    const names = Array.from(cards).map(c => c.dataset.name || c.textContent);
+    return { count: cards.length, names };
+}
+""")
+```
+
+### Check API Response
+
+```python
+# Trigger API call
+evaluate_script(function="""
+() => window.loraApiCallPromise = fetch('/loras/api/list').then(r => r.json())
+""")
+
+# Wait and get result
+import time
+time.sleep(1)
+
+result = evaluate_script(function="""
+async () => await window.loraApiCallPromise
+""")
+```
+
+### Monitor Console for Errors
+
+```python
+# Before test: clear console (navigate reloads)
+navigate_page(type="reload")
+
+# ... perform actions ...
+
+# Check for errors
+errors = list_console_messages(types=["error"])
+assert len(errors) == 0, f"Console errors: {errors}"
+```

.agents/skills/lora-manager-e2e/references/test-scenarios.md

@@ -0,0 +1,272 @@
+# LoRa Manager E2E Test Scenarios
+
+This document provides detailed test scenarios for end-to-end validation of LoRa Manager features.
+
+## Table of Contents
+
+1. [LoRA List Page](#lora-list-page)
+2. [Model Details](#model-details)
+3. [Recipes](#recipes)
+4. [Settings](#settings)
+5. [Import/Export](#importexport)
+
+---
+
+## LoRA List Page
+
+### Scenario: Page Load and Display
+
+**Objective**: Verify the LoRA list page loads correctly and displays models.
+
+**Steps**:
+1. Navigate to `http://127.0.0.1:8188/loras`
+2. Wait for page title "LoRAs" to appear
+3. Take snapshot to verify:
+   - Header with "LoRAs" title is visible
+   - Search/filter controls are present
+   - Grid/list view toggle exists
+   - LoRA cards are displayed (if models exist)
+   - Pagination controls (if applicable)
+
+**Expected Result**: Page loads without errors, UI elements are present.
+
+### Scenario: Search Functionality
+
+**Objective**: Verify search filters LoRA models correctly.
+
+**Steps**:
+1. Ensure at least one LoRA exists with known name (e.g., "test-character")
+2. Navigate to LoRA list page
+3. Enter search term in search box: "test"
+4. Press Enter or click search button
+5. Wait for results to update
+
+**Expected Result**: Only LoRAs matching search term are displayed.
+
+**Verification Script**:
+```python
+# After search, verify filtered results
+evaluate_script(function="""
+() => {
+  const cards = document.querySelectorAll('.lora-card');
+  const names = Array.from(cards).map(c => c.dataset.name);
+  return { count: cards.length, names };
+}
+""")
+```
+
+### Scenario: Filter by Tags
+
+**Objective**: Verify tag filtering works correctly.
+
+**Steps**:
+1. Navigate to LoRA list page
+2. Click on a tag (e.g., "character", "style")
+3. Wait for filtered results
+
+**Expected Result**: Only LoRAs with selected tag are displayed.
+
+### Scenario: View Mode Toggle
+
+**Objective**: Verify grid/list view toggle works.
+
+**Steps**:
+1. Navigate to LoRA list page
+2. Click list view button
+3. Verify list layout
+4. Click grid view button
+5. Verify grid layout
+
+**Expected Result**: View mode changes correctly, layout updates.
+
+---
+
+## Model Details
+
+### Scenario: Open Model Details
+
+**Objective**: Verify clicking a LoRA opens its details.
+
+**Steps**:
+1. Navigate to LoRA list page
+2. Click on a LoRA card
+3. Wait for details panel/modal to open
+
+**Expected Result**: Details panel shows:
+- Model name
+- Preview image
+- Metadata (trigger words, tags, etc.)
+- Action buttons (edit, delete, etc.)
+
+### Scenario: Edit Model Metadata
+
+**Objective**: Verify metadata editing works end-to-end.
+
+**Steps**:
+1. Open a LoRA's details
+2. Click "Edit" button
+3. Modify trigger words field
+4. Add/remove tags
+5. Save changes
+6. Refresh page
+7. Reopen the same LoRA
+
+**Expected Result**: Changes persist after refresh.
+
+### Scenario: Delete Model
+
+**Objective**: Verify model deletion works.
+
+**Steps**:
+1. Open a LoRA's details
+2. Click "Delete" button
+3. Confirm deletion in dialog
+4. Wait for removal
+
+**Expected Result**: Model removed from list, success message shown.
+
+---
+
+## Recipes
+
+### Scenario: Recipe List Display
+
+**Objective**: Verify recipes page loads and displays recipes.
+
+**Steps**:
+1. Navigate to `http://127.0.0.1:8188/recipes`
+2. Wait for "Recipes" title
+3. Take snapshot
+
+**Expected Result**: Recipe list displayed with cards/items.
+
+### Scenario: Create New Recipe
+
+**Objective**: Verify recipe creation workflow.
+
+**Steps**:
+1. Navigate to recipes page
+2. Click "New Recipe" button
+3. Fill recipe form:
+   - Name: "Test Recipe"
+   - Description: "E2E test recipe"
+   - Add LoRA models
+4. Save recipe
+5. Verify recipe appears in list
+
+**Expected Result**: New recipe created and displayed.
+
+### Scenario: Apply Recipe
+
+**Objective**: Verify applying a recipe to ComfyUI.
+
+**Steps**:
+1. Open a recipe
+2. Click "Apply" or "Load in ComfyUI"
+3. Verify action completes
+
+**Expected Result**: Recipe applied successfully.
+
+---
+
+## Settings
+
+### Scenario: Settings Page Load
+
+**Objective**: Verify settings page displays correctly.
+
+**Steps**:
+1. Navigate to `http://127.0.0.1:8188/settings`
+2. Wait for "Settings" title
+3. Take snapshot
+
+**Expected Result**: Settings form with various options displayed.
+
+### Scenario: Change Setting and Restart
+
+**Objective**: Verify settings persist after restart.
+
+**Steps**:
+1. Navigate to settings page
+2. Change a setting (e.g., default view mode)
+3. Save settings
+4. Restart server: `python scripts/start_server.py --restart --wait`
+5. Refresh browser page
+6. Navigate to settings
+
+**Expected Result**: Changed setting value persists.
+
+---
+
+## Import/Export
+
+### Scenario: Export Models List
+
+**Objective**: Verify export functionality.
+
+**Steps**:
+1. Navigate to LoRA list
+2. Click "Export" button
+3. Select format (JSON/CSV)
+4. Download file
+
+**Expected Result**: File downloaded with correct data.
+
+### Scenario: Import Models
+
+**Objective**: Verify import functionality.
+
+**Steps**:
+1. Prepare import file
+2. Navigate to import page
+3. Upload file
+4. Verify import results
+
+**Expected Result**: Models imported successfully, confirmation shown.
+
+---
+
+## API Integration Tests
+
+### Scenario: Verify API Endpoints
+
+**Objective**: Verify backend API responds correctly.
+
+**Test via browser console**:
+```javascript
+// List LoRAs
+fetch('/loras/api/list').then(r => r.json()).then(console.log)
+
+// Get LoRA details
+fetch('/loras/api/detail/<id>').then(r => r.json()).then(console.log)
+
+// Search LoRAs
+fetch('/loras/api/search?q=test').then(r => r.json()).then(console.log)
+```
+
+**Expected Result**: APIs return valid JSON with expected structure.
+
+---
+
+## Console Error Monitoring
+
+During all tests, monitor browser console for errors:
+
+```python
+# Check for JavaScript errors
+messages = list_console_messages(types=["error"])
+assert len(messages) == 0, f"Console errors found: {messages}"
+```
+
+## Network Request Verification
+
+Verify key API calls are made:
+
+```python
+# List XHR requests
+requests = list_network_requests(resourceTypes=["xhr", "fetch"])
+
+# Look for specific endpoints
+lora_list_requests = [r for r in requests if "/api/list" in r.get("url", "")]
+assert len(lora_list_requests) > 0, "LoRA list API not called"
+```

.agents/skills/lora-manager-e2e/scripts/example_e2e_test.py

@@ -0,0 +1,193 @@
+#!/usr/bin/env python3
+"""
+Example E2E test demonstrating LoRa Manager testing workflow.
+
+This script shows how to:
+1. Start the standalone server
+2. Use Chrome DevTools MCP to interact with the UI
+3. Verify functionality end-to-end
+
+Note: This is a template. Actual execution requires Chrome DevTools MCP.
+"""
+
+import subprocess
+import sys
+import time
+
+
+def run_test():
+    """Run example E2E test flow."""
+    
+    print("=" * 60)
+    print("LoRa Manager E2E Test Example")
+    print("=" * 60)
+    
+    # Step 1: Start server
+    print("\n[1/5] Starting LoRa Manager standalone server...")
+    result = subprocess.run(
+        [sys.executable, "start_server.py", "--port", "8188", "--wait", "--timeout", "30"],
+        capture_output=True,
+        text=True
+    )
+    if result.returncode != 0:
+        print(f"Failed to start server: {result.stderr}")
+        return 1
+    print("Server ready!")
+    
+    # Step 2: Open Chrome (manual step - show command)
+    print("\n[2/5] Open Chrome with debug mode:")
+    print("google-chrome --remote-debugging-port=9222 --user-data-dir=/tmp/chrome-lora-manager http://127.0.0.1:8188/loras")
+    print("(In actual test, this would be automated via MCP)")
+    
+    # Step 3: Navigate and verify page load
+    print("\n[3/5] Page Load Verification:")
+    print("""
+    MCP Commands to execute:
+    1. navigate_page(type="url", url="http://127.0.0.1:8188/loras")
+    2. wait_for(text="LoRAs", timeout=10000)
+    3. snapshot = take_snapshot()
+    """)
+    
+    # Step 4: Test search functionality
+    print("\n[4/5] Search Functionality Test:")
+    print("""
+    MCP Commands to execute:
+    1. fill(uid="search-input", value="test")
+    2. press_key(key="Enter")
+    3. wait_for(text="Results", timeout=5000)
+    4. result = evaluate_script(function="""
+        () => {
+            const cards = document.querySelectorAll('.lora-card');
+            return { count: cards.length };
+        }
+        """)
+    """)
+    
+    # Step 5: Verify API
+    print("\n[5/5] API Verification:")
+    print("""
+    MCP Commands to execute:
+    1. api_result = evaluate_script(function="""
+        async () => {
+            const response = await fetch('/loras/api/list');
+            const data = await response.json();
+            return { count: data.length, status: response.status };
+        }
+        """)
+    2. Verify api_result['status'] == 200
+    """)
+    
+    print("\n" + "=" * 60)
+    print("Test flow completed!")
+    print("=" * 60)
+    
+    return 0
+
+
+def example_restart_flow():
+    """Example: Testing configuration change that requires restart."""
+    
+    print("\n" + "=" * 60)
+    print("Example: Server Restart Flow")
+    print("=" * 60)
+    
+    print("""
+    Scenario: Change setting and verify after restart
+    
+    Steps:
+    1. Navigate to settings page
+       - navigate_page(type="url", url="http://127.0.0.1:8188/settings")
+    
+    2. Change a setting (e.g., theme)
+       - fill(uid="theme-select", value="dark")
+       - click(uid="save-settings-button")
+    
+    3. Restart server
+       - subprocess.run([python, "start_server.py", "--restart", "--wait"])
+    
+    4. Refresh browser
+       - navigate_page(type="reload", ignoreCache=True)
+       - wait_for(text="LoRAs", timeout=15000)
+    
+    5. Verify setting persisted
+       - navigate_page(type="url", url="http://127.0.0.1:8188/settings")
+       - theme = evaluate_script(function="() => document.querySelector('#theme-select').value")
+       - assert theme == "dark"
+    """)
+
+
+def example_modal_interaction():
+    """Example: Testing modal dialog interaction."""
+    
+    print("\n" + "=" * 60)
+    print("Example: Modal Dialog Interaction")
+    print("=" * 60)
+    
+    print("""
+    Scenario: Add new LoRA via modal
+    
+    Steps:
+    1. Open modal
+       - click(uid="add-lora-button")
+       - wait_for(text="Add LoRA", timeout=3000)
+    
+    2. Fill form
+       - fill_form(elements=[
+           {"uid": "lora-name", "value": "Test Character"},
+           {"uid": "lora-path", "value": "/models/test.safetensors"},
+       ])
+    
+    3. Submit
+       - click(uid="modal-submit-button")
+    
+    4. Verify success
+       - wait_for(text="Successfully added", timeout=5000)
+       - snapshot = take_snapshot()
+    """)
+
+
+def example_network_monitoring():
+    """Example: Network request monitoring."""
+    
+    print("\n" + "=" * 60)
+    print("Example: Network Request Monitoring")
+    print("=" * 60)
+    
+    print("""
+    Scenario: Verify API calls during user interaction
+    
+    Steps:
+    1. Clear network log (implicit on navigation)
+       - navigate_page(type="url", url="http://127.0.0.1:8188/loras")
+    
+    2. Perform action that triggers API call
+       - fill(uid="search-input", value="character")
+       - press_key(key="Enter")
+    
+    3. List network requests
+       - requests = list_network_requests(resourceTypes=["xhr", "fetch"])
+    
+    4. Find search API call
+       - search_requests = [r for r in requests if "/api/search" in r.get("url", "")]
+       - assert len(search_requests) > 0, "Search API was not called"
+    
+    5. Get request details
+       - if search_requests:
+           details = get_network_request(reqid=search_requests[0]["reqid"])
+           - Verify request method, response status, etc.
+    """)
+
+
+if __name__ == "__main__":
+    print("LoRa Manager E2E Test Examples\n")
+    print("This script demonstrates E2E testing patterns.\n")
+    print("Note: Actual execution requires Chrome DevTools MCP connection.\n")
+    
+    run_test()
+    example_restart_flow()
+    example_modal_interaction()
+    example_network_monitoring()
+    
+    print("\n" + "=" * 60)
+    print("All examples shown!")
+    print("=" * 60)

.agents/skills/lora-manager-e2e/scripts/start_server.py

@@ -0,0 +1,169 @@
+#!/usr/bin/env python3
+"""
+Start or restart LoRa Manager standalone server for E2E testing.
+"""
+
+import argparse
+import subprocess
+import sys
+import time
+import socket
+import signal
+import os
+
+
+def find_server_process(port: int) -> list[int]:
+    """Find PIDs of processes listening on the given port."""
+    try:
+        result = subprocess.run(
+            ["lsof", "-ti", f":{port}"],
+            capture_output=True,
+            text=True,
+            check=False
+        )
+        if result.returncode == 0 and result.stdout.strip():
+            return [int(pid) for pid in result.stdout.strip().split("\n") if pid]
+    except FileNotFoundError:
+        # lsof not available, try netstat
+        try:
+            result = subprocess.run(
+                ["netstat", "-tlnp"],
+                capture_output=True,
+                text=True,
+                check=False
+            )
+            pids = []
+            for line in result.stdout.split("\n"):
+                if f":{port}" in line:
+                    parts = line.split()
+                    for part in parts:
+                        if "/" in part:
+                            try:
+                                pid = int(part.split("/")[0])
+                                pids.append(pid)
+                            except ValueError:
+                                pass
+            return pids
+        except FileNotFoundError:
+            pass
+    return []
+
+
+def kill_server(port: int) -> None:
+    """Kill processes using the specified port."""
+    pids = find_server_process(port)
+    for pid in pids:
+        try:
+            os.kill(pid, signal.SIGTERM)
+            print(f"Sent SIGTERM to process {pid}")
+        except ProcessLookupError:
+            pass
+    
+    # Wait for processes to terminate
+    time.sleep(1)
+    
+    # Force kill if still running
+    pids = find_server_process(port)
+    for pid in pids:
+        try:
+            os.kill(pid, signal.SIGKILL)
+            print(f"Sent SIGKILL to process {pid}")
+        except ProcessLookupError:
+            pass
+
+
+def is_server_ready(port: int, timeout: float = 0.5) -> bool:
+    """Check if server is accepting connections."""
+    try:
+        with socket.create_connection(("127.0.0.1", port), timeout=timeout):
+            return True
+    except (socket.timeout, ConnectionRefusedError, OSError):
+        return False
+
+
+def wait_for_server(port: int, timeout: int = 30) -> bool:
+    """Wait for server to become ready."""
+    start = time.time()
+    while time.time() - start < timeout:
+        if is_server_ready(port):
+            return True
+        time.sleep(0.5)
+    return False
+
+
+def main() -> int:
+    parser = argparse.ArgumentParser(
+        description="Start LoRa Manager standalone server for E2E testing"
+    )
+    parser.add_argument(
+        "--port",
+        type=int,
+        default=8188,
+        help="Server port (default: 8188)"
+    )
+    parser.add_argument(
+        "--restart",
+        action="store_true",
+        help="Kill existing server before starting"
+    )
+    parser.add_argument(
+        "--wait",
+        action="store_true",
+        help="Wait for server to be ready before exiting"
+    )
+    parser.add_argument(
+        "--timeout",
+        type=int,
+        default=30,
+        help="Timeout for waiting (default: 30)"
+    )
+    
+    args = parser.parse_args()
+    
+    # Get project root (parent of .agents directory)
+    script_dir = os.path.dirname(os.path.abspath(__file__))
+    skill_dir = os.path.dirname(script_dir)
+    project_root = os.path.dirname(os.path.dirname(os.path.dirname(skill_dir)))
+    
+    # Restart if requested
+    if args.restart:
+        print(f"Killing existing server on port {args.port}...")
+        kill_server(args.port)
+        time.sleep(1)
+    
+    # Check if already running
+    if is_server_ready(args.port):
+        print(f"Server already running on port {args.port}")
+        return 0
+    
+    # Start server
+    print(f"Starting LoRa Manager standalone server on port {args.port}...")
+    cmd = [sys.executable, "standalone.py", "--port", str(args.port)]
+    
+    # Start in background
+    process = subprocess.Popen(
+        cmd,
+        cwd=project_root,
+        stdout=subprocess.PIPE,
+        stderr=subprocess.PIPE,
+        start_new_session=True
+    )
+    
+    print(f"Server process started with PID {process.pid}")
+    
+    # Wait for ready if requested
+    if args.wait:
+        print(f"Waiting for server to be ready (timeout: {args.timeout}s)...")
+        if wait_for_server(args.port, args.timeout):
+            print(f"Server ready at http://127.0.0.1:{args.port}/loras")
+            return 0
+        else:
+            print(f"Timeout waiting for server")
+            return 1
+    
+    print(f"Server starting at http://127.0.0.1:{args.port}/loras")
+    return 0
+
+
+if __name__ == "__main__":
+    sys.exit(main())

.agents/skills/lora-manager-e2e/scripts/wait_for_server.py

@@ -0,0 +1,61 @@
+#!/usr/bin/env python3
+"""
+Wait for LoRa Manager server to become ready.
+"""
+
+import argparse
+import socket
+import sys
+import time
+
+
+def is_server_ready(port: int, timeout: float = 0.5) -> bool:
+    """Check if server is accepting connections."""
+    try:
+        with socket.create_connection(("127.0.0.1", port), timeout=timeout):
+            return True
+    except (socket.timeout, ConnectionRefusedError, OSError):
+        return False
+
+
+def wait_for_server(port: int, timeout: int = 30) -> bool:
+    """Wait for server to become ready."""
+    start = time.time()
+    while time.time() - start < timeout:
+        if is_server_ready(port):
+            return True
+        time.sleep(0.5)
+    return False
+
+
+def main() -> int:
+    parser = argparse.ArgumentParser(
+        description="Wait for LoRa Manager server to become ready"
+    )
+    parser.add_argument(
+        "--port",
+        type=int,
+        default=8188,
+        help="Server port (default: 8188)"
+    )
+    parser.add_argument(
+        "--timeout",
+        type=int,
+        default=30,
+        help="Timeout in seconds (default: 30)"
+    )
+    
+    args = parser.parse_args()
+    
+    print(f"Waiting for server on port {args.port} (timeout: {args.timeout}s)...")
+    
+    if wait_for_server(args.port, args.timeout):
+        print(f"Server ready at http://127.0.0.1:{args.port}/loras")
+        return 0
+    else:
+        print(f"Timeout: Server not ready after {args.timeout}s")
+        return 1
+
+
+if __name__ == "__main__":
+    sys.exit(main())

README.md

@@ -34,6 +34,15 @@ Enhance your Civitai browsing experience with our companion browser extension! S
 
 ## Release Notes
 
+### v0.9.15
+* **Filter Presets** - Save filter combinations as presets for quick switching and reapplication.
+* **Bug Fixes** - Fixed various bugs for improved stability.
+
+### v0.9.14
+* **LoRA Cycler Node** - Introduced a new LoRA Cycler node that enables iteration through specified LoRAs with support for repeat count and pause iteration functionality. Refer to the new "Lora Cycler" template workflow for concrete example.
+* **Enhanced Prompt Node with Tag Autocomplete** - Enhanced the Prompt node with comprehensive tag autocomplete based on merged Danbooru + e621 tags. Supports tag search and autocomplete functionality. Implemented a command system with shortcuts like `/char` or `/artist` for category-specific tag searching. Added `/ac` or `/noac` commands to quickly enable or disable autocomplete. Refer to the "Lora Manager Basic" template workflow in ComfyUI -> Templates -> ComfyUI-Lora-Manager for detailed tips.
+* **Bug Fixes & Stability** - Addressed multiple bugs and improved overall stability.
+
 ### v0.9.12
 * **LoRA Randomizer System** - Introduced a comprehensive LoRA randomization system featuring LoRA Pool and LoRA Randomizer nodes for flexible and dynamic generation workflows.
 * **LoRA Randomizer Template** - Refer to the new "LoRA Randomizer" template workflow for detailed examples of flexible randomization modes, lock & reuse options, and other features.

__init__.py

@@ -1,15 +1,17 @@
 try:  # pragma: no cover - import fallback for pytest collection
     from .py.lora_manager import LoraManager
-    from .py.nodes.lora_loader import LoraManagerLoader, LoraManagerTextLoader
-    from .py.nodes.trigger_word_toggle import TriggerWordToggle
-    from .py.nodes.prompt import PromptLoraManager
-    from .py.nodes.lora_stacker import LoraStacker
+    from .py.nodes.lora_loader import LoraLoaderLM, LoraTextLoaderLM
+    from .py.nodes.trigger_word_toggle import TriggerWordToggleLM
+    from .py.nodes.prompt import PromptLM
+    from .py.nodes.text import TextLM
+    from .py.nodes.lora_stacker import LoraStackerLM
     from .py.nodes.save_image import SaveImageLM
-    from .py.nodes.debug_metadata import DebugMetadata
+    from .py.nodes.debug_metadata import DebugMetadataLM
     from .py.nodes.wanvideo_lora_select import WanVideoLoraSelectLM
-    from .py.nodes.wanvideo_lora_select_from_text import WanVideoLoraSelectFromText
-    from .py.nodes.lora_pool import LoraPoolNode
-    from .py.nodes.lora_randomizer import LoraRandomizerNode
+    from .py.nodes.wanvideo_lora_select_from_text import WanVideoLoraTextSelectLM
+    from .py.nodes.lora_pool import LoraPoolLM
+    from .py.nodes.lora_randomizer import LoraRandomizerLM
+    from .py.nodes.lora_cycler import LoraCyclerLM
     from .py.metadata_collector import init as init_metadata_collector
 except (
     ImportError
@@ -22,44 +24,50 @@ except (
     if str(package_root) not in sys.path:
         sys.path.append(str(package_root))
 
-    PromptLoraManager = importlib.import_module("py.nodes.prompt").PromptLoraManager
+    PromptLM = importlib.import_module("py.nodes.prompt").PromptLM
+    TextLM = importlib.import_module("py.nodes.text").TextLM
     LoraManager = importlib.import_module("py.lora_manager").LoraManager
-    LoraManagerLoader = importlib.import_module(
+    LoraLoaderLM = importlib.import_module(
         "py.nodes.lora_loader"
-    ).LoraManagerLoader
-    LoraManagerTextLoader = importlib.import_module(
+    ).LoraLoaderLM
+    LoraTextLoaderLM = importlib.import_module(
         "py.nodes.lora_loader"
-    ).LoraManagerTextLoader
-    TriggerWordToggle = importlib.import_module(
+    ).LoraTextLoaderLM
+    TriggerWordToggleLM = importlib.import_module(
         "py.nodes.trigger_word_toggle"
-    ).TriggerWordToggle
-    LoraStacker = importlib.import_module("py.nodes.lora_stacker").LoraStacker
+    ).TriggerWordToggleLM
+    LoraStackerLM = importlib.import_module("py.nodes.lora_stacker").LoraStackerLM
     SaveImageLM = importlib.import_module("py.nodes.save_image").SaveImageLM
-    DebugMetadata = importlib.import_module("py.nodes.debug_metadata").DebugMetadata
+    DebugMetadataLM = importlib.import_module("py.nodes.debug_metadata").DebugMetadataLM
     WanVideoLoraSelectLM = importlib.import_module(
         "py.nodes.wanvideo_lora_select"
     ).WanVideoLoraSelectLM
-    WanVideoLoraSelectFromText = importlib.import_module(
+    WanVideoLoraTextSelectLM = importlib.import_module(
         "py.nodes.wanvideo_lora_select_from_text"
-    ).WanVideoLoraSelectFromText
-    LoraPoolNode = importlib.import_module("py.nodes.lora_pool").LoraPoolNode
-    LoraRandomizerNode = importlib.import_module(
+    ).WanVideoLoraTextSelectLM
+    LoraPoolLM = importlib.import_module("py.nodes.lora_pool").LoraPoolLM
+    LoraRandomizerLM = importlib.import_module(
         "py.nodes.lora_randomizer"
-    ).LoraRandomizerNode
+    ).LoraRandomizerLM
+    LoraCyclerLM = importlib.import_module(
+        "py.nodes.lora_cycler"
+    ).LoraCyclerLM
     init_metadata_collector = importlib.import_module("py.metadata_collector").init
 
 NODE_CLASS_MAPPINGS = {
-    PromptLoraManager.NAME: PromptLoraManager,
-    LoraManagerLoader.NAME: LoraManagerLoader,
-    LoraManagerTextLoader.NAME: LoraManagerTextLoader,
-    TriggerWordToggle.NAME: TriggerWordToggle,
-    LoraStacker.NAME: LoraStacker,
+    PromptLM.NAME: PromptLM,
+    TextLM.NAME: TextLM,
+    LoraLoaderLM.NAME: LoraLoaderLM,
+    LoraTextLoaderLM.NAME: LoraTextLoaderLM,
+    TriggerWordToggleLM.NAME: TriggerWordToggleLM,
+    LoraStackerLM.NAME: LoraStackerLM,
     SaveImageLM.NAME: SaveImageLM,
-    DebugMetadata.NAME: DebugMetadata,
+    DebugMetadataLM.NAME: DebugMetadataLM,
     WanVideoLoraSelectLM.NAME: WanVideoLoraSelectLM,
-    WanVideoLoraSelectFromText.NAME: WanVideoLoraSelectFromText,
-    LoraPoolNode.NAME: LoraPoolNode,
-    LoraRandomizerNode.NAME: LoraRandomizerNode,
+    WanVideoLoraTextSelectLM.NAME: WanVideoLoraTextSelectLM,
+    LoraPoolLM.NAME: LoraPoolLM,
+    LoraRandomizerLM.NAME: LoraRandomizerLM,
+    LoraCyclerLM.NAME: LoraCyclerLM,
 }
 
 WEB_DIRECTORY = "./web/comfyui"

docs/dom-widgets/README.md

@@ -0,0 +1,28 @@
+# DOM Widgets Documentation
+
+Documentation for custom DOM widget development in ComfyUI LoRA Manager.
+
+## Files
+
+- **[Value Persistence Best Practices](value-persistence-best-practices.md)** - Essential guide for implementing text input DOM widgets that persist values correctly
+
+## Key Lessons
+
+### Common Anti-Patterns
+
+❌ **Don't**: Create internal state variables  
+❌ **Don't**: Use v-model for text inputs  
+❌ **Don't**: Add serializeValue, onSetValue callbacks  
+❌ **Don't**: Watch props.widget.value  
+
+### Best Practices
+
+✅ **Do**: Use DOM element as single source of truth  
+✅ **Do**: Store DOM reference on widget.inputEl  
+✅ **Do**: Direct getValue/setValue to DOM  
+✅ **Do**: Clean up reference on unmount
+
+## Related Documentation
+
+- [DOM Widget Development Guide](../dom_widget_dev_guide.md) - Comprehensive guide for building DOM widgets
+- [ComfyUI Built-in Example](../../../../code/ComfyUI_frontend/src/renderer/extensions/vueNodes/widgets/composables/useStringWidget.ts) - Reference implementation

docs/dom-widgets/value-persistence-best-practices.md

@@ -0,0 +1,225 @@
+# DOM Widget Value Persistence - Best Practices
+
+## Overview
+
+DOM widgets require different persistence patterns depending on their complexity. This document covers two patterns:
+
+1. **Simple Text Widgets**: DOM element as source of truth (e.g., textarea, input)
+2. **Complex Widgets**: Internal value with `widget.callback` (e.g., LoraPoolWidget, RandomizerWidget)
+
+## Understanding ComfyUI's Built-in Callback Mechanism
+
+When `widget.value` is set (e.g., during workflow load), ComfyUI's `domWidget.ts` triggers this flow:
+
+```typescript
+// From ComfyUI_frontend/src/scripts/domWidget.ts:146-149
+set value(v: V) {
+  this.options.setValue?.(v)      // 1. Update internal state
+  this.callback?.(this.value)     // 2. Notify listeners for UI updates
+}
+```
+
+This means:
+- `setValue()` handles storing the value
+- `widget.callback()` is automatically called to notify the UI
+- You don't need custom callback mechanisms like `onSetValue`
+
+---
+
+## Pattern 1: Simple Text Input Widgets
+
+For widgets where the value IS the DOM element's text content (textarea, input fields).
+
+### When to Use
+
+- Single text input/textarea widgets
+- Value is a simple string
+- No complex state management needed
+
+### Implementation
+
+**main.ts:**
+```typescript
+const widget = node.addDOMWidget(name, type, container, {
+  getValue() {
+    return widget.inputEl?.value ?? ''
+  },
+  setValue(v: string) {
+    if (widget.inputEl) {
+      widget.inputEl.value = v ?? ''
+    }
+  }
+})
+```
+
+**Vue Component:**
+```typescript
+onMounted(() => {
+  if (textareaRef.value) {
+    props.widget.inputEl = textareaRef.value
+  }
+})
+
+onUnmounted(() => {
+  if (props.widget.inputEl === textareaRef.value) {
+    props.widget.inputEl = undefined
+  }
+})
+```
+
+### Why This Works
+
+- Single source of truth: the DOM element
+- `getValue()` reads directly from DOM
+- `setValue()` writes directly to DOM
+- No sync issues between multiple state variables
+
+---
+
+## Pattern 2: Complex Widgets
+
+For widgets with structured data (JSON configs, arrays, objects) where the value cannot be stored in a DOM element.
+
+### When to Use
+
+- Value is a complex object/array (e.g., `{ loras: [...], settings: {...} }`)
+- Multiple UI elements contribute to the value
+- Vue reactive state manages the UI
+
+### Implementation
+
+**main.ts:**
+```typescript
+let internalValue: MyConfig | undefined
+
+const widget = node.addDOMWidget(name, type, container, {
+  getValue() {
+    return internalValue
+  },
+  setValue(v: MyConfig) {
+    internalValue = v
+    // NO custom onSetValue needed - widget.callback is called automatically
+  },
+  serialize: true  // Ensure value is saved with workflow
+})
+```
+
+**Vue Component:**
+```typescript
+const config = ref<MyConfig>(getDefaultConfig())
+
+onMounted(() => {
+  // Set up callback for UI updates when widget.value changes externally
+  // (e.g., workflow load, undo/redo)
+  props.widget.callback = (newValue: MyConfig) => {
+    if (newValue) {
+      config.value = newValue
+    }
+  }
+
+  // Restore initial value if workflow was already loaded
+  if (props.widget.value) {
+    config.value = props.widget.value
+  }
+})
+
+// When UI changes, update widget value
+function onConfigChange(newConfig: MyConfig) {
+  config.value = newConfig
+  props.widget.value = newConfig  // This also triggers callback
+}
+```
+
+### Why This Works
+
+1. **Clear separation**: `internalValue` stores the data, Vue ref manages the UI
+2. **Built-in callback**: ComfyUI calls `widget.callback()` automatically after `setValue()`
+3. **Bidirectional sync**:
+   - External → UI: `setValue()` updates `internalValue`, `callback()` updates Vue ref
+   - UI → External: User interaction updates Vue ref, which updates `widget.value`
+
+---
+
+## Common Mistakes
+
+### ❌ Creating custom callback mechanisms
+
+```typescript
+// Wrong - unnecessary complexity
+setValue(v: MyConfig) {
+  internalValue = v
+  widget.onSetValue?.(v)  // Don't add this - use widget.callback instead
+}
+```
+
+Use the built-in `widget.callback` instead.
+
+### ❌ Using v-model for simple text inputs in DOM widgets
+
+```html
+<!-- Wrong - creates sync issues -->
+<textarea v-model="textValue" />
+
+<!-- Right for simple text widgets -->
+<textarea ref="textareaRef" @input="onInput" />
+```
+
+### ❌ Watching props.widget.value
+
+```typescript
+// Wrong - creates race conditions
+watch(() => props.widget.value, (newValue) => {
+  config.value = newValue
+})
+```
+
+Use `widget.callback` instead - it's called at the right time in the lifecycle.
+
+### ❌ Multiple sources of truth
+
+```typescript
+// Wrong - who is the source of truth?
+let internalValue = ''        // State 1
+const textValue = ref('')     // State 2
+const domElement = textarea   // State 3
+props.widget.value            // State 4
+```
+
+Choose ONE source of truth:
+- **Simple widgets**: DOM element
+- **Complex widgets**: `internalValue` (with Vue ref as derived UI state)
+
+### ❌ Adding serializeValue for simple widgets
+
+```typescript
+// Wrong - getValue/setValue handle serialization
+props.widget.serializeValue = async () => textValue.value
+```
+
+---
+
+## Decision Guide
+
+| Widget Type | Source of Truth | Use `widget.callback` | Example |
+|-------------|-----------------|----------------------|---------|
+| Simple text input | DOM element (`inputEl`) | Optional | AutocompleteTextWidget |
+| Complex config | `internalValue` | Yes, for UI sync | LoraPoolWidget |
+| Vue component widget | Vue ref + `internalValue` | Yes | RandomizerWidget |
+
+---
+
+## Testing Checklist
+
+- [ ] Load workflow - value restores correctly
+- [ ] Switch workflow - value persists
+- [ ] Reload page - value persists
+- [ ] UI interaction - value updates
+- [ ] Undo/redo - value syncs with UI
+- [ ] No console errors
+
+---
+
+## References
+
+- ComfyUI DOMWidget implementation: `ComfyUI_frontend/src/scripts/domWidget.ts`
+- Simple text widget example: `ComfyUI_frontend/src/renderer/extensions/vueNodes/widgets/composables/useStringWidget.ts`

docs/dom_widget_dev_guide.md

@@ -240,6 +240,8 @@ inputEl.addEventListener("change", () => {
 });
 ```
 
+> **⚠️ Important**: For Vue-based DOM widgets with text inputs, follow the [Value Persistence Best Practices](dom-widgets/value-persistence-best-practices.md) to avoid sync issues. Key takeaway: use DOM element as single source of truth, avoid internal state variables and v-model.
+
 ### 5.3 The Restoration Mechanism (`configure`)
 
 *   **`configure(data)`**: When a Workflow is loaded, `LGraphNode` calls its `configure(data)` method.

docs/reference/danbooru_e621_categories.md

@@ -0,0 +1,69 @@
+# Danbooru/E621 Tag Categories Reference
+
+Reference for category values used in `danbooru_e621_merged.csv` tag files.
+
+## Category Value Mapping
+
+### Danbooru Categories
+
+| Value | Description |
+|-------|-------------|
+| 0 | General |
+| 1 | Artist |
+| 2 | *(unused)* |
+| 3 | Copyright |
+| 4 | Character |
+| 5 | Meta |
+
+### e621 Categories
+
+| Value | Description |
+|-------|-------------|
+| 6 | *(unused)* |
+| 7 | General |
+| 8 | Artist |
+| 9 | Contributor |
+| 10 | Copyright |
+| 11 | Character |
+| 12 | Species |
+| 13 | *(unused)* |
+| 14 | Meta |
+| 15 | Lore |
+
+## Danbooru Category Colors
+
+| Description | Normal Color | Hover Color |
+|-------------|--------------|-------------|
+| General | #009be6 | #4bb4ff |
+| Artist | #ff8a8b | #ffc3c3 |
+| Copyright | #c797ff | #ddc9fb |
+| Character | #35c64a | #93e49a |
+| Meta | #ead084 | #f7e7c3 |
+
+## CSV Column Structure
+
+Each row in the merged CSV file contains 4 columns:
+
+| Column | Description | Example |
+|--------|-------------|---------|
+| 1 | Tag name | `1girl`, `highres`, `solo` |
+| 2 | Category value (0-15) | `0`, `5`, `7` |
+| 3 | Post count | `6008644`, `5256195` |
+| 4 | Aliases (comma-separated, quoted) | `"1girls,sole_female"`, empty string |
+
+### Sample Data
+
+```
+1girl,0,6008644,"1girls,sole_female"
+highres,5,5256195,"high_res,high_resolution,hires"
+solo,0,5000954,"alone,female_solo,single,solo_female"
+long_hair,0,4350743,"/lh,longhair"
+mammal,12,3437444,"cetancodont,cetancodontamorph,feralmammal"
+anthro,7,3381927,"adult_anthro,anhtro,antho,anthro_horse"
+skirt,0,1557883,
+```
+
+## Source
+
+- [PR #312: Add danbooru_e621_merged.csv](https://github.com/DominikDoom/a1111-sd-webui-tagcomplete/pull/312)
+- [DraconicDragon/dbr-e621-lists-archive](https://github.com/DraconicDragon/dbr-e621-lists-archive)

docs/technical/model_type_refactoring_todo.md

@@ -0,0 +1,191 @@
+# Model Type 字段重构 - 遗留工作清单
+
+> **状态**: Phase 1-4 已完成 | **创建日期**: 2026-01-30  
+> **相关文件**: `py/utils/models.py`, `py/services/model_query.py`, `py/services/checkpoint_scanner.py`, etc.
+
+---
+
+## 概述
+
+本次重构旨在解决 `model_type` 字段语义不统一的问题。系统中有两个层面的"类型"概念：
+
+1. **Scanner Type** (`scanner_type`): 架构层面的大类 - `lora`, `checkpoint`, `embedding`
+2. **Sub Type** (`sub_type`): 业务层面的细分类型 - `lora`/`locon`/`dora`, `checkpoint`/`diffusion_model`, `embedding`
+
+重构目标是统一使用 `sub_type` 表示细分类型，保留 `model_type` 作为向后兼容的别名。
+
+---
+
+## 已完成工作 ✅
+
+### Phase 1: 后端字段重命名
+- [x] `CheckpointMetadata.model_type` → `sub_type`
+- [x] `EmbeddingMetadata.model_type` → `sub_type`
+- [x] `model_scanner.py` `_build_cache_entry()` 同时处理 `sub_type` 和 `model_type`
+
+### Phase 2: 查询逻辑更新
+- [x] `model_query.py` 新增 `resolve_sub_type()` 和 `normalize_sub_type()`
+- [x] ~~保持向后兼容的别名 `resolve_civitai_model_type`, `normalize_civitai_model_type`~~ (已在 Phase 5 移除)
+- [x] `ModelFilterSet.apply()` 更新为使用新的解析函数
+
+### Phase 3: API 响应更新
+- [x] `LoraService.format_response()` 返回 `sub_type` ~~+ `model_type`~~ (已移除 `model_type`)
+- [x] `CheckpointService.format_response()` 返回 `sub_type` ~~+ `model_type`~~ (已移除 `model_type`)
+- [x] `EmbeddingService.format_response()` 返回 `sub_type` ~~+ `model_type`~~ (已移除 `model_type`)
+
+### Phase 4: 前端更新
+- [x] `constants.js` 新增 `MODEL_SUBTYPE_DISPLAY_NAMES`
+- [x] `MODEL_TYPE_DISPLAY_NAMES` 作为别名保留
+
+### Phase 5: 清理废弃代码 ✅
+- [x] 从 `ModelScanner._build_cache_entry()` 中移除 `model_type` 向后兼容代码
+- [x] 从 `CheckpointScanner` 中移除 `model_type` 兼容处理
+- [x] 从 `model_query.py` 中移除 `resolve_civitai_model_type` 和 `normalize_civitai_model_type` 别名
+- [x] 更新前端 `FilterManager.js` 使用 `sub_type` (已在使用 `MODEL_SUBTYPE_DISPLAY_NAMES`)
+- [x] 更新所有相关测试
+
+---
+
+## 遗留工作 ⏳
+
+### Phase 5: 清理废弃代码 ✅ **已完成**
+
+所有 Phase 5 的清理工作已完成：
+
+#### 5.1 移除 `model_type` 字段的向后兼容代码 ✅
+- 从 `ModelScanner._build_cache_entry()` 中移除了 `model_type` 的设置
+- 现在只设置 `sub_type` 字段
+
+#### 5.2 移除 CheckpointScanner 的 model_type 兼容处理 ✅
+- `adjust_metadata()` 现在只处理 `sub_type`
+- `adjust_cached_entry()` 现在只设置 `sub_type`
+
+#### 5.3 移除 model_query 中的向后兼容别名 ✅
+- 移除了 `resolve_civitai_model_type = resolve_sub_type`
+- 移除了 `normalize_civitai_model_type = normalize_sub_type`
+
+#### 5.4 前端清理 ✅
+- `FilterManager.js` 已经在使用 `MODEL_SUBTYPE_DISPLAY_NAMES` (通过别名 `MODEL_TYPE_DISPLAY_NAMES`)
+- API list endpoint 现在只返回 `sub_type`，不再返回 `model_type`
+- `ModelCard.js` 现在设置 `card.dataset.sub_type` (所有模型类型通用)
+- `CheckpointContextMenu.js` 现在读取 `card.dataset.sub_type`
+- `MoveManager.js` 现在处理 `cache_entry.sub_type`
+- `RecipeModal.js` 现在读取 `checkpoint.sub_type`
+
+---
+
+## 数据库迁移评估
+
+### 当前状态
+- `persistent_model_cache.py` 使用 `civitai_model_type` 列存储 CivitAI 原始类型
+- 缓存 entry 中的 `sub_type` 在运行期动态计算
+- 数据库 schema **无需立即修改**
+
+### 未来可选优化
+```sql
+-- 可选：在 models 表中添加 sub_type 列（与 civitai_model_type 保持一致但语义更清晰）
+ALTER TABLE models ADD COLUMN sub_type TEXT;
+
+-- 数据迁移
+UPDATE models SET sub_type = civitai_model_type WHERE sub_type IS NULL;
+```
+
+**建议**: 如果决定添加 `sub_type` 列，应与 Phase 5 一起进行。
+
+---
+
+## 测试覆盖率
+
+### 新增/更新测试文件（已全部通过 ✅）
+
+| 测试文件 | 数量 | 覆盖内容 |
+|---------|------|---------|
+| `tests/utils/test_models_sub_type.py` | 7 | Metadata sub_type 字段 |
+| `tests/services/test_model_query_sub_type.py` | 19 | sub_type 解析和过滤 |
+| `tests/services/test_checkpoint_scanner_sub_type.py` | 6 | CheckpointScanner sub_type |
+| `tests/services/test_service_format_response_sub_type.py` | 6 | API 响应 sub_type 包含 |
+| `tests/services/test_checkpoint_scanner.py` | 1 | Checkpoint 缓存 sub_type |
+| `tests/services/test_model_scanner.py` | 1 | adjust_cached_entry hook |
+| `tests/services/test_download_manager.py` | 1 | Checkpoint 下载 sub_type |
+
+### 需要补充的测试（可选）
+
+- [ ] 集成测试：验证前端过滤使用 sub_type 字段
+- [ ] 数据库迁移测试（如果执行可选优化）
+- [ ] 性能测试：确认 resolve_sub_type 的优先级查找没有显著性能影响
+
+---
+
+## 兼容性检查清单
+
+### 已完成 ✅
+
+- [x] 前端代码已全部改用 `sub_type` 字段
+- [x] API list endpoint 已移除 `model_type`，只返回 `sub_type`
+- [x] 后端 cache entry 已移除 `model_type`，只保留 `sub_type`
+- [x] 所有测试已更新通过
+- [x] 文档已更新
+
+---
+
+## 相关文件清单
+
+### 核心文件
+```
+py/utils/models.py
+py/utils/constants.py
+py/services/model_scanner.py
+py/services/model_query.py
+py/services/checkpoint_scanner.py
+py/services/base_model_service.py
+py/services/lora_service.py
+py/services/checkpoint_service.py
+py/services/embedding_service.py
+```
+
+### 前端文件
+```
+static/js/utils/constants.js
+static/js/managers/FilterManager.js
+static/js/managers/MoveManager.js
+static/js/components/shared/ModelCard.js
+static/js/components/ContextMenu/CheckpointContextMenu.js
+static/js/components/RecipeModal.js
+```
+
+### 测试文件
+```
+tests/utils/test_models_sub_type.py
+tests/services/test_model_query_sub_type.py
+tests/services/test_checkpoint_scanner_sub_type.py
+tests/services/test_service_format_response_sub_type.py
+```
+
+---
+
+## 风险评估
+
+| 风险项 | 影响 | 缓解措施 |
+|-------|------|---------|
+| ~~第三方代码依赖 `model_type`~~ | ~~高~~ | ~~保持别名至少 1 个 major 版本~~ ✅ 已完成移除 |
+| ~~数据库 schema 变更~~ | ~~中~~ | ~~暂缓 schema 变更，仅运行时计算~~ ✅ 无需变更 |
+| ~~前端过滤失效~~ | ~~中~~ | ~~全面的集成测试覆盖~~ ✅ 测试通过 |
+| CivitAI API 变化 | 低 | 保持多源解析策略 |
+
+---
+
+## 时间线
+
+- **v1.x**: Phase 1-4 已完成，保持向后兼容
+- **v2.0 (当前)**: ✅ Phase 5 已完成 - `model_type` 兼容代码已移除
+  - API list endpoint 只返回 `sub_type`
+  - Cache entry 只保留 `sub_type`
+  - 移除了 `resolve_civitai_model_type` 和 `normalize_civitai_model_type` 别名
+
+---
+
+## 备注
+
+- 重构期间发现 `civitai_model_type` 数据库列命名尚可，但语义上应理解为存储 CivitAI API 返回的原始类型值
+- Checkpoint 的 `diffusion_model` sub_type 不能通过 CivitAI API 获取，必须通过文件路径（model root）判断
+- LoRA 的 sub_type（lora/locon/dora）直接来自 CivitAI API 的 `version_info.model.type`

locales/de.json

@@ -10,7 +10,8 @@
             "next": "Weiter",
             "backToTop": "Nach oben",
             "settings": "Einstellungen",
-            "help": "Hilfe"
+            "help": "Hilfe",
+            "add": "Hinzufügen"
         },
         "status": {
             "loading": "Wird geladen...",
@@ -204,6 +205,17 @@
         },
         "filter": {
             "title": "Modelle filtern",
+            "presets": "Voreinstellungen",
+            "savePreset": "Aktive Filter als neue Voreinstellung speichern.",
+            "savePresetDisabledActive": "Speichern nicht möglich: Eine Voreinstellung ist bereits aktiv. Ändern Sie die Filter, um eine neue Voreinstellung zu speichern",
+            "savePresetDisabledNoFilters": "Wählen Sie zuerst Filter aus, um als Voreinstellung zu speichern",
+            "savePresetPrompt": "Voreinstellungsname eingeben:",
+            "presetClickTooltip": "Voreinstellung \"{name}\" anwenden",
+            "presetDeleteTooltip": "Voreinstellung löschen",
+            "presetDeleteConfirm": "Voreinstellung \"{name}\" löschen?",
+            "presetDeleteConfirmClick": "Zum Bestätigen erneut klicken",
+            "presetOverwriteConfirm": "Voreinstellung \"{name}\" existiert bereits. Überschreiben?",
+            "presetNamePlaceholder": "Voreinstellungsname...",
             "baseModel": "Basis-Modell",
             "modelTags": "Tags (Top 20)",
             "modelTypes": "Model Types",
@@ -1165,7 +1177,11 @@
         "exampleImages": {
             "opened": "Beispielbilder-Ordner geöffnet",
             "openingFolder": "Beispielbilder-Ordner wird geöffnet",
-            "failedToOpen": "Fehler beim Öffnen des Beispielbilder-Ordners"
+            "failedToOpen": "Fehler beim Öffnen des Beispielbilder-Ordners",
+            "setupRequired": "Beispielbilder-Speicher",
+            "setupDescription": "Um benutzerdefinierte Beispielbilder hinzuzufügen, müssen Sie zuerst einen Download-Speicherort festlegen.",
+            "setupUsage": "Dieser Pfad wird sowohl für heruntergeladene als auch für benutzerdefinierte Beispielbilder verwendet.",
+            "openSettings": "Einstellungen öffnen"
         }
     },
     "help": {
@@ -1214,6 +1230,7 @@
         "checkingUpdates": "Nach Updates wird gesucht...",
         "checkingMessage": "Bitte warten Sie, während wir nach der neuesten Version suchen.",
         "showNotifications": "Update-Benachrichtigungen anzeigen",
+        "latestBadge": "Neueste",
         "updateProgress": {
             "preparing": "Update wird vorbereitet...",
             "installing": "Update wird installiert...",
@@ -1414,7 +1431,26 @@
         "filters": {
             "applied": "{message}",
             "cleared": "Filter gelöscht",
-            "noCustomFilterToClear": "Kein benutzerdefinierter Filter zum Löschen"
+            "noCustomFilterToClear": "Kein benutzerdefinierter Filter zum Löschen",
+            "noActiveFilters": "Keine aktiven Filter zum Speichern"
+        },
+        "presets": {
+            "created": "Voreinstellung \"{name}\" erstellt",
+            "deleted": "Voreinstellung \"{name}\" gelöscht",
+            "applied": "Voreinstellung \"{name}\" angewendet",
+            "overwritten": "Voreinstellung \"{name}\" überschrieben",
+            "restored": "Standard-Voreinstellungen wiederhergestellt"
+        },
+        "error": {
+            "presetNameEmpty": "Voreinstellungsname darf nicht leer sein",
+            "presetNameTooLong": "Voreinstellungsname darf maximal {max} Zeichen haben",
+            "presetNameInvalidChars": "Voreinstellungsname enthält ungültige Zeichen",
+            "presetNameExists": "Eine Voreinstellung mit diesem Namen existiert bereits",
+            "maxPresetsReached": "Maximal {max} Voreinstellungen erlaubt. Löschen Sie eine, um weitere hinzuzufügen.",
+            "presetNotFound": "Voreinstellung nicht gefunden",
+            "invalidPreset": "Ungültige Voreinstellungsdaten",
+            "deletePresetFailed": "Fehler beim Löschen der Voreinstellung",
+            "applyPresetFailed": "Fehler beim Anwenden der Voreinstellung"
         },
         "downloads": {
             "imagesCompleted": "Beispielbilder {action} abgeschlossen",
@@ -1536,6 +1572,20 @@
             "content": "LoRA Manager is a passion project maintained full-time by a solo developer. Your support on Ko-fi helps cover development costs, keeps new updates coming, and unlocks a license key for the LM Civitai Extension as a thank-you gift. Every contribution truly makes a difference.",
             "supportCta": "Support on Ko-fi",
             "learnMore": "LM Civitai Extension Tutorial"
+        },
+        "cacheHealth": {
+            "corrupted": {
+                "title": "Cache-Korruption erkannt"
+            },
+            "degraded": {
+                "title": "Cache-Probleme erkannt"
+            },
+            "content": "{invalid} von {total} Cache-Einträgen sind ungültig ({rate}). Dies kann zu fehlenden Modellen oder Fehlern führen. Ein Neuaufbau des Caches wird empfohlen.",
+            "rebuildCache": "Cache neu aufbauen",
+            "dismiss": "Verwerfen",
+            "rebuilding": "Cache wird neu aufgebaut...",
+            "rebuildFailed": "Fehler beim Neuaufbau des Caches: {error}",
+            "retry": "Wiederholen"
         }
     }
 }

locales/en.json

@@ -10,7 +10,8 @@
             "next": "Next",
             "backToTop": "Back to top",
             "settings": "Settings",
-            "help": "Help"
+            "help": "Help",
+            "add": "Add"
         },
         "status": {
             "loading": "Loading...",
@@ -204,6 +205,17 @@
         },
         "filter": {
             "title": "Filter Models",
+            "presets": "Presets",
+            "savePreset": "Save current active filters as a new preset.",
+            "savePresetDisabledActive": "Cannot save: A preset is already active. Modify filters to save new preset.",
+            "savePresetDisabledNoFilters": "Select filters first to save as preset",
+            "savePresetPrompt": "Enter preset name:",
+            "presetClickTooltip": "Click to apply preset \"{name}\"",
+            "presetDeleteTooltip": "Delete preset",
+            "presetDeleteConfirm": "Delete preset \"{name}\"?",
+            "presetDeleteConfirmClick": "Click again to confirm",
+            "presetOverwriteConfirm": "Preset \"{name}\" already exists. Overwrite?",
+            "presetNamePlaceholder": "Preset name...",
             "baseModel": "Base Model",
             "modelTags": "Tags (Top 20)",
             "modelTypes": "Model Types",
@@ -1165,7 +1177,11 @@
         "exampleImages": {
             "opened": "Example images folder opened",
             "openingFolder": "Opening example images folder",
-            "failedToOpen": "Failed to open example images folder"
+            "failedToOpen": "Failed to open example images folder",
+            "setupRequired": "Example Images Storage",
+            "setupDescription": "To add custom example images, you need to set a download location first.",
+            "setupUsage": "This path is used for both downloaded and custom example images.",
+            "openSettings": "Open Settings"
         }
     },
     "help": {
@@ -1214,6 +1230,7 @@
         "checkingUpdates": "Checking for updates...",
         "checkingMessage": "Please wait while we check for the latest version.",
         "showNotifications": "Show update notifications",
+        "latestBadge": "Latest",
         "updateProgress": {
             "preparing": "Preparing update...",
             "installing": "Installing update...",
@@ -1414,7 +1431,26 @@
         "filters": {
             "applied": "{message}",
             "cleared": "Filters cleared",
-            "noCustomFilterToClear": "No custom filter to clear"
+            "noCustomFilterToClear": "No custom filter to clear",
+            "noActiveFilters": "No active filters to save"
+        },
+        "presets": {
+            "created": "Preset \"{name}\" created",
+            "deleted": "Preset \"{name}\" deleted",
+            "applied": "Preset \"{name}\" applied",
+            "overwritten": "Preset \"{name}\" overwritten",
+            "restored": "Default presets restored"
+        },
+        "error": {
+            "presetNameEmpty": "Preset name cannot be empty",
+            "presetNameTooLong": "Preset name must be {max} characters or less",
+            "presetNameInvalidChars": "Preset name contains invalid characters",
+            "presetNameExists": "A preset with this name already exists",
+            "maxPresetsReached": "Maximum {max} presets allowed. Delete one to add more.",
+            "presetNotFound": "Preset not found",
+            "invalidPreset": "Invalid preset data",
+            "deletePresetFailed": "Failed to delete preset",
+            "applyPresetFailed": "Failed to apply preset"
         },
         "downloads": {
             "imagesCompleted": "Example images {action} completed",
@@ -1536,6 +1572,20 @@
             "content": "LoRA Manager is a passion project maintained full-time by a solo developer. Your support on Ko-fi helps cover development costs, keeps new updates coming, and unlocks a license key for the LM Civitai Extension as a thank-you gift. Every contribution truly makes a difference.",
             "supportCta": "Support on Ko-fi",
             "learnMore": "LM Civitai Extension Tutorial"
+        },
+        "cacheHealth": {
+            "corrupted": {
+                "title": "Cache Corruption Detected"
+            },
+            "degraded": {
+                "title": "Cache Issues Detected"
+            },
+            "content": "{invalid} of {total} cache entries are invalid ({rate}). This may cause missing models or errors. Rebuilding the cache is recommended.",
+            "rebuildCache": "Rebuild Cache",
+            "dismiss": "Dismiss",
+            "rebuilding": "Rebuilding cache...",
+            "rebuildFailed": "Failed to rebuild cache: {error}",
+            "retry": "Retry"
         }
     }
 }

locales/es.json

@@ -10,7 +10,8 @@
             "next": "Siguiente",
             "backToTop": "Volver arriba",
             "settings": "Configuración",
-            "help": "Ayuda"
+            "help": "Ayuda",
+            "add": "Añadir"
         },
         "status": {
             "loading": "Cargando...",
@@ -204,6 +205,17 @@
         },
         "filter": {
             "title": "Filtrar modelos",
+            "presets": "Preajustes",
+            "savePreset": "Guardar filtros activos como nuevo preajuste.",
+            "savePresetDisabledActive": "No se puede guardar: Ya hay un preajuste activo. Modifique los filtros para guardar un nuevo preajuste",
+            "savePresetDisabledNoFilters": "Seleccione filtros primero para guardar como preajuste",
+            "savePresetPrompt": "Ingrese el nombre del preajuste:",
+            "presetClickTooltip": "Hacer clic para aplicar preajuste \"{name}\"",
+            "presetDeleteTooltip": "Eliminar preajuste",
+            "presetDeleteConfirm": "¿Eliminar preajuste \"{name}\"?",
+            "presetDeleteConfirmClick": "Haga clic de nuevo para confirmar",
+            "presetOverwriteConfirm": "El preset \"{name}\" ya existe. ¿Sobrescribir?",
+            "presetNamePlaceholder": "Nombre del preajuste...",
             "baseModel": "Modelo base",
             "modelTags": "Etiquetas (Top 20)",
             "modelTypes": "Model Types",
@@ -1165,7 +1177,11 @@
         "exampleImages": {
             "opened": "Carpeta de imágenes de ejemplo abierta",
             "openingFolder": "Abriendo carpeta de imágenes de ejemplo",
-            "failedToOpen": "Error al abrir carpeta de imágenes de ejemplo"
+            "failedToOpen": "Error al abrir carpeta de imágenes de ejemplo",
+            "setupRequired": "Almacenamiento de imágenes de ejemplo",
+            "setupDescription": "Para agregar imágenes de ejemplo personalizadas, primero necesita establecer una ubicación de descarga.",
+            "setupUsage": "Esta ruta se utiliza tanto para imágenes de ejemplo descargadas como personalizadas.",
+            "openSettings": "Abrir configuración"
         }
     },
     "help": {
@@ -1214,6 +1230,7 @@
         "checkingUpdates": "Comprobando actualizaciones...",
         "checkingMessage": "Por favor espera mientras comprobamos la última versión.",
         "showNotifications": "Mostrar notificaciones de actualización",
+        "latestBadge": "Último",
         "updateProgress": {
             "preparing": "Preparando actualización...",
             "installing": "Instalando actualización...",
@@ -1414,7 +1431,26 @@
         "filters": {
             "applied": "{message}",
             "cleared": "Filtros limpiados",
-            "noCustomFilterToClear": "No hay filtro personalizado para limpiar"
+            "noCustomFilterToClear": "No hay filtro personalizado para limpiar",
+            "noActiveFilters": "No hay filtros activos para guardar"
+        },
+        "presets": {
+            "created": "Preajuste \"{name}\" creado",
+            "deleted": "Preajuste \"{name}\" eliminado",
+            "applied": "Preajuste \"{name}\" aplicado",
+            "overwritten": "Preset \"{name}\" sobrescrito",
+            "restored": "Presets predeterminados restaurados"
+        },
+        "error": {
+            "presetNameEmpty": "El nombre del preajuste no puede estar vacío",
+            "presetNameTooLong": "El nombre del preajuste debe tener {max} caracteres o menos",
+            "presetNameInvalidChars": "El nombre del preajuste contiene caracteres inválidos",
+            "presetNameExists": "Ya existe un preajuste con este nombre",
+            "maxPresetsReached": "Máximo {max} preajustes permitidos. Elimine uno para agregar más.",
+            "presetNotFound": "Preajuste no encontrado",
+            "invalidPreset": "Datos de preajuste inválidos",
+            "deletePresetFailed": "Error al eliminar el preajuste",
+            "applyPresetFailed": "Error al aplicar el preajuste"
         },
         "downloads": {
             "imagesCompleted": "Imágenes de ejemplo {action} completadas",
@@ -1536,6 +1572,20 @@
             "content": "LoRA Manager is a passion project maintained full-time by a solo developer. Your support on Ko-fi helps cover development costs, keeps new updates coming, and unlocks a license key for the LM Civitai Extension as a thank-you gift. Every contribution truly makes a difference.",
             "supportCta": "Support on Ko-fi",
             "learnMore": "LM Civitai Extension Tutorial"
+        },
+        "cacheHealth": {
+            "corrupted": {
+                "title": "Corrupción de caché detectada"
+            },
+            "degraded": {
+                "title": "Problemas de caché detectados"
+            },
+            "content": "{invalid} de {total} entradas de caché son inválidas ({rate}). Esto puede causar modelos faltantes o errores. Se recomienda reconstruir la caché.",
+            "rebuildCache": "Reconstruir caché",
+            "dismiss": "Descartar",
+            "rebuilding": "Reconstruyendo caché...",
+            "rebuildFailed": "Error al reconstruir la caché: {error}",
+            "retry": "Reintentar"
         }
     }
 }

locales/fr.json

@@ -10,7 +10,8 @@
             "next": "Suivant",
             "backToTop": "Retour en haut",
             "settings": "Paramètres",
-            "help": "Aide"
+            "help": "Aide",
+            "add": "Ajouter"
         },
         "status": {
             "loading": "Chargement...",
@@ -204,6 +205,17 @@
         },
         "filter": {
             "title": "Filtrer les modèles",
+            "presets": "Préréglages",
+            "savePreset": "Enregistrer les filtres actifs comme nouveau préréglage.",
+            "savePresetDisabledActive": "Impossible d'enregistrer : Un préréglage est déjà actif. Modifiez les filtres pour enregistrer un nouveau préréglage",
+            "savePresetDisabledNoFilters": "Sélectionnez d'abord des filtres à enregistrer comme préréglage",
+            "savePresetPrompt": "Entrez le nom du préréglage :",
+            "presetClickTooltip": "Cliquer pour appliquer le préréglage \"{name}\"",
+            "presetDeleteTooltip": "Supprimer le préréglage",
+            "presetDeleteConfirm": "Supprimer le préréglage \"{name}\" ?",
+            "presetDeleteConfirmClick": "Cliquez à nouveau pour confirmer",
+            "presetOverwriteConfirm": "Le préréglage \"{name}\" existe déjà. Remplacer?",
+            "presetNamePlaceholder": "Nom du préréglage...",
             "baseModel": "Modèle de base",
             "modelTags": "Tags (Top 20)",
             "modelTypes": "Model Types",
@@ -1165,7 +1177,11 @@
         "exampleImages": {
             "opened": "Dossier d'images d'exemple ouvert",
             "openingFolder": "Ouverture du dossier d'images d'exemple",
-            "failedToOpen": "Échec de l'ouverture du dossier d'images d'exemple"
+            "failedToOpen": "Échec de l'ouverture du dossier d'images d'exemple",
+            "setupRequired": "Stockage d'images d'exemple",
+            "setupDescription": "Pour ajouter des images d'exemple personnalisées, vous devez d'abord définir un emplacement de téléchargement.",
+            "setupUsage": "Ce chemin est utilisé pour les images d'exemple téléchargées et personnalisées.",
+            "openSettings": "Ouvrir les paramètres"
         }
     },
     "help": {
@@ -1214,6 +1230,7 @@
         "checkingUpdates": "Vérification des mises à jour...",
         "checkingMessage": "Veuillez patienter pendant la vérification de la dernière version.",
         "showNotifications": "Afficher les notifications de mise à jour",
+        "latestBadge": "Dernier",
         "updateProgress": {
             "preparing": "Préparation de la mise à jour...",
             "installing": "Installation de la mise à jour...",
@@ -1414,7 +1431,26 @@
         "filters": {
             "applied": "{message}",
             "cleared": "Filtres effacés",
-            "noCustomFilterToClear": "Aucun filtre personnalisé à effacer"
+            "noCustomFilterToClear": "Aucun filtre personnalisé à effacer",
+            "noActiveFilters": "Aucun filtre actif à enregistrer"
+        },
+        "presets": {
+            "created": "Préréglage \"{name}\" créé",
+            "deleted": "Préréglage \"{name}\" supprimé",
+            "applied": "Préréglage \"{name}\" appliqué",
+            "overwritten": "Préréglage \"{name}\" remplacé",
+            "restored": "Paramètres par défaut restaurés"
+        },
+        "error": {
+            "presetNameEmpty": "Le nom du préréglage ne peut pas être vide",
+            "presetNameTooLong": "Le nom du préréglage doit contenir au maximum {max} caractères",
+            "presetNameInvalidChars": "Le nom du préréglage contient des caractères invalides",
+            "presetNameExists": "Un préréglage avec ce nom existe déjà",
+            "maxPresetsReached": "Maximum {max} préréglages autorisés. Supprimez-en un pour en ajouter plus.",
+            "presetNotFound": "Préréglage non trouvé",
+            "invalidPreset": "Données de préréglage invalides",
+            "deletePresetFailed": "Échec de la suppression du préréglage",
+            "applyPresetFailed": "Échec de l'application du préréglage"
         },
         "downloads": {
             "imagesCompleted": "Images d'exemple {action} terminées",
@@ -1536,6 +1572,20 @@
             "content": "LoRA Manager is a passion project maintained full-time by a solo developer. Your support on Ko-fi helps cover development costs, keeps new updates coming, and unlocks a license key for the LM Civitai Extension as a thank-you gift. Every contribution truly makes a difference.",
             "supportCta": "Support on Ko-fi",
             "learnMore": "LM Civitai Extension Tutorial"
+        },
+        "cacheHealth": {
+            "corrupted": {
+                "title": "Corruption du cache détectée"
+            },
+            "degraded": {
+                "title": "Problèmes de cache détectés"
+            },
+            "content": "{invalid} des {total} entrées de cache sont invalides ({rate}). Cela peut provoquer des modèles manquants ou des erreurs. Il est recommandé de reconstruire le cache.",
+            "rebuildCache": "Reconstruire le cache",
+            "dismiss": "Ignorer",
+            "rebuilding": "Reconstruction du cache...",
+            "rebuildFailed": "Échec de la reconstruction du cache : {error}",
+            "retry": "Réessayer"
         }
     }
 }

locales/he.json

@@ -10,7 +10,8 @@
             "next": "הבא",
             "backToTop": "חזור למעלה",
             "settings": "הגדרות",
-            "help": "עזרה"
+            "help": "עזרה",
+            "add": "הוסף"
         },
         "status": {
             "loading": "טוען...",
@@ -204,6 +205,17 @@
         },
         "filter": {
             "title": "סנן מודלים",
+            "presets": "קביעות מראש",
+            "savePreset": "שמור מסננים פעילים כקביעה מראש חדשה.",
+            "savePresetDisabledActive": "לא ניתן לשמור: קביעה מראש כבר פעילה. שנה מסננים כדי לשמור קביעה מראש חדשה",
+            "savePresetDisabledNoFilters": "בחר מסננים תחילה כדי לשמור כקביעה מראש",
+            "savePresetPrompt": "הזן שם קביעה מראש:",
+            "presetClickTooltip": "לחץ כדי להפעיל קביעה מראש \"{name}\"",
+            "presetDeleteTooltip": "מחק קביעה מראש",
+            "presetDeleteConfirm": "למחוק קביעה מראש \"{name}\"?",
+            "presetDeleteConfirmClick": "לחץ שוב לאישור",
+            "presetOverwriteConfirm": "הפריסט \"{name}\" כבר קיים. לדרוס?",
+            "presetNamePlaceholder": "שם קביעה מראש...",
             "baseModel": "מודל בסיס",
             "modelTags": "תגיות (20 המובילות)",
             "modelTypes": "Model Types",
@@ -1165,7 +1177,11 @@
         "exampleImages": {
             "opened": "תיקיית תמונות הדוגמה נפתחה",
             "openingFolder": "פותח תיקיית תמונות דוגמה",
-            "failedToOpen": "פתיחת תיקיית תמונות הדוגמה נכשלה"
+            "failedToOpen": "פתיחת תיקיית תמונות הדוגמה נכשלה",
+            "setupRequired": "אחסון תמונות דוגמה",
+            "setupDescription": "כדי להוסיף תמונות דוגמה מותאמות אישית, עליך קודם להגדיר מיקום הורדה.",
+            "setupUsage": "נתיב זה משמש הן עבור תמונות דוגמה שהורדו והן עבור תמונות מותאמות אישית.",
+            "openSettings": "פתח הגדרות"
         }
     },
     "help": {
@@ -1214,6 +1230,7 @@
         "checkingUpdates": "בודק עדכונים...",
         "checkingMessage": "אנא המתן בזמן שאנו בודקים את הגרסה האחרונה.",
         "showNotifications": "הצג התראות עדכון",
+        "latestBadge": "עדכן",
         "updateProgress": {
             "preparing": "מכין עדכון...",
             "installing": "מתקין עדכון...",
@@ -1414,7 +1431,26 @@
         "filters": {
             "applied": "{message}",
             "cleared": "המסננים נוקו",
-            "noCustomFilterToClear": "אין מסנן מותאם אישית לניקוי"
+            "noCustomFilterToClear": "אין מסנן מותאם אישית לניקוי",
+            "noActiveFilters": "אין מסננים פעילים לשמירה"
+        },
+        "presets": {
+            "created": "קביעה מראש \"{name}\" נוצרה",
+            "deleted": "קביעה מראש \"{name}\" נמחקה",
+            "applied": "קביעה מראש \"{name}\" הופעלה",
+            "overwritten": "קביעה מראש \"{name}\" נדרסה",
+            "restored": "ברירות המחדל שוחזרו"
+        },
+        "error": {
+            "presetNameEmpty": "שם קביעה מראש לא יכול להיות ריק",
+            "presetNameTooLong": "שם קביעה מראש חייב להיות {max} תווים או פחות",
+            "presetNameInvalidChars": "שם קביעה מראש מכיל תווים לא חוקיים",
+            "presetNameExists": "קביעה מראש עם שם זה כבר קיימת",
+            "maxPresetsReached": "מותר מקסימום {max} קביעות מראש. מחק אחת כדי להוסיף עוד.",
+            "presetNotFound": "קביעה מראש לא נמצאה",
+            "invalidPreset": "נתוני קביעה מראש לא חוקיים",
+            "deletePresetFailed": "מחיקת קביעה מראש נכשלה",
+            "applyPresetFailed": "הפעלת קביעה מראש נכשלה"
         },
         "downloads": {
             "imagesCompleted": "{action} תמונות הדוגמה הושלם",
@@ -1536,6 +1572,20 @@
             "content": "LoRA Manager is a passion project maintained full-time by a solo developer. Your support on Ko-fi helps cover development costs, keeps new updates coming, and unlocks a license key for the LM Civitai Extension as a thank-you gift. Every contribution truly makes a difference.",
             "supportCta": "Support on Ko-fi",
             "learnMore": "LM Civitai Extension Tutorial"
+        },
+        "cacheHealth": {
+            "corrupted": {
+                "title": "זוהתה שחיתות במטמון"
+            },
+            "degraded": {
+                "title": "זוהו בעיות במטמון"
+            },
+            "content": "{invalid} מתוך {total} רשומות מטמון אינן תקינות ({rate}). זה עלול לגרום לדגמים חסרים או לשגיאות. מומלץ לבנות מחדש את המטמון.",
+            "rebuildCache": "בניית מטמון מחדש",
+            "dismiss": "ביטול",
+            "rebuilding": "בונה מחדש את המטמון...",
+            "rebuildFailed": "נכשלה בניית המטמון מחדש: {error}",
+            "retry": "נסה שוב"
         }
     }
 }

locales/ja.json

@@ -10,7 +10,8 @@
             "next": "次へ",
             "backToTop": "トップに戻る",
             "settings": "設定",
-            "help": "ヘルプ"
+            "help": "ヘルプ",
+            "add": "追加"
         },
         "status": {
             "loading": "読み込み中...",
@@ -204,6 +205,17 @@
         },
         "filter": {
             "title": "モデルをフィルタ",
+            "presets": "プリセット",
+            "savePreset": "現在のアクティブフィルタを新しいプリセットとして保存。",
+            "savePresetDisabledActive": "保存できません：プリセットがすでにアクティブです。フィルタを変更して新しいプリセットを保存してください",
+            "savePresetDisabledNoFilters": "先にフィルタを選択してからプリセットとして保存",
+            "savePresetPrompt": "プリセット名を入力：",
+            "presetClickTooltip": "プリセット \"{name}\" を適用するにはクリック",
+            "presetDeleteTooltip": "プリセットを削除",
+            "presetDeleteConfirm": "プリセット \"{name}\" を削除しますか？",
+            "presetDeleteConfirmClick": "もう一度クリックして確認",
+            "presetOverwriteConfirm": "プリセット「{name}」は既に存在します。上書きしますか？",
+            "presetNamePlaceholder": "プリセット名...",
             "baseModel": "ベースモデル",
             "modelTags": "タグ（上位20）",
             "modelTypes": "Model Types",
@@ -1165,7 +1177,11 @@
         "exampleImages": {
             "opened": "例画像フォルダが開かれました",
             "openingFolder": "例画像フォルダを開いています",
-            "failedToOpen": "例画像フォルダを開くのに失敗しました"
+            "failedToOpen": "例画像フォルダを開くのに失敗しました",
+            "setupRequired": "例画像ストレージ",
+            "setupDescription": "カスタム例画像を追加するには、まずダウンロード場所を設定する必要があります。",
+            "setupUsage": "このパスは、ダウンロードした例画像とカスタム画像の両方に使用されます。",
+            "openSettings": "設定を開く"
         }
     },
     "help": {
@@ -1214,6 +1230,7 @@
         "checkingUpdates": "更新を確認中...",
         "checkingMessage": "最新バージョンを確認しています。お待ちください。",
         "showNotifications": "更新通知を表示",
+        "latestBadge": "最新",
         "updateProgress": {
             "preparing": "更新を準備中...",
             "installing": "更新をインストール中...",
@@ -1414,7 +1431,26 @@
         "filters": {
             "applied": "{message}",
             "cleared": "フィルタがクリアされました",
-            "noCustomFilterToClear": "クリアするカスタムフィルタがありません"
+            "noCustomFilterToClear": "クリアするカスタムフィルタがありません",
+            "noActiveFilters": "保存するアクティブフィルタがありません"
+        },
+        "presets": {
+            "created": "プリセット \"{name}\" が作成されました",
+            "deleted": "プリセット \"{name}\" が削除されました",
+            "applied": "プリセット \"{name}\" が適用されました",
+            "overwritten": "プリセット「{name}」を上書きしました",
+            "restored": "デフォルトのプリセットを復元しました"
+        },
+        "error": {
+            "presetNameEmpty": "プリセット名を入力してください",
+            "presetNameTooLong": "プリセット名は{max}文字以内にしてください",
+            "presetNameInvalidChars": "プリセット名に使用できない文字が含まれています",
+            "presetNameExists": "同じ名前のプリセットが既に存在します",
+            "maxPresetsReached": "プリセットは最大{max}個までです。追加するには既存のものを削除してください。",
+            "presetNotFound": "プリセットが見つかりません",
+            "invalidPreset": "無効なプリセットデータです",
+            "deletePresetFailed": "プリセットの削除に失敗しました",
+            "applyPresetFailed": "プリセットの適用に失敗しました"
         },
         "downloads": {
             "imagesCompleted": "例画像 {action} が完了しました",
@@ -1536,6 +1572,20 @@
             "content": "LoRA Manager is a passion project maintained full-time by a solo developer. Your support on Ko-fi helps cover development costs, keeps new updates coming, and unlocks a license key for the LM Civitai Extension as a thank-you gift. Every contribution truly makes a difference.",
             "supportCta": "Support on Ko-fi",
             "learnMore": "LM Civitai Extension Tutorial"
+        },
+        "cacheHealth": {
+            "corrupted": {
+                "title": "キャッシュの破損が検出されました"
+            },
+            "degraded": {
+                "title": "キャッシュの問題が検出されました"
+            },
+            "content": "{total}個のキャッシュエントリのうち{invalid}個が無効です（{rate}）。モデルが見つからない原因になったり、エラーが発生する可能性があります。キャッシュの再構築を推奨します。",
+            "rebuildCache": "キャッシュを再構築",
+            "dismiss": "閉じる",
+            "rebuilding": "キャッシュを再構築中...",
+            "rebuildFailed": "キャッシュの再構築に失敗しました: {error}",
+            "retry": "再試行"
         }
     }
 }

locales/ko.json

@@ -10,7 +10,8 @@
             "next": "다음",
             "backToTop": "맨 위로",
             "settings": "설정",
-            "help": "도움말"
+            "help": "도움말",
+            "add": "추가"
         },
         "status": {
             "loading": "로딩 중...",
@@ -204,6 +205,17 @@
         },
         "filter": {
             "title": "모델 필터",
+            "presets": "프리셋",
+            "savePreset": "현재 활성 필터를 새 프리셋으로 저장.",
+            "savePresetDisabledActive": "저장할 수 없음: 프리셋이 이미 활성화되어 있습니다. 필터를 수정한 후 새 프리셋을 저장하세요",
+            "savePresetDisabledNoFilters": "먼저 필터를 선택한 후 프리셋으로 저장",
+            "savePresetPrompt": "프리셋 이름 입력:",
+            "presetClickTooltip": "프리셋 \"{name}\" 적용하려면 클릭",
+            "presetDeleteTooltip": "프리셋 삭제",
+            "presetDeleteConfirm": "프리셋 \"{name}\" 삭제하시겠습니까?",
+            "presetDeleteConfirmClick": "다시 클릭하여 확인",
+            "presetOverwriteConfirm": "프리셋 \"{name}\"이(가) 이미 존재합니다. 덮어쓰시겠습니까?",
+            "presetNamePlaceholder": "프리셋 이름...",
             "baseModel": "베이스 모델",
             "modelTags": "태그 (상위 20개)",
             "modelTypes": "Model Types",
@@ -1165,7 +1177,11 @@
         "exampleImages": {
             "opened": "예시 이미지 폴더가 열렸습니다",
             "openingFolder": "예시 이미지 폴더를 여는 중",
-            "failedToOpen": "예시 이미지 폴더 열기 실패"
+            "failedToOpen": "예시 이미지 폴더 열기 실패",
+            "setupRequired": "예시 이미지 저장소",
+            "setupDescription": "사용자 지정 예시 이미지를 추가하려면 먼저 다운로드 위치를 설정해야 합니다.",
+            "setupUsage": "이 경로는 다운로드한 예시 이미지와 사용자 지정 이미지 모두에 사용됩니다.",
+            "openSettings": "설정 열기"
         }
     },
     "help": {
@@ -1214,6 +1230,7 @@
         "checkingUpdates": "업데이트 확인 중...",
         "checkingMessage": "최신 버전을 확인하는 동안 잠시 기다려주세요.",
         "showNotifications": "업데이트 알림 표시",
+        "latestBadge": "최신",
         "updateProgress": {
             "preparing": "업데이트 준비 중...",
             "installing": "업데이트 설치 중...",
@@ -1414,7 +1431,26 @@
         "filters": {
             "applied": "{message}",
             "cleared": "필터가 지워졌습니다",
-            "noCustomFilterToClear": "지울 사용자 정의 필터가 없습니다"
+            "noCustomFilterToClear": "지울 사용자 정의 필터가 없습니다",
+            "noActiveFilters": "저장할 활성 필터가 없습니다"
+        },
+        "presets": {
+            "created": "프리셋 \"{name}\" 생성됨",
+            "deleted": "프리셋 \"{name}\" 삭제됨",
+            "applied": "프리셋 \"{name}\" 적용됨",
+            "overwritten": "프리셋 \"{name}\" 덮어쓰기 완료",
+            "restored": "기본 프리셋 복원 완료"
+        },
+        "error": {
+            "presetNameEmpty": "프리셋 이름을 입력하세요",
+            "presetNameTooLong": "프리셋 이름은 {max}자 이하여야 합니다",
+            "presetNameInvalidChars": "프리셋 이름에 유효하지 않은 문자가 포함되어 있습니다",
+            "presetNameExists": "동일한 이름의 프리셋이 이미 존재합니다",
+            "maxPresetsReached": "최대 {max}개의 프리셋만 허용됩니다. 더 추가하려면 기존 것을 삭제하세요.",
+            "presetNotFound": "프리셋을 찾을 수 없습니다",
+            "invalidPreset": "잘못된 프리셋 데이터입니다",
+            "deletePresetFailed": "프리셋 삭제에 실패했습니다",
+            "applyPresetFailed": "프리셋 적용에 실패했습니다"
         },
         "downloads": {
             "imagesCompleted": "예시 이미지 {action}이(가) 완료되었습니다",
@@ -1536,6 +1572,20 @@
             "content": "LoRA Manager is a passion project maintained full-time by a solo developer. Your support on Ko-fi helps cover development costs, keeps new updates coming, and unlocks a license key for the LM Civitai Extension as a thank-you gift. Every contribution truly makes a difference.",
             "supportCta": "Support on Ko-fi",
             "learnMore": "LM Civitai Extension Tutorial"
+        },
+        "cacheHealth": {
+            "corrupted": {
+                "title": "캐시 손상이 감지되었습니다"
+            },
+            "degraded": {
+                "title": "캐시 문제가 감지되었습니다"
+            },
+            "content": "{total}개의 캐시 항목 중 {invalid}개가 유효하지 않습니다 ({rate}). 모델 누락이나 오류가 발생할 수 있습니다. 캐시를 재구축하는 것이 좋습니다.",
+            "rebuildCache": "캐시 재구축",
+            "dismiss": "무시",
+            "rebuilding": "캐시 재구축 중...",
+            "rebuildFailed": "캐시 재구축 실패: {error}",
+            "retry": "다시 시도"
         }
     }
 }

locales/ru.json

@@ -10,7 +10,8 @@
             "next": "Далее",
             "backToTop": "Наверх",
             "settings": "Настройки",
-            "help": "Справка"
+            "help": "Справка",
+            "add": "Добавить"
         },
         "status": {
             "loading": "Загрузка...",
@@ -204,6 +205,17 @@
         },
         "filter": {
             "title": "Фильтр моделей",
+            "presets": "Пресеты",
+            "savePreset": "Сохранить текущие активные фильтры как новый пресет.",
+            "savePresetDisabledActive": "Невозможно сохранить: Пресет уже активен. Измените фильтры, чтобы сохранить новый пресет",
+            "savePresetDisabledNoFilters": "Сначала выберите фильтры для сохранения как пресет",
+            "savePresetPrompt": "Введите имя пресета:",
+            "presetClickTooltip": "Нажмите чтобы применить пресет \"{name}\"",
+            "presetDeleteTooltip": "Удалить пресет",
+            "presetDeleteConfirm": "Удалить пресет \"{name}\"?",
+            "presetDeleteConfirmClick": "Нажмите еще раз для подтверждения",
+            "presetOverwriteConfirm": "Пресет \"{name}\" уже существует. Перезаписать?",
+            "presetNamePlaceholder": "Имя пресета...",
             "baseModel": "Базовая модель",
             "modelTags": "Теги (Топ 20)",
             "modelTypes": "Model Types",
@@ -1165,7 +1177,11 @@
         "exampleImages": {
             "opened": "Папка с примерами изображений открыта",
             "openingFolder": "Открытие папки с примерами изображений",
-            "failedToOpen": "Не удалось открыть папку с примерами изображений"
+            "failedToOpen": "Не удалось открыть папку с примерами изображений",
+            "setupRequired": "Хранилище примеров изображений",
+            "setupDescription": "Чтобы добавить собственные примеры изображений, сначала нужно установить место загрузки.",
+            "setupUsage": "Этот путь используется как для загруженных, так и для пользовательских примеров изображений.",
+            "openSettings": "Открыть настройки"
         }
     },
     "help": {
@@ -1214,6 +1230,7 @@
         "checkingUpdates": "Проверка обновлений...",
         "checkingMessage": "Пожалуйста, подождите, пока мы проверяем последнюю версию.",
         "showNotifications": "Показывать уведомления об обновлениях",
+        "latestBadge": "Последний",
         "updateProgress": {
             "preparing": "Подготовка обновления...",
             "installing": "Установка обновления...",
@@ -1414,7 +1431,26 @@
         "filters": {
             "applied": "{message}",
             "cleared": "Фильтры очищены",
-            "noCustomFilterToClear": "Нет пользовательского фильтра для очистки"
+            "noCustomFilterToClear": "Нет пользовательского фильтра для очистки",
+            "noActiveFilters": "Нет активных фильтров для сохранения"
+        },
+        "presets": {
+            "created": "Пресет \"{name}\" создан",
+            "deleted": "Пресет \"{name}\" удален",
+            "applied": "Пресет \"{name}\" применен",
+            "overwritten": "Пресет \"{name}\" перезаписан",
+            "restored": "Пресеты по умолчанию восстановлены"
+        },
+        "error": {
+            "presetNameEmpty": "Имя пресета не может быть пустым",
+            "presetNameTooLong": "Имя пресета должно содержать не более {max} символов",
+            "presetNameInvalidChars": "Имя пресета содержит недопустимые символы",
+            "presetNameExists": "Пресет с таким именем уже существует",
+            "maxPresetsReached": "Допустимо максимум {max} пресетов. Удалите один, чтобы добавить больше.",
+            "presetNotFound": "Пресет не найден",
+            "invalidPreset": "Недопустимые данные пресета",
+            "deletePresetFailed": "Не удалось удалить пресет",
+            "applyPresetFailed": "Не удалось применить пресет"
         },
         "downloads": {
             "imagesCompleted": "Примеры изображений {action} завершены",
@@ -1536,6 +1572,20 @@
             "content": "LoRA Manager is a passion project maintained full-time by a solo developer. Your support on Ko-fi helps cover development costs, keeps new updates coming, and unlocks a license key for the LM Civitai Extension as a thank-you gift. Every contribution truly makes a difference.",
             "supportCta": "Support on Ko-fi",
             "learnMore": "LM Civitai Extension Tutorial"
+        },
+        "cacheHealth": {
+            "corrupted": {
+                "title": "Обнаружено повреждение кэша"
+            },
+            "degraded": {
+                "title": "Обнаружены проблемы с кэшем"
+            },
+            "content": "{invalid} из {total} записей кэша недействительны ({rate}). Это может привести к отсутствию моделей или ошибкам. Рекомендуется перестроить кэш.",
+            "rebuildCache": "Перестроить кэш",
+            "dismiss": "Отклонить",
+            "rebuilding": "Перестроение кэша...",
+            "rebuildFailed": "Не удалось перестроить кэш: {error}",
+            "retry": "Повторить"
         }
     }
 }

locales/zh-CN.json

@@ -10,7 +10,8 @@
             "next": "下一步",
             "backToTop": "返回顶部",
             "settings": "设置",
-            "help": "帮助"
+            "help": "帮助",
+            "add": "添加"
         },
         "status": {
             "loading": "加载中...",
@@ -204,6 +205,17 @@
         },
         "filter": {
             "title": "筛选模型",
+            "presets": "预设",
+            "savePreset": "将当前激活的筛选器保存为新预设。",
+            "savePresetDisabledActive": "无法保存：已有预设处于激活状态。修改筛选器后可保存新预设",
+            "savePresetDisabledNoFilters": "先选择筛选器，然后保存为预设",
+            "savePresetPrompt": "输入预设名称：",
+            "presetClickTooltip": "点击应用预设 \"{name}\"",
+            "presetDeleteTooltip": "删除预设",
+            "presetDeleteConfirm": "删除预设 \"{name}\"？",
+            "presetDeleteConfirmClick": "再次点击确认",
+            "presetOverwriteConfirm": "预设 \"{name}\" 已存在。是否覆盖？",
+            "presetNamePlaceholder": "预设名称...",
             "baseModel": "基础模型",
             "modelTags": "标签（前20）",
             "modelTypes": "Model Types",
@@ -1165,7 +1177,11 @@
         "exampleImages": {
             "opened": "示例图片文件夹已打开",
             "openingFolder": "正在打开示例图片文件夹",
-            "failedToOpen": "打开示例图片文件夹失败"
+            "failedToOpen": "打开示例图片文件夹失败",
+            "setupRequired": "示例图片存储",
+            "setupDescription": "要添加自定义示例图片，您需要先设置下载位置。",
+            "setupUsage": "此路径用于存储下载的示例图片和自定义图片。",
+            "openSettings": "打开设置"
         }
     },
     "help": {
@@ -1214,6 +1230,7 @@
         "checkingUpdates": "正在检查更新...",
         "checkingMessage": "请稍候，正在检查最新版本。",
         "showNotifications": "显示更新通知",
+        "latestBadge": "最新",
         "updateProgress": {
             "preparing": "正在准备更新...",
             "installing": "正在安装更新...",
@@ -1414,7 +1431,26 @@
         "filters": {
             "applied": "{message}",
             "cleared": "筛选已清除",
-            "noCustomFilterToClear": "没有自定义筛选可清除"
+            "noCustomFilterToClear": "没有自定义筛选可清除",
+            "noActiveFilters": "没有可保存的激活筛选"
+        },
+        "presets": {
+            "created": "预设 \"{name}\" 已创建",
+            "deleted": "预设 \"{name}\" 已删除",
+            "applied": "预设 \"{name}\" 已应用",
+            "overwritten": "预设 \"{name}\" 已覆盖",
+            "restored": "默认预设已恢复"
+        },
+        "error": {
+            "presetNameEmpty": "预设名称不能为空",
+            "presetNameTooLong": "预设名称不能超过 {max} 个字符",
+            "presetNameInvalidChars": "预设名称包含无效字符",
+            "presetNameExists": "已存在同名预设",
+            "maxPresetsReached": "最多允许 {max} 个预设。删除一个以添加更多。",
+            "presetNotFound": "预设未找到",
+            "invalidPreset": "无效的预设数据",
+            "deletePresetFailed": "删除预设失败",
+            "applyPresetFailed": "应用预设失败"
         },
         "downloads": {
             "imagesCompleted": "示例图片{action}完成",
@@ -1536,6 +1572,20 @@
             "content": "来爱发电为Lora Manager项目发电，支持项目持续开发的同时，获取浏览器插件验证码，按季支付更优惠！支付宝/微信方便支付。感谢支持！🚀",
             "supportCta": "为LM发电",
             "learnMore": "浏览器插件教程"
+        },
+        "cacheHealth": {
+            "corrupted": {
+                "title": "检测到缓存损坏"
+            },
+            "degraded": {
+                "title": "检测到缓存问题"
+            },
+            "content": "{total} 个缓存条目中有 {invalid} 个无效（{rate}）。这可能导致模型丢失或错误。建议重建缓存。",
+            "rebuildCache": "重建缓存",
+            "dismiss": "忽略",
+            "rebuilding": "正在重建缓存...",
+            "rebuildFailed": "重建缓存失败：{error}",
+            "retry": "重试"
         }
     }
-}
+}

locales/zh-TW.json

@@ -10,7 +10,8 @@
             "next": "下一步",
             "backToTop": "回到頂部",
             "settings": "設定",
-            "help": "說明"
+            "help": "說明",
+            "add": "新增"
         },
         "status": {
             "loading": "載入中...",
@@ -204,6 +205,17 @@
         },
         "filter": {
             "title": "篩選模型",
+            "presets": "預設",
+            "savePreset": "將目前啟用的篩選器儲存為新預設。",
+            "savePresetDisabledActive": "無法儲存：已有預設處於啟用狀態。修改篩選器後可儲存新預設",
+            "savePresetDisabledNoFilters": "先選擇篩選器，然後儲存為預設",
+            "savePresetPrompt": "輸入預設名稱：",
+            "presetClickTooltip": "點擊套用預設 \"{name}\"",
+            "presetDeleteTooltip": "刪除預設",
+            "presetDeleteConfirm": "刪除預設 \"{name}\"？",
+            "presetDeleteConfirmClick": "再次點擊確認",
+            "presetOverwriteConfirm": "預設 \"{name}\" 已存在。是否覆蓋？",
+            "presetNamePlaceholder": "預設名稱...",
             "baseModel": "基礎模型",
             "modelTags": "標籤（前 20）",
             "modelTypes": "Model Types",
@@ -1165,7 +1177,11 @@
         "exampleImages": {
             "opened": "範例圖片資料夾已開啟",
             "openingFolder": "正在開啟範例圖片資料夾",
-            "failedToOpen": "開啟範例圖片資料夾失敗"
+            "failedToOpen": "開啟範例圖片資料夾失敗",
+            "setupRequired": "範例圖片儲存",
+            "setupDescription": "要新增自訂範例圖片，您需要先設定下載位置。",
+            "setupUsage": "此路徑用於儲存下載的範例圖片和自訂圖片。",
+            "openSettings": "開啟設定"
         }
     },
     "help": {
@@ -1214,6 +1230,7 @@
         "checkingUpdates": "正在檢查更新...",
         "checkingMessage": "請稍候，正在檢查最新版本。",
         "showNotifications": "顯示更新通知",
+        "latestBadge": "最新",
         "updateProgress": {
             "preparing": "正在準備更新...",
             "installing": "正在安裝更新...",
@@ -1414,7 +1431,26 @@
         "filters": {
             "applied": "{message}",
             "cleared": "篩選已清除",
-            "noCustomFilterToClear": "無自訂篩選可清除"
+            "noCustomFilterToClear": "無自訂篩選可清除",
+            "noActiveFilters": "沒有可儲存的啟用篩選"
+        },
+        "presets": {
+            "created": "預設 \"{name}\" 已建立",
+            "deleted": "預設 \"{name}\" 已刪除",
+            "applied": "預設 \"{name}\" 已套用",
+            "overwritten": "預設 \"{name}\" 已覆蓋",
+            "restored": "預設設定已恢復"
+        },
+        "error": {
+            "presetNameEmpty": "預設名稱不能為空",
+            "presetNameTooLong": "預設名稱不能超過 {max} 個字元",
+            "presetNameInvalidChars": "預設名稱包含無效字元",
+            "presetNameExists": "已存在同名預設",
+            "maxPresetsReached": "最多允許 {max} 個預設。刪除一個以新增更多。",
+            "presetNotFound": "預設未找到",
+            "invalidPreset": "無效的預設資料",
+            "deletePresetFailed": "刪除預設失敗",
+            "applyPresetFailed": "套用預設失敗"
         },
         "downloads": {
             "imagesCompleted": "範例圖片{action}完成",
@@ -1536,6 +1572,20 @@
             "content": "LoRA Manager is a passion project maintained full-time by a solo developer. Your support on Ko-fi helps cover development costs, keeps new updates coming, and unlocks a license key for the LM Civitai Extension as a thank-you gift. Every contribution truly makes a difference.",
             "supportCta": "Support on Ko-fi",
             "learnMore": "LM Civitai Extension Tutorial"
+        },
+        "cacheHealth": {
+            "corrupted": {
+                "title": "檢測到快取損壞"
+            },
+            "degraded": {
+                "title": "檢測到快取問題"
+            },
+            "content": "{total} 個快取項目中有 {invalid} 個無效（{rate}）。這可能會導致模型遺失或錯誤。建議重建快取。",
+            "rebuildCache": "重建快取",
+            "dismiss": "關閉",
+            "rebuilding": "重建快取中...",
+            "rebuildFailed": "重建快取失敗：{error}",
+            "retry": "重試"
         }
     }
 }

package.json

@@ -4,7 +4,9 @@
   "private": true,
   "type": "module",
   "scripts": {
-    "test": "vitest run",
+    "test": "npm run test:js && npm run test:vue",
+    "test:js": "vitest run",
+    "test:vue": "cd vue-widgets && npx vitest run",
     "test:watch": "vitest",
     "test:coverage": "node scripts/run_frontend_coverage.js"
   },

py/config.py

@@ -9,6 +9,7 @@ import json
 import urllib.parse
 import time
 
+from .utils.cache_paths import CacheType, get_cache_file_path, get_legacy_cache_paths
 from .utils.settings_paths import ensure_settings_file, get_settings_dir, load_settings_template
 
 # Use an environment variable to control standalone mode
@@ -223,13 +224,26 @@ class Config:
             logger.error(f"Error checking link status for {path}: {e}")
             return False
 
+    def _entry_is_symlink(self, entry: os.DirEntry) -> bool:
+        """Check if a directory entry is a symlink, including Windows junctions."""
+        if entry.is_symlink():
+            return True
+        if platform.system() == 'Windows':
+            try:
+                import ctypes
+                FILE_ATTRIBUTE_REPARSE_POINT = 0x400
+                attrs = ctypes.windll.kernel32.GetFileAttributesW(entry.path)
+                return attrs != -1 and (attrs & FILE_ATTRIBUTE_REPARSE_POINT)
+            except Exception:
+                pass
+        return False
+
     def _normalize_path(self, path: str) -> str:
         return os.path.normpath(path).replace(os.sep, '/')
 
     def _get_symlink_cache_path(self) -> Path:
-        cache_dir = Path(get_settings_dir(create=True)) / "cache"
-        cache_dir.mkdir(parents=True, exist_ok=True)
-        return cache_dir / "symlink_map.json"
+        canonical_path = get_cache_file_path(CacheType.SYMLINK, create_dir=True)
+        return Path(canonical_path)
 
     def _symlink_roots(self) -> List[str]:
         roots: List[str] = []
@@ -241,8 +255,32 @@ class Config:
     def _build_symlink_fingerprint(self) -> Dict[str, object]:
         roots = [self._normalize_path(path) for path in self._symlink_roots() if path]
         unique_roots = sorted(set(roots))
-        # Fingerprint now only contains the root paths to avoid sensitivity to folder content changes.
-        return {"roots": unique_roots}
+
+        # Include first-level symlinks in fingerprint for change detection.
+        # This ensures new symlinks under roots trigger a cache invalidation.
+        # Use lists (not tuples) for JSON serialization compatibility.
+        direct_symlinks: List[List[str]] = []
+        for root in unique_roots:
+            try:
+                if os.path.isdir(root):
+                    with os.scandir(root) as it:
+                        for entry in it:
+                            if self._entry_is_symlink(entry):
+                                try:
+                                    target = os.path.realpath(entry.path)
+                                    direct_symlinks.append([
+                                        self._normalize_path(entry.path),
+                                        self._normalize_path(target)
+                                    ])
+                                except OSError:
+                                    pass
+            except (OSError, PermissionError):
+                pass
+
+        return {
+            "roots": unique_roots,
+            "direct_symlinks": sorted(direct_symlinks)
+        }
 
     def _initialize_symlink_mappings(self) -> None:
         start = time.perf_counter()
@@ -255,15 +293,19 @@ class Config:
             )
             self._rebuild_preview_roots()
 
-            # Only rescan if target roots have changed. 
-            # This is stable across file additions/deletions.
             current_fingerprint = self._build_symlink_fingerprint()
             cached_fingerprint = self._cached_fingerprint
-            
-            if cached_fingerprint and current_fingerprint == cached_fingerprint:
+
+            # Check 1: First-level symlinks unchanged (catches new symlinks at root)
+            fingerprint_valid = cached_fingerprint and current_fingerprint == cached_fingerprint
+
+            # Check 2: All cached mappings still valid (catches changes at any depth)
+            mappings_valid = self._validate_cached_mappings() if fingerprint_valid else False
+
+            if fingerprint_valid and mappings_valid:
                 return
 
-            logger.info("Symlink root paths changed; rescanning symbolic links")
+            logger.info("Symlink configuration changed; rescanning symbolic links")
 
         self.rebuild_symlink_cache()
         logger.info(
@@ -280,14 +322,28 @@ class Config:
     def _load_persisted_cache_into_mappings(self) -> bool:
         """Load the symlink cache and store its fingerprint for comparison."""
         cache_path = self._get_symlink_cache_path()
-        if not cache_path.exists():
-            return False
 
-        try:
-            with cache_path.open("r", encoding="utf-8") as handle:
-                payload = json.load(handle)
-        except Exception as exc:
-            logger.info("Failed to load symlink cache %s: %s", cache_path, exc)
+        # Check canonical path first, then legacy paths for migration
+        paths_to_check = [cache_path]
+        legacy_paths = get_legacy_cache_paths(CacheType.SYMLINK)
+        paths_to_check.extend(Path(p) for p in legacy_paths if p != str(cache_path))
+
+        loaded_path = None
+        payload = None
+
+        for check_path in paths_to_check:
+            if not check_path.exists():
+                continue
+            try:
+                with check_path.open("r", encoding="utf-8") as handle:
+                    payload = json.load(handle)
+                    loaded_path = check_path
+                    break
+            except Exception as exc:
+                logger.info("Failed to load symlink cache %s: %s", check_path, exc)
+                continue
+
+        if payload is None:
             return False
 
         if not isinstance(payload, dict):
@@ -307,7 +363,67 @@ class Config:
             normalized_mappings[self._normalize_path(target)] = self._normalize_path(link)
 
         self._path_mappings = normalized_mappings
-        logger.info("Symlink cache loaded with %d mappings", len(self._path_mappings))
+
+        # Log migration if loaded from legacy path
+        if loaded_path is not None and loaded_path != cache_path:
+            logger.info(
+                "Symlink cache migrated from %s (will save to %s)",
+                loaded_path,
+                cache_path,
+            )
+
+            try:
+                if loaded_path.exists():
+                    loaded_path.unlink()
+                    logger.info("Cleaned up legacy symlink cache: %s", loaded_path)
+
+                    try:
+                        parent_dir = loaded_path.parent
+                        if parent_dir.name == "cache" and not any(parent_dir.iterdir()):
+                            parent_dir.rmdir()
+                            logger.info("Removed empty legacy cache directory: %s", parent_dir)
+                    except Exception:
+                        pass
+
+            except Exception as exc:
+                logger.warning(
+                    "Failed to cleanup legacy symlink cache %s: %s",
+                    loaded_path,
+                    exc,
+                )
+        else:
+            logger.info("Symlink cache loaded with %d mappings", len(self._path_mappings))
+
+        return True
+
+    def _validate_cached_mappings(self) -> bool:
+        """Verify all cached symlink mappings are still valid.
+
+        Returns True if all mappings are valid, False if rescan is needed.
+        This catches removed or retargeted symlinks at ANY depth.
+        """
+        for target, link in self._path_mappings.items():
+            # Convert normalized paths back to OS paths
+            link_path = link.replace('/', os.sep)
+
+            # Check if symlink still exists
+            if not self._is_link(link_path):
+                logger.debug("Cached symlink no longer exists: %s", link_path)
+                return False
+
+            # Check if target is still the same
+            try:
+                actual_target = self._normalize_path(os.path.realpath(link_path))
+                if actual_target != target:
+                    logger.debug(
+                        "Symlink target changed: %s -> %s (cached: %s)",
+                        link_path, actual_target, target
+                    )
+                    return False
+            except OSError:
+                logger.debug("Cannot resolve symlink: %s", link_path)
+                return False
+
         return True
 
     def _save_symlink_cache(self) -> None:
@@ -325,83 +441,53 @@ class Config:
             logger.info("Failed to write symlink cache %s: %s", cache_path, exc)
 
     def _scan_symbolic_links(self):
-        """Scan all symbolic links in LoRA, Checkpoint, and Embedding root directories"""
+        """Scan symbolic links in LoRA, Checkpoint, and Embedding root directories.
+
+        Only scans the first level of each root directory to avoid performance
+        issues with large file systems. Detects symlinks and Windows junctions
+        at the root level only (not nested symlinks in subdirectories).
+        """
         start = time.perf_counter()
         
         # Reset mappings before rescanning to avoid stale entries
         self._path_mappings.clear()
         self._seed_root_symlink_mappings()
-        visited_dirs: Set[str] = set()
         for root in self._symlink_roots():
-            self._scan_directory_links(root, visited_dirs)
+            self._scan_first_level_symlinks(root)
         logger.debug(
             "Symlink scan finished in %.2f ms with %d mappings",
             (time.perf_counter() - start) * 1000,
             len(self._path_mappings),
         )
 
-    def _scan_directory_links(self, root: str, visited_dirs: Set[str]):
-        """Iteratively scan directory symlinks to avoid deep recursion."""
+    def _scan_first_level_symlinks(self, root: str):
+        """Scan only the first level of a directory for symlinks.
+        
+        This avoids traversing the entire directory tree which can be extremely
+        slow for large model collections. Only symlinks directly under the root
+        are detected.
+        """
         try:
-            # Note: We only use realpath for the initial root if it's not already resolved
-            # to ensure we have a valid entry point.
-            root_real = self._normalize_path(os.path.realpath(root))
-        except OSError:
-            root_real = self._normalize_path(root)
+            with os.scandir(root) as it:
+                for entry in it:
+                    try:
+                        # Only detect symlinks including Windows junctions
+                        # Skip normal directories to avoid deep traversal
+                        if not self._entry_is_symlink(entry):
+                            continue
 
-        if root_real in visited_dirs:
-            return
+                        # Resolve the symlink target
+                        target_path = os.path.realpath(entry.path)
+                        if not os.path.isdir(target_path):
+                            continue
 
-        visited_dirs.add(root_real)
-        # Stack entries: (display_path, real_resolved_path)
-        stack: List[Tuple[str, str]] = [(root, root_real)]
-
-        while stack:
-            current_display, current_real = stack.pop()
-            try:
-                with os.scandir(current_display) as it:
-                    for entry in it:
-                        try:
-                            # 1. High speed detection using dirent data (is_symlink)
-                            is_link = entry.is_symlink()
-                            
-                            # On Windows, is_symlink handles reparse points
-                            if is_link:
-                                # Only resolve realpath when we actually find a link
-                                target_path = os.path.realpath(entry.path)
-                                if not os.path.isdir(target_path):
-                                    continue
-
-                                normalized_target = self._normalize_path(target_path)
-                                self.add_path_mapping(entry.path, target_path)
-                                
-                                if normalized_target in visited_dirs:
-                                    continue
-                                    
-                                visited_dirs.add(normalized_target)
-                                stack.append((target_path, normalized_target))
-                                continue
-
-                            # 2. Process normal directories
-                            if not entry.is_dir(follow_symlinks=False):
-                                continue
-
-                            # For normal directories, we avoid realpath() call by 
-                            # incrementally building the real path relative to current_real.
-                            # This is safe because 'entry' is NOT a symlink.
-                            entry_real = self._normalize_path(os.path.join(current_real, entry.name))
-                            
-                            if entry_real in visited_dirs:
-                                continue
-                            
-                            visited_dirs.add(entry_real)
-                            stack.append((entry.path, entry_real))
-                        except Exception as inner_exc:
-                            logger.debug(
-                                "Error processing directory entry %s: %s", entry.path, inner_exc
-                            )
-            except Exception as e:
-                logger.error(f"Error scanning links in {current_display}: {e}")
+                        self.add_path_mapping(entry.path, target_path)
+                    except Exception as inner_exc:
+                        logger.debug(
+                            "Error processing directory entry %s: %s", entry.path, inner_exc
+                        )
+        except Exception as e:
+            logger.error(f"Error scanning links in {root}: {e}")
 
 
 
@@ -559,6 +645,23 @@ class Config:
         checkpoint_map = self._dedupe_existing_paths(checkpoint_paths)
         unet_map = self._dedupe_existing_paths(unet_paths)
 
+        # Detect when checkpoints and unet share the same physical location
+        # This is a configuration issue that can cause duplicate model entries
+        overlapping_real_paths = set(checkpoint_map.keys()) & set(unet_map.keys())
+        if overlapping_real_paths:
+            logger.warning(
+                "Detected overlapping paths between 'checkpoints' and 'diffusion_models' (unet). "
+                "They should not point to the same physical folder as they are different model types. "
+                "Please fix your ComfyUI path configuration to separate these folders. "
+                "Falling back to 'checkpoints' for backward compatibility. "
+                "Overlapping real paths: %s",
+                [checkpoint_map.get(rp, rp) for rp in overlapping_real_paths]
+            )
+            # Remove overlapping paths from unet_map to prioritize checkpoints
+            for rp in overlapping_real_paths:
+                if rp in unet_map:
+                    del unet_map[rp]
+
         merged_map: Dict[str, str] = {}
         for real_path, original in {**checkpoint_map, **unet_map}.items():
             if real_path not in merged_map:
@@ -663,7 +766,23 @@ class Config:
         return f'/api/lm/previews?path={encoded_path}'
 
     def is_preview_path_allowed(self, preview_path: str) -> bool:
-        """Return ``True`` if ``preview_path`` is within an allowed directory."""
+        """Return ``True`` if ``preview_path`` is within an allowed directory.
+        
+        If the path is initially rejected, attempts to discover deep symlinks
+        that were not scanned during initialization. If a symlink is found,
+        updates the in-memory path mappings and retries the check.
+        """
+
+        if self._is_path_in_allowed_roots(preview_path):
+            return True
+
+        if self._try_discover_deep_symlink(preview_path):
+            return self._is_path_in_allowed_roots(preview_path)
+
+        return False
+
+    def _is_path_in_allowed_roots(self, preview_path: str) -> bool:
+        """Check if preview_path is within allowed preview roots without modification."""
 
         if not preview_path:
             return False
@@ -673,29 +792,72 @@ class Config:
         except Exception:
             return False
 
-        # Use os.path.normcase for case-insensitive comparison on Windows.
-        # On Windows, Path.relative_to() is case-sensitive for drive letters,
-        # causing paths like 'a:/folder' to not match 'A:/folder'.
         candidate_str = os.path.normcase(str(candidate))
         for root in self._preview_root_paths:
             root_str = os.path.normcase(str(root))
-            # Check if candidate is equal to or under the root directory
             if candidate_str == root_str or candidate_str.startswith(root_str + os.sep):
                 return True
 
-        if self._preview_root_paths:
-            logger.debug(
-                "Preview path rejected: %s (candidate=%s, num_roots=%d, first_root=%s)",
-                preview_path,
-                candidate_str,
-                len(self._preview_root_paths),
-                os.path.normcase(str(next(iter(self._preview_root_paths)))),
-            )
-        else:
-            logger.debug(
-                "Preview path rejected (no roots configured): %s",
-                preview_path,
-            )
+        logger.debug(
+            "Path not in allowed roots: %s (candidate=%s, num_roots=%d)",
+            preview_path,
+            candidate_str,
+            len(self._preview_root_paths),
+        )
+
+        return False
+
+    def _try_discover_deep_symlink(self, preview_path: str) -> bool:
+        """Attempt to discover a deep symlink that contains the preview_path.
+
+        Walks up from the preview path to the root directories, checking each
+        parent directory for symlinks. If a symlink is found, updates the
+        in-memory path mappings and preview roots.
+
+        Only updates in-memory state (self._path_mappings and self._preview_root_paths),
+        does not modify the persistent cache file.
+
+        Returns:
+            True if a symlink was discovered and mappings updated, False otherwise.
+        """
+        if not preview_path:
+            return False
+
+        try:
+            candidate = Path(preview_path).expanduser()
+        except Exception:
+            return False
+
+        current = candidate
+        while True:
+            try:
+                if self._is_link(str(current)):
+                    try:
+                        target = os.path.realpath(str(current))
+                        normalized_target = self._normalize_path(target)
+                        normalized_link = self._normalize_path(str(current))
+
+                        self._path_mappings[normalized_target] = normalized_link
+                        self._preview_root_paths.update(self._expand_preview_root(normalized_target))
+                        self._preview_root_paths.update(self._expand_preview_root(normalized_link))
+
+                        logger.debug(
+                            "Discovered deep symlink: %s -> %s (preview path: %s)",
+                            normalized_link,
+                            normalized_target,
+                            preview_path
+                        )
+
+                        return True
+                    except OSError:
+                        pass
+            except OSError:
+                pass
+
+            parent = current.parent
+            if parent == current:
+                break
+            current = parent
 
         return False
 

py/metadata_collector/__init__.py

@@ -1,4 +1,7 @@
 import os
+import logging
+
+logger = logging.getLogger(__name__)
 
 # Check if running in standalone mode
 standalone_mode = os.environ.get("LORA_MANAGER_STANDALONE", "0") == "1" or os.environ.get("HF_HUB_DISABLE_TELEMETRY", "0") == "0"
@@ -14,7 +17,7 @@ if not standalone_mode:
         # Initialize registry
         registry = MetadataRegistry()
         
-        print("ComfyUI Metadata Collector initialized")
+        logger.info("ComfyUI Metadata Collector initialized")
         
     def get_metadata(prompt_id=None):
         """Helper function to get metadata from the registry"""
@@ -23,7 +26,7 @@ if not standalone_mode:
 else:
     # Standalone mode - provide dummy implementations
     def init():
-        print("ComfyUI Metadata Collector disabled in standalone mode")
+        logger.info("ComfyUI Metadata Collector disabled in standalone mode")
         
     def get_metadata(prompt_id=None):
         """Dummy implementation for standalone mode"""

py/metadata_collector/metadata_hook.py

@@ -1,7 +1,10 @@
 import sys
 import inspect
+import logging
 from .metadata_registry import MetadataRegistry
 
+logger = logging.getLogger(__name__)
+
 class MetadataHook:
     """Install hooks for metadata collection"""
     
@@ -23,7 +26,7 @@ class MetadataHook:
                     
             # If we can't find the execution module, we can't install hooks
             if execution is None:
-                print("Could not locate ComfyUI execution module, metadata collection disabled")
+                logger.warning("Could not locate ComfyUI execution module, metadata collection disabled")
                 return
             
             # Detect whether we're using the new async version of ComfyUI
@@ -37,16 +40,16 @@ class MetadataHook:
                 is_async = inspect.iscoroutinefunction(execution._map_node_over_list)
             
             if is_async:
-                print("Detected async ComfyUI execution, installing async metadata hooks")
+                logger.info("Detected async ComfyUI execution, installing async metadata hooks")
                 MetadataHook._install_async_hooks(execution, map_node_func_name)
             else:
-                print("Detected sync ComfyUI execution, installing sync metadata hooks")
+                logger.info("Detected sync ComfyUI execution, installing sync metadata hooks")
                 MetadataHook._install_sync_hooks(execution)
             
-            print("Metadata collection hooks installed for runtime values")
+            logger.info("Metadata collection hooks installed for runtime values")
             
         except Exception as e:
-            print(f"Error installing metadata hooks: {str(e)}")
+            logger.error(f"Error installing metadata hooks: {str(e)}")
     
     @staticmethod
     def _install_sync_hooks(execution):
@@ -82,7 +85,7 @@ class MetadataHook:
                         if node_id is not None:
                             registry.record_node_execution(node_id, class_type, input_data_all, None)
                 except Exception as e:
-                    print(f"Error collecting metadata (pre-execution): {str(e)}")
+                    logger.error(f"Error collecting metadata (pre-execution): {str(e)}")
             
             # Execute the original function
             results = original_map_node_over_list(obj, input_data_all, func, allow_interrupt, execution_block_cb, pre_execute_cb)
@@ -113,7 +116,7 @@ class MetadataHook:
                         if node_id is not None:
                             registry.update_node_execution(node_id, class_type, results)
                 except Exception as e:
-                    print(f"Error collecting metadata (post-execution): {str(e)}")
+                    logger.error(f"Error collecting metadata (post-execution): {str(e)}")
             
             return results
             
@@ -159,7 +162,7 @@ class MetadataHook:
                         if node_id is not None:
                             registry.record_node_execution(node_id, class_type, input_data_all, None)
                 except Exception as e:
-                    print(f"Error collecting metadata (pre-execution): {str(e)}")
+                    logger.error(f"Error collecting metadata (pre-execution): {str(e)}")
             
             # Call original function with all args/kwargs
             results = await original_map_node_over_list(
@@ -176,7 +179,7 @@ class MetadataHook:
                         if node_id is not None:
                             registry.update_node_execution(node_id, class_type, results)
                 except Exception as e:
-                    print(f"Error collecting metadata (post-execution): {str(e)}")
+                    logger.error(f"Error collecting metadata (post-execution): {str(e)}")
             
             return results
         

py/metadata_collector/node_extractors.py

@@ -714,10 +714,10 @@ NODE_EXTRACTORS = {
     "UNETLoader": UNETLoaderExtractor,          # Updated to use dedicated extractor
     "UnetLoaderGGUF": UNETLoaderExtractor,  # Updated to use dedicated extractor
     "LoraLoader": LoraLoaderExtractor,
-    "LoraManagerLoader": LoraLoaderManagerExtractor,
+    "LoraLoaderLM": LoraLoaderManagerExtractor,
     # Conditioning
     "CLIPTextEncode": CLIPTextEncodeExtractor,
-    "PromptLoraManager": CLIPTextEncodeExtractor,
+    "PromptLM": CLIPTextEncodeExtractor,
     "CLIPTextEncodeFlux": CLIPTextEncodeFluxExtractor,  # Add CLIPTextEncodeFlux
     "WAS_Text_to_Conditioning": CLIPTextEncodeExtractor,
     "AdvancedCLIPTextEncode": CLIPTextEncodeExtractor,  # From https://github.com/BlenderNeko/ComfyUI_ADV_CLIP_emb

py/nodes/debug_metadata.py

@@ -4,7 +4,7 @@ from ..metadata_collector.metadata_processor import MetadataProcessor
 logger = logging.getLogger(__name__)
 
 
-class DebugMetadata:
+class DebugMetadataLM:
     NAME = "Debug Metadata (LoraManager)"
     CATEGORY = "Lora Manager/utils"
     DESCRIPTION = "Debug node to verify metadata_processor functionality"

py/nodes/lora_cycler.py

@@ -0,0 +1,134 @@
+"""
+Lora Cycler Node - Sequentially cycles through LoRAs from a pool.
+
+This node accepts optional pool_config input to filter available LoRAs, and outputs
+a LORA_STACK with one LoRA at a time. Returns UI updates with current/next LoRA info
+and tracks the cycle progress which persists across workflow save/load.
+"""
+
+import logging
+import os
+from ..utils.utils import get_lora_info
+
+logger = logging.getLogger(__name__)
+
+
+class LoraCyclerLM:
+    """Node that sequentially cycles through LoRAs from a pool"""
+
+    NAME = "Lora Cycler (LoraManager)"
+    CATEGORY = "Lora Manager/randomizer"
+
+    @classmethod
+    def INPUT_TYPES(cls):
+        return {
+            "required": {
+                "cycler_config": ("CYCLER_CONFIG", {}),
+            },
+            "optional": {
+                "pool_config": ("POOL_CONFIG", {}),
+            },
+        }
+
+    RETURN_TYPES = ("LORA_STACK",)
+    RETURN_NAMES = ("LORA_STACK",)
+
+    FUNCTION = "cycle"
+    OUTPUT_NODE = False
+
+    async def cycle(self, cycler_config, pool_config=None):
+        """
+        Cycle through LoRAs based on configuration and pool filters.
+
+        Args:
+            cycler_config: Dict with cycler settings (current_index, model_strength, clip_strength, sort_by)
+            pool_config: Optional config from LoRA Pool node for filtering
+
+        Returns:
+            Dictionary with 'result' (LORA_STACK tuple) and 'ui' (for widget display)
+        """
+        from ..services.service_registry import ServiceRegistry
+        from ..services.lora_service import LoraService
+
+        # Extract settings from cycler_config
+        current_index = cycler_config.get("current_index", 1)  # 1-based
+        model_strength = float(cycler_config.get("model_strength", 1.0))
+        clip_strength = float(cycler_config.get("clip_strength", 1.0))
+        sort_by = "filename"
+
+        # Dual-index mechanism for batch queue synchronization
+        execution_index = cycler_config.get("execution_index")  # Can be None
+        # next_index_from_config = cycler_config.get("next_index")  # Not used on backend
+
+        # Get scanner and service
+        scanner = await ServiceRegistry.get_lora_scanner()
+        lora_service = LoraService(scanner)
+
+        # Get filtered and sorted LoRA list
+        lora_list = await lora_service.get_cycler_list(
+            pool_config=pool_config, sort_by=sort_by
+        )
+
+        total_count = len(lora_list)
+
+        if total_count == 0:
+            logger.warning("[LoraCyclerLM] No LoRAs available in pool")
+            return {
+                "result": ([],),
+                "ui": {
+                    "current_index": [1],
+                    "next_index": [1],
+                    "total_count": [0],
+                    "current_lora_name": [""],
+                    "current_lora_filename": [""],
+                    "error": ["No LoRAs available in pool"],
+                },
+            }
+
+        # Determine which index to use for this execution
+        # If execution_index is provided (batch queue case), use it
+        # Otherwise use current_index (first execution or non-batch case)
+        if execution_index is not None:
+            actual_index = execution_index
+        else:
+            actual_index = current_index
+
+        # Clamp index to valid range (1-based)
+        clamped_index = max(1, min(actual_index, total_count))
+
+        # Get LoRA at current index (convert to 0-based for list access)
+        current_lora = lora_list[clamped_index - 1]
+
+        # Build LORA_STACK with single LoRA
+        lora_path, _ = get_lora_info(current_lora["file_name"])
+        if not lora_path:
+            logger.warning(
+                f"[LoraCyclerLM] Could not find path for LoRA: {current_lora['file_name']}"
+            )
+            lora_stack = []
+        else:
+            # Normalize path separators
+            lora_path = lora_path.replace("/", os.sep)
+            lora_stack = [(lora_path, model_strength, clip_strength)]
+
+        # Calculate next index (wrap to 1 if at end)
+        next_index = clamped_index + 1
+        if next_index > total_count:
+            next_index = 1
+
+        # Get next LoRA for UI display (what will be used next generation)
+        next_lora = lora_list[next_index - 1]
+        next_display_name = next_lora["file_name"]
+
+        return {
+            "result": (lora_stack,),
+            "ui": {
+                "current_index": [clamped_index],
+                "next_index": [next_index],
+                "total_count": [total_count],
+                "current_lora_name": [current_lora["file_name"]],
+                "current_lora_filename": [current_lora["file_name"]],
+                "next_lora_name": [next_display_name],
+                "next_lora_filename": [next_lora["file_name"]],
+            },
+        }

py/nodes/lora_loader.py

@@ -6,7 +6,7 @@ from .utils import FlexibleOptionalInputType, any_type, extract_lora_name, get_l
 
 logger = logging.getLogger(__name__)
 
-class LoraManagerLoader:
+class LoraLoaderLM:
     NAME = "Lora Loader (LoraManager)"
     CATEGORY = "Lora Manager/loaders"
     
@@ -16,12 +16,9 @@ class LoraManagerLoader:
             "required": {
                 "model": ("MODEL",),
                 # "clip": ("CLIP",),
-                "text": ("STRING", {
-                    "multiline": True,
-                    "pysssss.autocomplete": False, 
-                    "dynamicPrompts": True, 
+                "text": ("AUTOCOMPLETE_TEXT_LORAS", {
+                    "placeholder": "Search LoRAs to add...",
                     "tooltip": "Format: <lora:lora_name:strength> separated by spaces or punctuation",
-                    "placeholder": "LoRA syntax input: <lora:name:strength>"
                 }),
             },
             "optional": FlexibleOptionalInputType(any_type),
@@ -131,7 +128,7 @@ class LoraManagerLoader:
 
         return (model, clip, trigger_words_text, formatted_loras_text)
 
-class LoraManagerTextLoader:
+class LoraTextLoaderLM:
     NAME = "LoRA Text Loader (LoraManager)"
     CATEGORY = "Lora Manager/loaders"
     

py/nodes/lora_pool.py

@@ -10,7 +10,7 @@ import logging
 logger = logging.getLogger(__name__)
 
 
-class LoraPoolNode:
+class LoraPoolLM:
     """
     A node that defines LoRA filter criteria through a Vue-based widget.
 
@@ -67,7 +67,7 @@ class LoraPoolNode:
         filters = pool_config.get("filters", self._default_config()["filters"])
 
         # Log for debugging
-        logger.debug(f"[LoraPoolNode] Processing filters: {filters}")
+        logger.debug(f"[LoraPoolLM] Processing filters: {filters}")
 
         return (filters,)
 

py/nodes/lora_randomizer.py

@@ -15,7 +15,7 @@ from .utils import extract_lora_name
 logger = logging.getLogger(__name__)
 
 
-class LoraRandomizerNode:
+class LoraRandomizerLM:
     """Node that randomly selects LoRAs from a pool"""
 
     NAME = "Lora Randomizer (LoraManager)"
@@ -72,7 +72,7 @@ class LoraRandomizerNode:
         loras = self._preprocess_loras_input(loras)
 
         roll_mode = randomizer_config.get("roll_mode", "always")
-        logger.debug(f"[LoraRandomizerNode] roll_mode: {roll_mode}")
+        logger.debug(f"[LoraRandomizerLM] roll_mode: {roll_mode}")
 
         # Dual seed mechanism for batch queue synchronization
         # execution_seed: seed for generating execution_stack (= previous next_seed)
@@ -127,7 +127,7 @@ class LoraRandomizerNode:
             lora_path, trigger_words = get_lora_info(lora["name"])
             if not lora_path:
                 logger.warning(
-                    f"[LoraRandomizerNode] Could not find path for LoRA: {lora['name']}"
+                    f"[LoraRandomizerLM] Could not find path for LoRA: {lora['name']}"
                 )
                 continue
 

py/nodes/lora_stacker.py

@@ -6,7 +6,7 @@ import logging
 
 logger = logging.getLogger(__name__)
 
-class LoraStacker:
+class LoraStackerLM:
     NAME = "Lora Stacker (LoraManager)"
     CATEGORY = "Lora Manager/stackers"
 
@@ -14,12 +14,9 @@ class LoraStacker:
     def INPUT_TYPES(cls):
         return {
             "required": {
-                "text": ("STRING", {
-                    "multiline": True, 
-                    "pysssss.autocomplete": False, 
-                    "dynamicPrompts": True, 
+                "text": ("AUTOCOMPLETE_TEXT_LORAS", {
+                    "placeholder": "Search LoRAs to add...",
                     "tooltip": "Format: <lora:lora_name:strength> separated by spaces or punctuation",
-                    "placeholder": "LoRA syntax input: <lora:name:strength>"
                 }),
             },
             "optional": FlexibleOptionalInputType(any_type),

py/nodes/prompt.py

@@ -1,6 +1,6 @@
 from typing import Any, Optional
 
-class PromptLoraManager:
+class PromptLM:
     """Encodes text (and optional trigger words) into CLIP conditioning."""
 
     NAME = "Prompt (LoraManager)"
@@ -15,11 +15,10 @@ class PromptLoraManager:
         return {
             "required": {
                 "text": (
-                    'STRING',
+                    "AUTOCOMPLETE_TEXT_PROMPT,STRING",
                     {
-                        "multiline": True,
-                        "pysssss.autocomplete": False, 
-                        "dynamicPrompts": True,
+                        "widgetType": "AUTOCOMPLETE_TEXT_PROMPT",
+                        "placeholder": "Enter prompt... /char, /artist for quick tag search",
                         "tooltip": "The text to be encoded.",
                     },
                 ),

py/nodes/save_image.py

@@ -8,6 +8,9 @@ from ..metadata_collector.metadata_processor import MetadataProcessor
 from ..metadata_collector import get_metadata
 from PIL import Image, PngImagePlugin
 import piexif
+import logging
+
+logger = logging.getLogger(__name__)
 
 class SaveImageLM:
     NAME = "Save Image (LoraManager)"
@@ -385,7 +388,7 @@ class SaveImageLM:
                             exif_bytes = piexif.dump(exif_dict)
                             save_kwargs["exif"] = exif_bytes
                         except Exception as e:
-                            print(f"Error adding EXIF data: {e}")
+                            logger.error(f"Error adding EXIF data: {e}")
                     img.save(file_path, format="JPEG", **save_kwargs)
                 elif file_format == "webp":
                     try:
@@ -403,7 +406,7 @@ class SaveImageLM:
                         exif_bytes = piexif.dump(exif_dict)
                         save_kwargs["exif"] = exif_bytes
                     except Exception as e:
-                        print(f"Error adding EXIF data: {e}")
+                        logger.error(f"Error adding EXIF data: {e}")
                     
                     img.save(file_path, format="WEBP", **save_kwargs)
                 
@@ -414,7 +417,7 @@ class SaveImageLM:
                 })
                 
             except Exception as e:
-                print(f"Error saving image: {e}")
+                logger.error(f"Error saving image: {e}")
         
         return results
 

py/nodes/text.py

@@ -0,0 +1,33 @@
+class TextLM:
+    """A simple text node with autocomplete support."""
+
+    NAME = "Text (LoraManager)"
+    CATEGORY = "Lora Manager/utils"
+    DESCRIPTION = (
+        "A simple text input node with autocomplete support for tags and styles."
+    )
+
+    @classmethod
+    def INPUT_TYPES(cls):
+        return {
+            "required": {
+                "text": (
+                    "AUTOCOMPLETE_TEXT_PROMPT,STRING",
+                    {
+                        "widgetType": "AUTOCOMPLETE_TEXT_PROMPT",
+                        "placeholder": "Enter text... /char, /artist for quick tag search",
+                        "tooltip": "The text output.",
+                    },
+                ),
+            },
+        }
+
+    RETURN_TYPES = ("STRING",)
+    RETURN_NAMES = ("STRING",)
+    OUTPUT_TOOLTIPS = (
+        "The text output.",
+    )
+    FUNCTION = "process"
+
+    def process(self, text: str):
+        return (text,)

py/nodes/trigger_word_toggle.py

@@ -6,27 +6,36 @@ import logging
 logger = logging.getLogger(__name__)
 
 
-class TriggerWordToggle:
+class TriggerWordToggleLM:
     NAME = "TriggerWord Toggle (LoraManager)"
     CATEGORY = "Lora Manager/utils"
     DESCRIPTION = "Toggle trigger words on/off"
-    
+
     @classmethod
     def INPUT_TYPES(cls):
         return {
             "required": {
-                "group_mode": ("BOOLEAN", {
-                    "default": True,
-                    "tooltip": "When enabled, treats each group of trigger words as a single toggleable unit."
-                }),
-                "default_active": ("BOOLEAN", {
-                    "default": True,
-                    "tooltip": "Sets the default initial state (active or inactive) when trigger words are added."
-                }),
-                "allow_strength_adjustment": ("BOOLEAN", {
-                    "default": False,
-                    "tooltip": "Enable mouse wheel adjustment of each trigger word's strength."
-                }),
+                "group_mode": (
+                    "BOOLEAN",
+                    {
+                        "default": True,
+                        "tooltip": "When enabled, treats each group of trigger words as a single toggleable unit.",
+                    },
+                ),
+                "default_active": (
+                    "BOOLEAN",
+                    {
+                        "default": True,
+                        "tooltip": "Sets the default initial state (active or inactive) when trigger words are added.",
+                    },
+                ),
+                "allow_strength_adjustment": (
+                    "BOOLEAN",
+                    {
+                        "default": False,
+                        "tooltip": "Enable mouse wheel adjustment of each trigger word's strength.",
+                    },
+                ),
             },
             "optional": FlexibleOptionalInputType(any_type),
             "hidden": {
@@ -38,19 +47,35 @@ class TriggerWordToggle:
     RETURN_NAMES = ("filtered_trigger_words",)
     FUNCTION = "process_trigger_words"
 
-    def _get_toggle_data(self, kwargs, key='toggle_trigger_words'):
+    def _get_toggle_data(self, kwargs, key="toggle_trigger_words"):
         """Helper to extract data from either old or new kwargs format"""
         if key not in kwargs:
             return None
-            
+
         data = kwargs[key]
         # Handle new format: {'key': {'__value__': ...}}
-        if isinstance(data, dict) and '__value__' in data:
-            return data['__value__']
+        if isinstance(data, dict) and "__value__" in data:
+            return data["__value__"]
         # Handle old format: {'key': ...}
         else:
             return data
 
+    def _normalize_trigger_words(self, trigger_words):
+        """Normalize trigger words by splitting by both single and double commas, stripping whitespace, and filtering empty strings"""
+        if not trigger_words or not isinstance(trigger_words, str):
+            return set()
+        
+        # Split by double commas first to preserve groups, then by single commas
+        groups = re.split(r",{2,}", trigger_words)
+        words = []
+        for group in groups:
+            # Split each group by single comma
+            group_words = [word.strip() for word in group.split(",")]
+            words.extend(group_words)
+        
+        # Filter out empty strings and return as set
+        return set(word for word in words if word)
+
     def process_trigger_words(
         self,
         id,
@@ -60,13 +85,25 @@ class TriggerWordToggle:
         **kwargs,
     ):
         # Handle both old and new formats for trigger_words
-        trigger_words_data = self._get_toggle_data(kwargs, 'orinalMessage')
-        trigger_words = trigger_words_data if isinstance(trigger_words_data, str) else ""
-        
+        trigger_words_data = self._get_toggle_data(kwargs, "orinalMessage")
+        trigger_words = (
+            trigger_words_data if isinstance(trigger_words_data, str) else ""
+        )
+
         filtered_triggers = trigger_words
-        
+
+        # Check if trigger_words is provided and different from orinalMessage
+        trigger_words_override = self._get_toggle_data(kwargs, "trigger_words")
+        if (
+            trigger_words_override
+            and isinstance(trigger_words_override, str)
+            and self._normalize_trigger_words(trigger_words_override) != self._normalize_trigger_words(trigger_words)
+        ):
+            filtered_triggers = trigger_words_override
+            return (filtered_triggers,)
+
         # Get toggle data with support for both formats
-        trigger_data = self._get_toggle_data(kwargs, 'toggle_trigger_words')
+        trigger_data = self._get_toggle_data(kwargs, "toggle_trigger_words")
         if trigger_data:
             try:
                 # Convert to list if it's a JSON string
@@ -77,7 +114,9 @@ class TriggerWordToggle:
                     if group_mode:
                         if allow_strength_adjustment:
                             parsed_items = [
-                                self._parse_trigger_item(item, allow_strength_adjustment)
+                                self._parse_trigger_item(
+                                    item, allow_strength_adjustment
+                                )
                                 for item in trigger_data
                             ]
                             filtered_groups = [
@@ -91,11 +130,14 @@ class TriggerWordToggle:
                             ]
                         else:
                             filtered_groups = [
-                                (item.get('text') or "").strip()
+                                (item.get("text") or "").strip()
                                 for item in trigger_data
-                                if (item.get('text') or "").strip() and item.get('active', False)
+                                if (item.get("text") or "").strip()
+                                and item.get("active", False)
                             ]
-                        filtered_triggers = ', '.join(filtered_groups) if filtered_groups else ""
+                        filtered_triggers = (
+                            ", ".join(filtered_groups) if filtered_groups else ""
+                        )
                     else:
                         parsed_items = [
                             self._parse_trigger_item(item, allow_strength_adjustment)
@@ -110,28 +152,34 @@ class TriggerWordToggle:
                             for item in parsed_items
                             if item["text"] and item["active"]
                         ]
-                        filtered_triggers = ', '.join(filtered_words) if filtered_words else ""
+                        filtered_triggers = (
+                            ", ".join(filtered_words) if filtered_words else ""
+                        )
                 else:
                     # Fallback to original message parsing if data is not in the expected list format
                     if group_mode:
-                        groups = re.split(r',{2,}', trigger_words)
+                        groups = re.split(r",{2,}", trigger_words)
                         groups = [group.strip() for group in groups if group.strip()]
-                        filtered_triggers = ', '.join(groups)
+                        filtered_triggers = ", ".join(groups)
                     else:
-                        words = [word.strip() for word in trigger_words.split(',') if word.strip()]
-                        filtered_triggers = ', '.join(words)
+                        words = [
+                            word.strip()
+                            for word in trigger_words.split(",")
+                            if word.strip()
+                        ]
+                        filtered_triggers = ", ".join(words)
 
             except Exception as e:
                 logger.error(f"Error processing trigger words: {e}")
-            
+
         return (filtered_triggers,)
 
     def _parse_trigger_item(self, item, allow_strength_adjustment):
-        text = (item.get('text') or "").strip()
-        active = bool(item.get('active', False))
-        strength = item.get('strength')
+        text = (item.get("text") or "").strip()
+        active = bool(item.get("active", False))
+        strength = item.get("strength")
 
-        strength_match = re.match(r'^\((.+):([\d.]+)\)$', text)
+        strength_match = re.match(r"^\((.+):([\d.]+)\)$", text)
         if strength_match:
             text = strength_match.group(1).strip()
             if strength is None:

py/nodes/wanvideo_lora_select.py

@@ -15,12 +15,9 @@ class WanVideoLoraSelectLM:
             "required": {
                 "low_mem_load": ("BOOLEAN", {"default": False, "tooltip": "Load LORA models with less VRAM usage, slower loading. This affects ALL LoRAs, not just the current ones. No effect if merge_loras is False"}),
                 "merge_loras": ("BOOLEAN", {"default": True, "tooltip": "Merge LoRAs into the model, otherwise they are loaded on the fly. Always disabled for GGUF and scaled fp8 models. This affects ALL LoRAs, not just the current one"}),
-                "text": ("STRING", {
-                    "multiline": True, 
-                    "pysssss.autocomplete": False, 
-                    "dynamicPrompts": True, 
+                "text": ("AUTOCOMPLETE_TEXT_LORAS", {
+                    "placeholder": "Search LoRAs to add...",
                     "tooltip": "Format: <lora:lora_name:strength> separated by spaces or punctuation",
-                    "placeholder": "LoRA syntax input: <lora:name:strength>"
                 }),
             },
             "optional": FlexibleOptionalInputType(any_type),

py/nodes/wanvideo_lora_select_from_text.py

@@ -7,7 +7,7 @@ import logging
 logger = logging.getLogger(__name__)
 
 # 定义新节点的类
-class WanVideoLoraSelectFromText:
+class WanVideoLoraTextSelectLM:
     # 节点在UI中显示的名称
     NAME = "WanVideo Lora Select From Text (LoraManager)"
     # 节点所属的分类
@@ -115,11 +115,3 @@ class WanVideoLoraSelectFromText:
         active_loras_text = " ".join(formatted_loras)
 
         return (loras_list, trigger_words_text, active_loras_text)
-
-NODE_CLASS_MAPPINGS = {
-    "WanVideoLoraSelectFromText": WanVideoLoraSelectFromText
-}
-
-NODE_DISPLAY_NAME_MAPPINGS = {
-    "WanVideoLoraSelectFromText": "WanVideo Lora Select From Text (LoraManager)"
-}

py/routes/example_images_route_registrar.py

@@ -30,6 +30,7 @@ ROUTE_DEFINITIONS: tuple[RouteDefinition, ...] = (
     RouteDefinition("POST", "/api/lm/force-download-example-images", "force_download_example_images"),
     RouteDefinition("POST", "/api/lm/cleanup-example-image-folders", "cleanup_example_image_folders"),
     RouteDefinition("POST", "/api/lm/example-images/set-nsfw-level", "set_example_image_nsfw_level"),
+    RouteDefinition("POST", "/api/lm/check-example-images-needed", "check_example_images_needed"),
 )
 
 

py/routes/handlers/example_images_handlers.py

@@ -92,6 +92,19 @@ class ExampleImagesDownloadHandler:
         except ExampleImagesDownloadError as exc:
             return web.json_response({'success': False, 'error': str(exc)}, status=500)
 
+    async def check_example_images_needed(self, request: web.Request) -> web.StreamResponse:
+        """Lightweight check to see if any models need example images downloaded."""
+        try:
+            payload = await request.json()
+            model_types = payload.get('model_types', ['lora', 'checkpoint', 'embedding'])
+            result = await self._download_manager.check_pending_models(model_types)
+            return web.json_response(result)
+        except Exception as exc:
+            return web.json_response(
+                {'success': False, 'error': str(exc)},
+                status=500
+            )
+
 
 class ExampleImagesManagementHandler:
     """HTTP adapters for import/delete endpoints."""
@@ -161,6 +174,7 @@ class ExampleImagesHandlerSet:
             "resume_example_images": self.download.resume_example_images,
             "stop_example_images": self.download.stop_example_images,
             "force_download_example_images": self.download.force_download_example_images,
+            "check_example_images_needed": self.download.check_example_images_needed,
             "import_example_images": self.management.import_example_images,
             "delete_example_image": self.management.delete_example_image,
             "set_example_image_nsfw_level": self.management.set_example_image_nsfw_level,

py/routes/handlers/misc_handlers.py

@@ -231,6 +231,8 @@ class SettingsHandler:
         "enable_metadata_archive_db",
         "language",
         "use_portable_settings",
+        "onboarding_completed",
+        "dismissed_banners",
         "proxy_enabled",
         "proxy_type",
         "proxy_host",
@@ -253,6 +255,7 @@ class SettingsHandler:
         "model_name_display",
         "update_flag_strategy",
         "auto_organize_exclusions",
+        "filter_presets",
     )
 
     _PROXY_KEYS = {
@@ -1201,6 +1204,80 @@ class FileSystemHandler:
             return web.json_response({"success": False, "error": str(exc)}, status=500)
 
 
+class CustomWordsHandler:
+    """Handler for autocomplete via TagFTSIndex."""
+
+    def __init__(self) -> None:
+        from ...services.custom_words_service import get_custom_words_service
+        self._service = get_custom_words_service()
+
+    async def search_custom_words(self, request: web.Request) -> web.Response:
+        """Search custom words with autocomplete.
+
+        Query parameters:
+            search: The search term to match against.
+            limit: Maximum number of results to return (default: 20).
+            category: Optional category filter. Can be:
+                - A category name (e.g., "character", "artist", "general")
+                - Comma-separated category IDs (e.g., "4,11" for character)
+            enriched: If "true", return enriched results with category and post_count
+                      even without category filtering.
+        """
+        try:
+            search_term = request.query.get("search", "")
+            limit = int(request.query.get("limit", "20"))
+            category_param = request.query.get("category", "")
+            enriched_param = request.query.get("enriched", "").lower() == "true"
+
+            # Parse category parameter
+            categories = None
+            if category_param:
+                categories = self._parse_category_param(category_param)
+
+            results = self._service.search_words(
+                search_term, limit, categories=categories, enriched=enriched_param
+            )
+
+            return web.json_response({
+                "success": True,
+                "words": results
+            })
+        except Exception as exc:
+            logger.error("Error searching custom words: %s", exc, exc_info=True)
+            return web.json_response({"error": str(exc)}, status=500)
+
+    def _parse_category_param(self, param: str) -> list[int] | None:
+        """Parse category parameter into list of category IDs.
+
+        Args:
+            param: Category parameter value (name or comma-separated IDs).
+
+        Returns:
+            List of category IDs, or None if parsing fails.
+        """
+        from ...services.tag_fts_index import CATEGORY_NAME_TO_IDS
+
+        param = param.strip().lower()
+        if not param:
+            return None
+
+        # Try to parse as category name first
+        if param in CATEGORY_NAME_TO_IDS:
+            return CATEGORY_NAME_TO_IDS[param]
+
+        # Try to parse as comma-separated integers
+        try:
+            category_ids = []
+            for part in param.split(","):
+                part = part.strip()
+                if part:
+                    category_ids.append(int(part))
+            return category_ids if category_ids else None
+        except ValueError:
+            logger.debug("Invalid category parameter: %s", param)
+            return None
+
+
 class NodeRegistryHandler:
     def __init__(
         self,
@@ -1427,6 +1504,7 @@ class MiscHandlerSet:
         model_library: ModelLibraryHandler,
         metadata_archive: MetadataArchiveHandler,
         filesystem: FileSystemHandler,
+        custom_words: CustomWordsHandler,
     ) -> None:
         self.health = health
         self.settings = settings
@@ -1438,6 +1516,7 @@ class MiscHandlerSet:
         self.model_library = model_library
         self.metadata_archive = metadata_archive
         self.filesystem = filesystem
+        self.custom_words = custom_words
 
     def to_route_mapping(
         self,
@@ -1465,6 +1544,7 @@ class MiscHandlerSet:
             "get_model_versions_status": self.model_library.get_model_versions_status,
             "open_file_location": self.filesystem.open_file_location,
             "open_settings_location": self.filesystem.open_settings_location,
+            "search_custom_words": self.custom_words.search_custom_words,
         }
 
 

py/routes/handlers/preview_handlers.py

@@ -33,6 +33,10 @@ class PreviewHandler:
             raise web.HTTPBadRequest(text="Invalid preview path encoding") from exc
 
         normalized = decoded_path.replace("\\", "/")
+
+        if not self._config.is_preview_path_allowed(normalized):
+            raise web.HTTPForbidden(text="Preview path is not within an allowed directory")
+
         candidate = Path(normalized)
         try:
             resolved = candidate.expanduser().resolve(strict=False)
@@ -40,14 +44,8 @@ class PreviewHandler:
             logger.debug("Failed to resolve preview path %s: %s", normalized, exc)
             raise web.HTTPBadRequest(text="Unable to resolve preview path") from exc
 
-        resolved_str = str(resolved)
-        # TODO: Temporarily disabled path validation due to issues #772 and #774
-        # Re-enable after fixing preview root path handling
-        # if not self._config.is_preview_path_allowed(resolved_str):
-        #     raise web.HTTPForbidden(text="Preview path is not within an allowed directory")
-
         if not resolved.is_file():
-            logger.debug("Preview file not found at %s", resolved_str)
+            logger.debug("Preview file not found at %s", str(resolved))
             raise web.HTTPNotFound(text="Preview file not found")
 
         # aiohttp's FileResponse handles range requests and content headers for us.

py/routes/handlers/recipe_handlers.py

@@ -412,10 +412,11 @@ class RecipeQueryHandler:
             if recipe_scanner is None:
                 raise RuntimeError("Recipe scanner unavailable")
 
-            duplicate_groups = await recipe_scanner.find_all_duplicate_recipes()
+            fingerprint_groups = await recipe_scanner.find_all_duplicate_recipes()
+            url_groups = await recipe_scanner.find_duplicate_recipes_by_source()
             response_data = []
 
-            for fingerprint, recipe_ids in duplicate_groups.items():
+            for fingerprint, recipe_ids in fingerprint_groups.items():
                 if len(recipe_ids) <= 1:
                     continue
 
@@ -439,12 +440,44 @@ class RecipeQueryHandler:
                     recipes.sort(key=lambda entry: entry.get("modified", 0), reverse=True)
                     response_data.append(
                         {
+                            "type": "fingerprint",
                             "fingerprint": fingerprint,
                             "count": len(recipes),
                             "recipes": recipes,
                         }
                     )
 
+            for url, recipe_ids in url_groups.items():
+                if len(recipe_ids) <= 1:
+                    continue
+
+                recipes = []
+                for recipe_id in recipe_ids:
+                    recipe = await recipe_scanner.get_recipe_by_id(recipe_id)
+                    if recipe:
+                        recipes.append(
+                            {
+                                "id": recipe.get("id"),
+                                "title": recipe.get("title"),
+                                "file_url": recipe.get("file_url")
+                                or self._format_recipe_file_url(recipe.get("file_path", "")),
+                                "modified": recipe.get("modified"),
+                                "created_date": recipe.get("created_date"),
+                                "lora_count": len(recipe.get("loras", [])),
+                            }
+                        )
+
+                if len(recipes) >= 2:
+                    recipes.sort(key=lambda entry: entry.get("modified", 0), reverse=True)
+                    response_data.append(
+                        {
+                            "type": "source_url",
+                            "fingerprint": url,
+                            "count": len(recipes),
+                            "recipes": recipes,
+                        }
+                    )
+
             response_data.sort(key=lambda entry: entry["count"], reverse=True)
             return web.json_response({"success": True, "duplicate_groups": response_data})
         except Exception as exc:
@@ -1021,7 +1054,7 @@ class RecipeManagementHandler:
             "exclude": False,
         }
 
-    async def _download_remote_media(self, image_url: str) -> tuple[bytes, str]:
+    async def _download_remote_media(self, image_url: str) -> tuple[bytes, str, Any]:
         civitai_client = self._civitai_client_getter()
         downloader = await self._downloader_factory()
         temp_path = None
@@ -1029,6 +1062,7 @@ class RecipeManagementHandler:
             with tempfile.NamedTemporaryFile(delete=False) as temp_file:
                 temp_path = temp_file.name
             download_url = image_url
+            image_info = None
             civitai_match = re.match(r"https://civitai\.com/images/(\d+)", image_url)
             if civitai_match:
                 if civitai_client is None:

py/routes/lora_routes.py

@@ -63,6 +63,11 @@ class LoraRoutes(BaseModelRoutes):
             "POST", "/api/lm/{prefix}/random-sample", prefix, self.get_random_loras
         )
 
+        # Cycler routes
+        registrar.add_prefixed_route(
+            "POST", "/api/lm/{prefix}/cycler-list", prefix, self.get_cycler_list
+        )
+
         # ComfyUI integration
         registrar.add_prefixed_route(
             "POST", "/api/lm/{prefix}/get_trigger_words", prefix, self.get_trigger_words
@@ -283,6 +288,29 @@ class LoraRoutes(BaseModelRoutes):
             logger.error(f"Error getting random LoRAs: {e}", exc_info=True)
             return web.json_response({"success": False, "error": str(e)}, status=500)
 
+    async def get_cycler_list(self, request: web.Request) -> web.Response:
+        """Get filtered and sorted LoRA list for cycler widget"""
+        try:
+            json_data = await request.json()
+
+            # Parse parameters
+            pool_config = json_data.get("pool_config")
+            sort_by = json_data.get("sort_by", "filename")
+
+            # Get cycler list from service
+            lora_list = await self.service.get_cycler_list(
+                pool_config=pool_config,
+                sort_by=sort_by
+            )
+
+            return web.json_response(
+                {"success": True, "loras": lora_list, "count": len(lora_list)}
+            )
+
+        except Exception as e:
+            logger.error(f"Error getting cycler list: {e}", exc_info=True)
+            return web.json_response({"success": False, "error": str(e)}, status=500)
+
     async def get_trigger_words(self, request: web.Request) -> web.Response:
         """Get trigger words for specified LoRA models"""
         try:

py/routes/misc_route_registrar.py

@@ -42,6 +42,7 @@ MISC_ROUTE_DEFINITIONS: tuple[RouteDefinition, ...] = (
     RouteDefinition("GET", "/api/lm/metadata-archive-status", "get_metadata_archive_status"),
     RouteDefinition("GET", "/api/lm/model-versions-status", "get_model_versions_status"),
     RouteDefinition("POST", "/api/lm/settings/open-location", "open_settings_location"),
+    RouteDefinition("GET", "/api/lm/custom-words/search", "search_custom_words"),
 )
 
 

py/routes/misc_routes.py

@@ -18,6 +18,7 @@ from ..services.settings_manager import get_settings_manager
 from ..services.downloader import get_downloader
 from ..utils.usage_stats import UsageStats
 from .handlers.misc_handlers import (
+    CustomWordsHandler,
     FileSystemHandler,
     HealthCheckHandler,
     LoraCodeHandler,
@@ -117,6 +118,7 @@ class MiscRoutes:
             service_registry=self._service_registry_adapter,
             metadata_provider_factory=self._metadata_provider_factory,
         )
+        custom_words = CustomWordsHandler()
 
         return self._handler_set_factory(
             health=health,
@@ -129,6 +131,7 @@ class MiscRoutes:
             model_library=model_library,
             metadata_archive=metadata_archive,
             filesystem=filesystem,
+            custom_words=custom_words,
         )
 
 

py/routes/update_routes.py

@@ -45,8 +45,9 @@ class UpdateRoutes:
             # Fetch remote version from GitHub
             if nightly:
                 remote_version, changelog = await UpdateRoutes._get_nightly_version()
+                releases = None
             else:
-                remote_version, changelog = await UpdateRoutes._get_remote_version()
+                remote_version, changelog, releases = await UpdateRoutes._get_remote_version()
             
             # Compare versions
             if nightly:
@@ -59,7 +60,7 @@ class UpdateRoutes:
                     remote_version.replace('v', '')
                 )
             
-            return web.json_response({
+            response_data = {
                 'success': True,
                 'current_version': local_version,
                 'latest_version': remote_version,
@@ -67,7 +68,13 @@ class UpdateRoutes:
                 'changelog': changelog,
                 'git_info': git_info,
                 'nightly': nightly
-            })
+            }
+            
+            # Include releases list for stable mode
+            if releases is not None:
+                response_data['releases'] = releases
+            
+            return web.json_response(response_data)
             
         except NETWORK_EXCEPTIONS as e:
             logger.warning("Network unavailable during update check: %s", e)
@@ -443,42 +450,58 @@ class UpdateRoutes:
         return git_info
     
     @staticmethod
-    async def _get_remote_version() -> tuple[str, List[str]]:
+    async def _get_remote_version() -> tuple[str, List[str], List[Dict]]:
         """
         Fetch remote version from GitHub
         Returns:
-            tuple: (version string, changelog list)
+            tuple: (version string, changelog list, releases list)
         """
         repo_owner = "willmiao"
         repo_name = "ComfyUI-Lora-Manager"
         
-        # Use GitHub API to fetch the latest release
-        github_url = f"https://api.github.com/repos/{repo_owner}/{repo_name}/releases/latest"
+        # Use GitHub API to fetch the last 5 releases
+        github_url = f"https://api.github.com/repos/{repo_owner}/{repo_name}/releases?per_page=5"
         
         try:
             downloader = await get_downloader()
             success, data = await downloader.make_request('GET', github_url, custom_headers={'Accept': 'application/vnd.github+json'})
             
             if not success:
-                logger.warning(f"Failed to fetch GitHub release: {data}")
-                return "v0.0.0", []
+                logger.warning(f"Failed to fetch GitHub releases: {data}")
+                return "v0.0.0", [], []
             
-            version = data.get('tag_name', '')
-            if not version.startswith('v'):
-                version = f"v{version}"
+            # Parse releases
+            releases = []
+            for i, release in enumerate(data):
+                version = release.get('tag_name', '')
+                if not version.startswith('v'):
+                    version = f"v{version}"
+                
+                # Extract changelog from release notes
+                body = release.get('body', '')
+                changelog = UpdateRoutes._parse_changelog(body)
+                
+                releases.append({
+                    'version': version,
+                    'changelog': changelog,
+                    'published_at': release.get('published_at', ''),
+                    'is_latest': i == 0
+                })
             
-            # Extract changelog from release notes
-            body = data.get('body', '')
-            changelog = UpdateRoutes._parse_changelog(body)
+            # Get latest version and its changelog
+            if releases:
+                latest_version = releases[0]['version']
+                latest_changelog = releases[0]['changelog']
+                return latest_version, latest_changelog, releases
             
-            return version, changelog
+            return "v0.0.0", [], []
         
         except NETWORK_EXCEPTIONS as e:
             logger.warning("Unable to reach GitHub for release info: %s", e)
-            return "v0.0.0", []
+            return "v0.0.0", [], []
         except Exception as e:
             logger.error(f"Error fetching remote version: {e}", exc_info=True)
-            return "v0.0.0", []
+            return "v0.0.0", [], []
     
     @staticmethod
     def _parse_changelog(release_notes: str) -> List[str]:

py/services/base_model_service.py

@@ -5,7 +5,7 @@ import logging
 import os
 import time
 
-from ..utils.constants import VALID_LORA_TYPES
+from ..utils.constants import VALID_LORA_SUB_TYPES, VALID_CHECKPOINT_SUB_TYPES
 from ..utils.models import BaseModelMetadata
 from ..utils.metadata_manager import MetadataManager
 from ..utils.usage_stats import UsageStats
@@ -15,8 +15,8 @@ from .model_query import (
     ModelFilterSet,
     SearchStrategy,
     SettingsProvider,
-    normalize_civitai_model_type,
-    resolve_civitai_model_type,
+    normalize_sub_type,
+    resolve_sub_type,
 )
 from .settings_manager import get_settings_manager
 
@@ -568,16 +568,21 @@ class BaseModelService(ABC):
         return await self.scanner.get_base_models(limit)
 
     async def get_model_types(self, limit: int = 20) -> List[Dict[str, Any]]:
-        """Get counts of normalized CivitAI model types present in the cache."""
+        """Get counts of sub-types present in the cache."""
         cache = await self.scanner.get_cached_data()
 
         type_counts: Dict[str, int] = {}
         for entry in cache.raw_data:
-            normalized_type = normalize_civitai_model_type(
-                resolve_civitai_model_type(entry)
-            )
-            if not normalized_type or normalized_type not in VALID_LORA_TYPES:
+            normalized_type = normalize_sub_type(resolve_sub_type(entry))
+            if not normalized_type:
                 continue
+            
+            # Filter by valid sub-types based on scanner type
+            if self.model_type == "lora" and normalized_type not in VALID_LORA_SUB_TYPES:
+                continue
+            if self.model_type == "checkpoint" and normalized_type not in VALID_CHECKPOINT_SUB_TYPES:
+                continue
+            
             type_counts[normalized_type] = type_counts.get(normalized_type, 0) + 1
 
         sorted_types = sorted(

py/services/cache_entry_validator.py

@@ -0,0 +1,259 @@
+"""
+Cache Entry Validator
+
+Validates and repairs cache entries to prevent runtime errors from
+missing or invalid critical fields.
+"""
+
+from dataclasses import dataclass, field
+from typing import Any, Dict, List, Optional, Tuple
+import logging
+import os
+
+logger = logging.getLogger(__name__)
+
+
+@dataclass
+class ValidationResult:
+    """Result of validating a single cache entry."""
+    is_valid: bool
+    repaired: bool
+    errors: List[str] = field(default_factory=list)
+    entry: Optional[Dict[str, Any]] = None
+
+
+class CacheEntryValidator:
+    """
+    Validates and repairs cache entry core fields.
+
+    Critical fields that cause runtime errors when missing:
+    - file_path: KeyError in multiple locations
+    - sha256: KeyError/AttributeError in hash operations
+
+    Medium severity fields that may cause sorting/display issues:
+    - size: KeyError during sorting
+    - modified: KeyError during sorting
+    - model_name: AttributeError on .lower() calls
+
+    Low severity fields:
+    - tags: KeyError/TypeError in recipe operations
+    """
+
+    # Field definitions: (default_value, is_required)
+    CORE_FIELDS: Dict[str, Tuple[Any, bool]] = {
+        'file_path': ('', True),
+        'sha256': ('', True),
+        'file_name': ('', False),
+        'model_name': ('', False),
+        'folder': ('', False),
+        'size': (0, False),
+        'modified': (0.0, False),
+        'tags': ([], False),
+        'preview_url': ('', False),
+        'base_model': ('', False),
+        'from_civitai': (True, False),
+        'favorite': (False, False),
+        'exclude': (False, False),
+        'db_checked': (False, False),
+        'preview_nsfw_level': (0, False),
+        'notes': ('', False),
+        'usage_tips': ('', False),
+    }
+
+    @classmethod
+    def validate(cls, entry: Dict[str, Any], *, auto_repair: bool = True) -> ValidationResult:
+        """
+        Validate a single cache entry.
+
+        Args:
+            entry: The cache entry dictionary to validate
+            auto_repair: If True, attempt to repair missing/invalid fields
+
+        Returns:
+            ValidationResult with validation status and optionally repaired entry
+        """
+        if entry is None:
+            return ValidationResult(
+                is_valid=False,
+                repaired=False,
+                errors=['Entry is None'],
+                entry=None
+            )
+
+        if not isinstance(entry, dict):
+            return ValidationResult(
+                is_valid=False,
+                repaired=False,
+                errors=[f'Entry is not a dict: {type(entry).__name__}'],
+                entry=None
+            )
+
+        errors: List[str] = []
+        repaired = False
+        working_entry = dict(entry) if auto_repair else entry
+
+        for field_name, (default_value, is_required) in cls.CORE_FIELDS.items():
+            value = working_entry.get(field_name)
+
+            # Check if field is missing or None
+            if value is None:
+                if is_required:
+                    errors.append(f"Required field '{field_name}' is missing or None")
+                if auto_repair:
+                    working_entry[field_name] = cls._get_default_copy(default_value)
+                    repaired = True
+                continue
+
+            # Validate field type and value
+            field_error = cls._validate_field(field_name, value, default_value)
+            if field_error:
+                errors.append(field_error)
+                if auto_repair:
+                    working_entry[field_name] = cls._get_default_copy(default_value)
+                    repaired = True
+
+        # Special validation: file_path must not be empty for required field
+        file_path = working_entry.get('file_path', '')
+        if not file_path or (isinstance(file_path, str) and not file_path.strip()):
+            errors.append("Required field 'file_path' is empty")
+            # Cannot repair empty file_path - entry is invalid
+            return ValidationResult(
+                is_valid=False,
+                repaired=repaired,
+                errors=errors,
+                entry=working_entry if auto_repair else None
+            )
+
+        # Special validation: sha256 must not be empty for required field
+        sha256 = working_entry.get('sha256', '')
+        if not sha256 or (isinstance(sha256, str) and not sha256.strip()):
+            errors.append("Required field 'sha256' is empty")
+            # Cannot repair empty sha256 - entry is invalid
+            return ValidationResult(
+                is_valid=False,
+                repaired=repaired,
+                errors=errors,
+                entry=working_entry if auto_repair else None
+            )
+
+        # Normalize sha256 to lowercase if needed
+        if isinstance(sha256, str):
+            normalized_sha = sha256.lower().strip()
+            if normalized_sha != sha256:
+                working_entry['sha256'] = normalized_sha
+                repaired = True
+
+        # Determine if entry is valid
+        # Entry is valid if no critical required field errors remain after repair
+        # Critical fields are file_path and sha256
+        CRITICAL_REQUIRED_FIELDS = {'file_path', 'sha256'}
+        has_critical_errors = any(
+            "Required field" in error and
+            any(f"'{field}'" in error for field in CRITICAL_REQUIRED_FIELDS)
+            for error in errors
+        )
+
+        is_valid = not has_critical_errors
+
+        return ValidationResult(
+            is_valid=is_valid,
+            repaired=repaired,
+            errors=errors,
+            entry=working_entry if auto_repair else entry
+        )
+
+    @classmethod
+    def validate_batch(
+        cls,
+        entries: List[Dict[str, Any]],
+        *,
+        auto_repair: bool = True
+    ) -> Tuple[List[Dict[str, Any]], List[Dict[str, Any]]]:
+        """
+        Validate a batch of cache entries.
+
+        Args:
+            entries: List of cache entry dictionaries to validate
+            auto_repair: If True, attempt to repair missing/invalid fields
+
+        Returns:
+            Tuple of (valid_entries, invalid_entries)
+        """
+        if not entries:
+            return [], []
+
+        valid_entries: List[Dict[str, Any]] = []
+        invalid_entries: List[Dict[str, Any]] = []
+
+        for entry in entries:
+            result = cls.validate(entry, auto_repair=auto_repair)
+
+            if result.is_valid:
+                # Use repaired entry if available, otherwise original
+                valid_entries.append(result.entry if result.entry else entry)
+            else:
+                invalid_entries.append(entry)
+                # Log invalid entries for debugging
+                file_path = entry.get('file_path', '<unknown>') if isinstance(entry, dict) else '<not a dict>'
+                logger.warning(
+                    f"Invalid cache entry for '{file_path}': {', '.join(result.errors)}"
+                )
+
+        return valid_entries, invalid_entries
+
+    @classmethod
+    def _validate_field(cls, field_name: str, value: Any, default_value: Any) -> Optional[str]:
+        """
+        Validate a specific field value.
+
+        Returns an error message if invalid, None if valid.
+        """
+        expected_type = type(default_value)
+
+        # Special handling for numeric types
+        if expected_type == int:
+            if not isinstance(value, (int, float)):
+                return f"Field '{field_name}' should be numeric, got {type(value).__name__}"
+        elif expected_type == float:
+            if not isinstance(value, (int, float)):
+                return f"Field '{field_name}' should be numeric, got {type(value).__name__}"
+        elif expected_type == bool:
+            # Be lenient with boolean fields - accept truthy/falsy values
+            pass
+        elif expected_type == str:
+            if not isinstance(value, str):
+                return f"Field '{field_name}' should be string, got {type(value).__name__}"
+        elif expected_type == list:
+            if not isinstance(value, (list, tuple)):
+                return f"Field '{field_name}' should be list, got {type(value).__name__}"
+
+        return None
+
+    @classmethod
+    def _get_default_copy(cls, default_value: Any) -> Any:
+        """Get a copy of the default value to avoid shared mutable state."""
+        if isinstance(default_value, list):
+            return list(default_value)
+        if isinstance(default_value, dict):
+            return dict(default_value)
+        return default_value
+
+    @classmethod
+    def get_file_path_safe(cls, entry: Dict[str, Any], default: str = '') -> str:
+        """Safely get file_path from an entry."""
+        if not isinstance(entry, dict):
+            return default
+        value = entry.get('file_path')
+        if isinstance(value, str):
+            return value
+        return default
+
+    @classmethod
+    def get_sha256_safe(cls, entry: Dict[str, Any], default: str = '') -> str:
+        """Safely get sha256 from an entry."""
+        if not isinstance(entry, dict):
+            return default
+        value = entry.get('sha256')
+        if isinstance(value, str):
+            return value.lower()
+        return default

py/services/cache_health_monitor.py

@@ -0,0 +1,201 @@
+"""
+Cache Health Monitor
+
+Monitors cache health status and determines when user intervention is needed.
+"""
+
+from dataclasses import dataclass, field
+from enum import Enum
+from typing import Any, Dict, List, Optional
+import logging
+
+from .cache_entry_validator import CacheEntryValidator, ValidationResult
+
+logger = logging.getLogger(__name__)
+
+
+class CacheHealthStatus(Enum):
+    """Health status of the cache."""
+    HEALTHY = "healthy"
+    DEGRADED = "degraded"
+    CORRUPTED = "corrupted"
+
+
+@dataclass
+class HealthReport:
+    """Report of cache health check."""
+    status: CacheHealthStatus
+    total_entries: int
+    valid_entries: int
+    invalid_entries: int
+    repaired_entries: int
+    invalid_paths: List[str] = field(default_factory=list)
+    message: str = ""
+
+    @property
+    def corruption_rate(self) -> float:
+        """Calculate the percentage of invalid entries."""
+        if self.total_entries <= 0:
+            return 0.0
+        return self.invalid_entries / self.total_entries
+
+    def to_dict(self) -> Dict[str, Any]:
+        """Convert to dictionary for JSON serialization."""
+        return {
+            'status': self.status.value,
+            'total_entries': self.total_entries,
+            'valid_entries': self.valid_entries,
+            'invalid_entries': self.invalid_entries,
+            'repaired_entries': self.repaired_entries,
+            'corruption_rate': f"{self.corruption_rate:.1%}",
+            'invalid_paths': self.invalid_paths[:10],  # Limit to first 10
+            'message': self.message,
+        }
+
+
+class CacheHealthMonitor:
+    """
+    Monitors cache health and determines appropriate status.
+
+    Thresholds:
+    - HEALTHY: 0% invalid entries
+    - DEGRADED: 0-5% invalid entries (auto-repaired, user should rebuild)
+    - CORRUPTED: >5% invalid entries (significant data loss likely)
+    """
+
+    # Threshold percentages
+    DEGRADED_THRESHOLD = 0.01   # 1% - show warning
+    CORRUPTED_THRESHOLD = 0.05  # 5% - critical warning
+
+    def __init__(
+        self,
+        *,
+        degraded_threshold: float = DEGRADED_THRESHOLD,
+        corrupted_threshold: float = CORRUPTED_THRESHOLD
+    ):
+        """
+        Initialize the health monitor.
+
+        Args:
+            degraded_threshold: Corruption rate threshold for DEGRADED status
+            corrupted_threshold: Corruption rate threshold for CORRUPTED status
+        """
+        self.degraded_threshold = degraded_threshold
+        self.corrupted_threshold = corrupted_threshold
+
+    def check_health(
+        self,
+        entries: List[Dict[str, Any]],
+        *,
+        auto_repair: bool = True
+    ) -> HealthReport:
+        """
+        Check the health of cache entries.
+
+        Args:
+            entries: List of cache entry dictionaries to check
+            auto_repair: If True, attempt to repair entries during validation
+
+        Returns:
+            HealthReport with status and statistics
+        """
+        if not entries:
+            return HealthReport(
+                status=CacheHealthStatus.HEALTHY,
+                total_entries=0,
+                valid_entries=0,
+                invalid_entries=0,
+                repaired_entries=0,
+                message="Cache is empty"
+            )
+
+        total_entries = len(entries)
+        valid_entries: List[Dict[str, Any]] = []
+        invalid_entries: List[Dict[str, Any]] = []
+        repaired_count = 0
+        invalid_paths: List[str] = []
+
+        for entry in entries:
+            result = CacheEntryValidator.validate(entry, auto_repair=auto_repair)
+
+            if result.is_valid:
+                valid_entries.append(result.entry if result.entry else entry)
+                if result.repaired:
+                    repaired_count += 1
+            else:
+                invalid_entries.append(entry)
+                # Extract file path for reporting
+                file_path = CacheEntryValidator.get_file_path_safe(entry, '<unknown>')
+                invalid_paths.append(file_path)
+
+        invalid_count = len(invalid_entries)
+        valid_count = len(valid_entries)
+
+        # Determine status based on corruption rate
+        corruption_rate = invalid_count / total_entries if total_entries > 0 else 0.0
+
+        if invalid_count == 0:
+            status = CacheHealthStatus.HEALTHY
+            message = "Cache is healthy"
+        elif corruption_rate >= self.corrupted_threshold:
+            status = CacheHealthStatus.CORRUPTED
+            message = (
+                f"Cache is corrupted: {invalid_count} invalid entries "
+                f"({corruption_rate:.1%}). Rebuild recommended."
+            )
+        elif corruption_rate >= self.degraded_threshold or invalid_count > 0:
+            status = CacheHealthStatus.DEGRADED
+            message = (
+                f"Cache has {invalid_count} invalid entries "
+                f"({corruption_rate:.1%}). Consider rebuilding cache."
+            )
+        else:
+            # This shouldn't happen, but handle gracefully
+            status = CacheHealthStatus.HEALTHY
+            message = "Cache is healthy"
+
+        # Log the health check result
+        if status != CacheHealthStatus.HEALTHY:
+            logger.warning(
+                f"Cache health check: {status.value} - "
+                f"{invalid_count}/{total_entries} invalid, "
+                f"{repaired_count} repaired"
+            )
+            if invalid_paths:
+                logger.debug(f"Invalid entry paths: {invalid_paths[:5]}")
+
+        return HealthReport(
+            status=status,
+            total_entries=total_entries,
+            valid_entries=valid_count,
+            invalid_entries=invalid_count,
+            repaired_entries=repaired_count,
+            invalid_paths=invalid_paths,
+            message=message
+        )
+
+    def should_notify_user(self, report: HealthReport) -> bool:
+        """
+        Determine if the user should be notified about cache health.
+
+        Args:
+            report: The health report to evaluate
+
+        Returns:
+            True if user should be notified
+        """
+        return report.status != CacheHealthStatus.HEALTHY
+
+    def get_notification_severity(self, report: HealthReport) -> str:
+        """
+        Get the severity level for user notification.
+
+        Args:
+            report: The health report to evaluate
+
+        Returns:
+            Severity string: 'warning' or 'error'
+        """
+        if report.status == CacheHealthStatus.CORRUPTED:
+            return 'error'
+        return 'warning'

py/services/checkpoint_scanner.py

@@ -21,7 +21,8 @@ class CheckpointScanner(ModelScanner):
             hash_index=ModelHashIndex()
         )
 
-    def _resolve_model_type(self, root_path: Optional[str]) -> Optional[str]:
+    def _resolve_sub_type(self, root_path: Optional[str]) -> Optional[str]:
+        """Resolve the sub-type based on the root path."""
         if not root_path:
             return None
 
@@ -34,18 +35,19 @@ class CheckpointScanner(ModelScanner):
         return None
 
     def adjust_metadata(self, metadata, file_path, root_path):
-        if hasattr(metadata, "model_type"):
-            model_type = self._resolve_model_type(root_path)
-            if model_type:
-                metadata.model_type = model_type
+        """Adjust metadata during scanning to set sub_type."""
+        sub_type = self._resolve_sub_type(root_path)
+        if sub_type:
+            metadata.sub_type = sub_type
         return metadata
 
     def adjust_cached_entry(self, entry: Dict[str, Any]) -> Dict[str, Any]:
-        model_type = self._resolve_model_type(
+        """Adjust entries loaded from the persisted cache to ensure sub_type is set."""
+        sub_type = self._resolve_sub_type(
             self._find_root_for_file(entry.get("file_path"))
         )
-        if model_type:
-            entry["model_type"] = model_type
+        if sub_type:
+            entry["sub_type"] = sub_type
         return entry
 
     def get_model_roots(self) -> List[str]:

py/services/checkpoint_service.py

@@ -22,6 +22,9 @@ class CheckpointService(BaseModelService):
     
     async def format_response(self, checkpoint_data: Dict) -> Dict:
         """Format Checkpoint data for API response"""
+        # Get sub_type from cache entry (new canonical field)
+        sub_type = checkpoint_data.get("sub_type", "checkpoint")
+        
         return {
             "model_name": checkpoint_data["model_name"],
             "file_name": checkpoint_data["file_name"],
@@ -37,7 +40,7 @@ class CheckpointService(BaseModelService):
             "from_civitai": checkpoint_data.get("from_civitai", True),
             "usage_count": checkpoint_data.get("usage_count", 0),
             "notes": checkpoint_data.get("notes", ""),
-            "model_type": checkpoint_data.get("model_type", "checkpoint"),
+            "sub_type": sub_type,
             "favorite": checkpoint_data.get("favorite", False),
             "update_available": bool(checkpoint_data.get("update_available", False)),
             "civitai": self.filter_civitai_data(checkpoint_data.get("civitai", {}), minimal=True)

py/services/custom_words_service.py

@@ -0,0 +1,91 @@
+"""Service for managing autocomplete via TagFTSIndex.
+
+This service provides full-text search capabilities for Danbooru/e621 tags
+with category filtering and enriched results including post counts.
+"""
+
+from __future__ import annotations
+
+import logging
+from typing import List, Dict, Any, Optional
+
+logger = logging.getLogger(__name__)
+
+
+class CustomWordsService:
+    """Service for autocomplete via TagFTSIndex.
+
+    This service:
+    - Uses TagFTSIndex for fast full-text search of Danbooru/e621 tags
+    - Supports category-based filtering
+    - Returns enriched results with category and post_count
+    - Provides sub-100ms search times for 221k+ tags
+    """
+
+    _instance: Optional[CustomWordsService] = None
+    _initialized: bool = False
+
+    def __new__(cls) -> CustomWordsService:
+        if cls._instance is None:
+            cls._instance = super().__new__(cls)
+        return cls._instance
+
+    def __init__(self) -> None:
+        if self._initialized:
+            return
+
+        self._tag_index: Optional[Any] = None
+        self._initialized = True
+
+    @classmethod
+    def get_instance(cls) -> CustomWordsService:
+        """Get the singleton instance of CustomWordsService."""
+        if cls._instance is None:
+            cls._instance = cls()
+        return cls._instance
+
+    def _get_tag_index(self):
+        """Get or create the TagFTSIndex instance (lazy initialization)."""
+        if self._tag_index is None:
+            try:
+                from .tag_fts_index import get_tag_fts_index
+                self._tag_index = get_tag_fts_index()
+            except Exception as e:
+                logger.warning(f"Failed to initialize TagFTSIndex: {e}")
+                self._tag_index = None
+        return self._tag_index
+
+    def search_words(
+        self,
+        search_term: str,
+        limit: int = 20,
+        categories: Optional[List[int]] = None,
+        enriched: bool = False
+    ) -> List[Dict[str, Any]]:
+        """Search tags using TagFTSIndex with category filtering.
+
+        Args:
+            search_term: The search term to match against.
+            limit: Maximum number of results to return.
+            categories: Optional list of category IDs to filter by.
+            enriched: If True, always return enriched results with category
+                       and post_count (default behavior now).
+
+        Returns:
+            List of dicts with tag_name, category, and post_count.
+        """
+        tag_index = self._get_tag_index()
+        if tag_index is not None:
+            results = tag_index.search(search_term, categories=categories, limit=limit)
+            return results
+
+        logger.debug("TagFTSIndex not available, returning empty results")
+        return []
+
+
+def get_custom_words_service() -> CustomWordsService:
+    """Factory function to get the CustomWordsService singleton."""
+    return CustomWordsService.get_instance()
+
+
+__all__ = ["CustomWordsService", "get_custom_words_service"]

py/services/embedding_service.py

@@ -22,6 +22,9 @@ class EmbeddingService(BaseModelService):
     
     async def format_response(self, embedding_data: Dict) -> Dict:
         """Format Embedding data for API response"""
+        # Get sub_type from cache entry (new canonical field)
+        sub_type = embedding_data.get("sub_type", "embedding")
+        
         return {
             "model_name": embedding_data["model_name"],
             "file_name": embedding_data["file_name"],
@@ -37,7 +40,7 @@ class EmbeddingService(BaseModelService):
             "from_civitai": embedding_data.get("from_civitai", True),
             # "usage_count": embedding_data.get("usage_count", 0), # TODO: Enable when embedding usage tracking is implemented
             "notes": embedding_data.get("notes", ""),
-            "model_type": embedding_data.get("model_type", "embedding"),
+            "sub_type": sub_type,
             "favorite": embedding_data.get("favorite", False),
             "update_available": bool(embedding_data.get("update_available", False)),
             "civitai": self.filter_civitai_data(embedding_data.get("civitai", {}), minimal=True)

py/services/lora_scanner.py

@@ -30,36 +30,36 @@ class LoraScanner(ModelScanner):
 
     async def diagnose_hash_index(self):
         """Diagnostic method to verify hash index functionality"""
-        print("\n\n*** DIAGNOSING LORA HASH INDEX ***\n\n", file=sys.stderr)
+        logger.debug("\n\n*** DIAGNOSING LORA HASH INDEX ***\n\n")
         
         # First check if the hash index has any entries
         if hasattr(self, '_hash_index'):
             index_entries = len(self._hash_index._hash_to_path)
-            print(f"Hash index has {index_entries} entries", file=sys.stderr)
+            logger.debug(f"Hash index has {index_entries} entries")
             
             # Print a few example entries if available
             if index_entries > 0:
-                print("\nSample hash index entries:", file=sys.stderr)
+                logger.debug("\nSample hash index entries:")
                 count = 0
                 for hash_val, path in self._hash_index._hash_to_path.items():
                     if count < 5:  # Just show the first 5
-                        print(f"Hash: {hash_val[:8]}... -> Path: {path}", file=sys.stderr)
+                        logger.debug(f"Hash: {hash_val[:8]}... -> Path: {path}")
                         count += 1
                     else:
                         break
         else:
-            print("Hash index not initialized", file=sys.stderr)
+            logger.debug("Hash index not initialized")
         
         # Try looking up by a known hash for testing
         if not hasattr(self, '_hash_index') or not self._hash_index._hash_to_path:
-            print("No hash entries to test lookup with", file=sys.stderr)
+            logger.debug("No hash entries to test lookup with")
             return
         
         test_hash = next(iter(self._hash_index._hash_to_path.keys()))
         test_path = self._hash_index.get_path(test_hash)
-        print(f"\nTest lookup by hash: {test_hash[:8]}... -> {test_path}", file=sys.stderr)
+        logger.debug(f"\nTest lookup by hash: {test_hash[:8]}... -> {test_path}")
         
         # Also test reverse lookup
         test_hash_result = self._hash_index.get_hash(test_path)
-        print(f"Test reverse lookup: {test_path} -> {test_hash_result[:8]}...\n\n", file=sys.stderr)
+        logger.debug(f"Test reverse lookup: {test_path} -> {test_hash_result[:8]}...\n\n")
 

py/services/lora_service.py

@@ -3,6 +3,7 @@ import logging
 from typing import Dict, List, Optional
 
 from .base_model_service import BaseModelService
+from .model_query import resolve_sub_type
 from ..utils.models import LoraMetadata
 from ..config import config
 
@@ -23,6 +24,10 @@ class LoraService(BaseModelService):
 
     async def format_response(self, lora_data: Dict) -> Dict:
         """Format LoRA data for API response"""
+        # Resolve sub_type using priority: sub_type > model_type > civitai.model.type > default
+        # Normalize to lowercase for consistent API responses
+        sub_type = resolve_sub_type(lora_data).lower()
+        
         return {
             "model_name": lora_data["model_name"],
             "file_name": lora_data["file_name"],
@@ -43,6 +48,7 @@ class LoraService(BaseModelService):
             "notes": lora_data.get("notes", ""),
             "favorite": lora_data.get("favorite", False),
             "update_available": bool(lora_data.get("update_available", False)),
+            "sub_type": sub_type,
             "civitai": self.filter_civitai_data(
                 lora_data.get("civitai", {}), minimal=True
             ),
@@ -479,3 +485,49 @@ class LoraService(BaseModelService):
             ]
 
         return available_loras
+
+    async def get_cycler_list(
+        self,
+        pool_config: Optional[Dict] = None,
+        sort_by: str = "filename"
+    ) -> List[Dict]:
+        """
+        Get filtered and sorted LoRA list for cycling.
+
+        Args:
+            pool_config: Optional pool config for filtering (filters dict)
+            sort_by: Sort field - 'filename' or 'model_name'
+
+        Returns:
+            List of LoRA dicts with file_name and model_name
+        """
+        # Get cached data
+        cache = await self.scanner.get_cached_data(force_refresh=False)
+        available_loras = cache.raw_data if cache else []
+
+        # Apply pool filters if provided
+        if pool_config:
+            available_loras = await self._apply_pool_filters(
+                available_loras, pool_config
+            )
+
+        # Sort by specified field
+        if sort_by == "model_name":
+            available_loras = sorted(
+                available_loras,
+                key=lambda x: (x.get("model_name") or x.get("file_name", "")).lower()
+            )
+        else:  # Default to filename
+            available_loras = sorted(
+                available_loras,
+                key=lambda x: x.get("file_name", "").lower()
+            )
+
+        # Return minimal data needed for cycling
+        return [
+            {
+                "file_name": lora["file_name"],
+                "model_name": lora.get("model_name", lora["file_name"]),
+            }
+            for lora in available_loras
+        ]

py/services/metadata_service.py

@@ -44,6 +44,8 @@ async def initialize_metadata_providers():
                 logger.debug(f"SQLite metadata provider registered with database: {db_path}")
             else:
                 logger.warning("Metadata archive database is enabled but database file not found")
+                logger.info("Automatically disabling enable_metadata_archive_db setting")
+                settings_manager.set('enable_metadata_archive_db', False)
         except Exception as e:
             logger.error(f"Failed to initialize SQLite metadata provider: {e}")
     

py/services/metadata_sync_service.py

@@ -243,17 +243,27 @@ class MetadataSyncService:
                 last_error = error or last_error
 
             if civitai_metadata is None or metadata_provider is None:
+                # Track if we need to save metadata
+                needs_save = False
+
                 if sqlite_attempted:
                     model_data["db_checked"] = True
+                    needs_save = True
 
                 if civitai_api_not_found:
                     model_data["from_civitai"] = False
                     model_data["civitai_deleted"] = True
                     model_data["db_checked"] = sqlite_attempted or (enable_archive and model_data.get("db_checked", False))
                     model_data["last_checked_at"] = datetime.now().timestamp()
+                    needs_save = True
 
+                # Save metadata if any state was updated
+                if needs_save:
                     data_to_save = model_data.copy()
                     data_to_save.pop("folder", None)
+                    # Update last_checked_at for sqlite-only attempts if not already set
+                    if "last_checked_at" not in data_to_save:
+                        data_to_save["last_checked_at"] = datetime.now().timestamp()
                     await self._metadata_manager.save_metadata(file_path, data_to_save)
 
                 default_error = (

py/services/model_cache.py

@@ -5,7 +5,6 @@ import logging
 logger = logging.getLogger(__name__)
 from typing import Any, Dict, List, Optional, Tuple
 from dataclasses import dataclass, field
-from operator import itemgetter
 from natsort import natsorted
 
 # Supported sort modes: (sort_key, order)
@@ -229,17 +228,17 @@ class ModelCache:
                 reverse=reverse
             )
         elif sort_key == 'date':
-            # Sort by modified timestamp
+            # Sort by modified timestamp (use .get() with default to handle missing fields)
             result = sorted(
                 data,
-                key=itemgetter('modified'),
+                key=lambda x: x.get('modified', 0.0),
                 reverse=reverse
             )
         elif sort_key == 'size':
-            # Sort by file size
+            # Sort by file size (use .get() with default to handle missing fields)
             result = sorted(
                 data,
-                key=itemgetter('size'),
+                key=lambda x: x.get('size', 0),
                 reverse=reverse
             )
         elif sort_key == 'usage':

py/services/model_metadata_provider.py

@@ -676,10 +676,12 @@ class ModelMetadataProviderManager:
         
     def _get_provider(self, provider_name: str = None) -> ModelMetadataProvider:
         """Get provider by name or default provider"""
-        if provider_name and provider_name in self.providers:
+        if provider_name:
+            if provider_name not in self.providers:
+                raise ValueError(f"Provider '{provider_name}' is not registered")
             return self.providers[provider_name]
-        
+
         if self.default_provider is None:
             raise ValueError("No default provider set and no valid provider specified")
-            
+
         return self.providers[self.default_provider]

py/services/model_query.py

@@ -33,28 +33,42 @@ def _coerce_to_str(value: Any) -> Optional[str]:
     return candidate if candidate else None
 
 
-def normalize_civitai_model_type(value: Any) -> Optional[str]:
-    """Return a lowercase string suitable for comparisons."""
+def normalize_sub_type(value: Any) -> Optional[str]:
+    """Return a lowercase string suitable for sub_type comparisons."""
     candidate = _coerce_to_str(value)
     return candidate.lower() if candidate else None
 
 
-def resolve_civitai_model_type(entry: Mapping[str, Any]) -> str:
-    """Extract the model type from CivitAI metadata, defaulting to LORA."""
+def resolve_sub_type(entry: Mapping[str, Any]) -> str:
+    """Extract the sub-type from metadata, checking multiple sources.
+    
+    Priority:
+    1. entry['sub_type'] - new canonical field
+    2. entry['model_type'] - backward compatibility
+    3. civitai.model.type - CivitAI API data
+    4. DEFAULT_CIVITAI_MODEL_TYPE - fallback
+    """
     if not isinstance(entry, Mapping):
         return DEFAULT_CIVITAI_MODEL_TYPE
 
+    # Priority 1: Check new canonical field 'sub_type'
+    sub_type = _coerce_to_str(entry.get("sub_type"))
+    if sub_type:
+        return sub_type
+
+    # Priority 2: Backward compatibility - check 'model_type' field
+    model_type = _coerce_to_str(entry.get("model_type"))
+    if model_type:
+        return model_type
+
+    # Priority 3: Extract from CivitAI metadata
     civitai = entry.get("civitai")
     if isinstance(civitai, Mapping):
         civitai_model = civitai.get("model")
         if isinstance(civitai_model, Mapping):
-            model_type = _coerce_to_str(civitai_model.get("type"))
-            if model_type:
-                return model_type
-
-    model_type = _coerce_to_str(entry.get("model_type"))
-    if model_type:
-        return model_type
+            civitai_type = _coerce_to_str(civitai_model.get("type"))
+            if civitai_type:
+                return civitai_type
 
     return DEFAULT_CIVITAI_MODEL_TYPE
 
@@ -313,7 +327,7 @@ class ModelFilterSet:
             normalized_model_types = {
                 model_type
                 for model_type in (
-                    normalize_civitai_model_type(value) for value in model_types
+                    normalize_sub_type(value) for value in model_types
                 )
                 if model_type
             }
@@ -321,7 +335,7 @@ class ModelFilterSet:
                 items = [
                     item
                     for item in items
-                    if normalize_civitai_model_type(resolve_civitai_model_type(item))
+                    if normalize_sub_type(resolve_sub_type(item))
                     in normalized_model_types
                 ]
             model_types_duration = time.perf_counter() - t0

py/services/model_scanner.py

@@ -20,6 +20,8 @@ from .service_registry import ServiceRegistry
 from .websocket_manager import ws_manager
 from .persistent_model_cache import get_persistent_cache
 from .settings_manager import get_settings_manager
+from .cache_entry_validator import CacheEntryValidator
+from .cache_health_monitor import CacheHealthMonitor, CacheHealthStatus
 
 logger = logging.getLogger(__name__)
 
@@ -275,9 +277,10 @@ class ModelScanner:
         _, license_flags = resolve_license_info(license_source or {})
         entry['license_flags'] = license_flags
 
-        model_type = get_value('model_type', None)
-        if model_type:
-            entry['model_type'] = model_type
+        # Handle sub_type (new canonical field)
+        sub_type = get_value('sub_type', None)
+        if sub_type:
+            entry['sub_type'] = sub_type
 
         return entry
 
@@ -467,6 +470,39 @@ class ModelScanner:
             for tag in adjusted_item.get('tags') or []:
                 tags_count[tag] = tags_count.get(tag, 0) + 1
 
+        # Validate cache entries and check health
+        valid_entries, invalid_entries = CacheEntryValidator.validate_batch(
+            adjusted_raw_data, auto_repair=True
+        )
+
+        if invalid_entries:
+            monitor = CacheHealthMonitor()
+            report = monitor.check_health(adjusted_raw_data, auto_repair=True)
+
+            if report.status != CacheHealthStatus.HEALTHY:
+                # Broadcast health warning to frontend
+                await ws_manager.broadcast_cache_health_warning(report, page_type)
+                logger.warning(
+                    f"{self.model_type.capitalize()} Scanner: Cache health issue detected - "
+                    f"{report.invalid_entries} invalid entries, {report.repaired_entries} repaired"
+                )
+
+            # Use only valid entries
+            adjusted_raw_data = valid_entries
+
+            # Rebuild tags count from valid entries only
+            tags_count = {}
+            for item in adjusted_raw_data:
+                for tag in item.get('tags') or []:
+                    tags_count[tag] = tags_count.get(tag, 0) + 1
+
+            # Remove invalid entries from hash index
+            for invalid_entry in invalid_entries:
+                file_path = CacheEntryValidator.get_file_path_safe(invalid_entry)
+                sha256 = CacheEntryValidator.get_sha256_safe(invalid_entry)
+                if file_path:
+                    hash_index.remove_by_path(file_path, sha256)
+
         scan_result = CacheBuildResult(
             raw_data=adjusted_raw_data,
             hash_index=hash_index,
@@ -650,7 +686,6 @@ class ModelScanner:
 
     async def _initialize_cache(self) -> None:
         """Initialize or refresh the cache"""
-        print("init start", flush=True)
         self._is_initializing = True  # Set flag
         try:
             start_time = time.time()
@@ -664,7 +699,6 @@ class ModelScanner:
             scan_result = await self._gather_model_data()
             await self._apply_scan_result(scan_result)
             await self._save_persistent_cache(scan_result)
-            print("init end", flush=True)
 
             logger.info(
                 f"{self.model_type.capitalize()} Scanner: Cache initialization completed in {time.time() - start_time:.2f} seconds, "
@@ -775,6 +809,18 @@ class ModelScanner:
                                     model_data = self.adjust_cached_entry(dict(model_data))
                                     if not model_data:
                                         continue
+
+                                    # Validate the new entry before adding
+                                    validation_result = CacheEntryValidator.validate(
+                                        model_data, auto_repair=True
+                                    )
+                                    if not validation_result.is_valid:
+                                        logger.warning(
+                                            f"Skipping invalid entry during reconcile: {path}"
+                                        )
+                                        continue
+                                    model_data = validation_result.entry
+
                                     self._ensure_license_flags(model_data)
                                     # Add to cache
                                     self._cache.raw_data.append(model_data)
@@ -1089,6 +1135,17 @@ class ModelScanner:
                             processed_files += 1
 
                             if result:
+                                # Validate the entry before adding
+                                validation_result = CacheEntryValidator.validate(
+                                    result, auto_repair=True
+                                )
+                                if not validation_result.is_valid:
+                                    logger.warning(
+                                        f"Skipping invalid scan result: {file_path}"
+                                    )
+                                    continue
+                                result = validation_result.entry
+
                                 self._ensure_license_flags(result)
                                 raw_data.append(result)
 

py/services/persistent_model_cache.py

@@ -1,13 +1,12 @@
 import json
 import logging
 import os
-import re
 import sqlite3
 import threading
 from dataclasses import dataclass
 from typing import Dict, List, Mapping, Optional, Sequence, Tuple
 
-from ..utils.settings_paths import get_project_root, get_settings_dir
+from ..utils.cache_paths import CacheType, resolve_cache_path_with_migration
 
 logger = logging.getLogger(__name__)
 
@@ -404,20 +403,12 @@ class PersistentModelCache:
     # Internal helpers -------------------------------------------------
 
     def _resolve_default_path(self, library_name: str) -> str:
-        override = os.environ.get("LORA_MANAGER_CACHE_DB")
-        if override:
-            return override
-        try:
-            settings_dir = get_settings_dir(create=True)
-        except Exception as exc:  # pragma: no cover - defensive guard
-            logger.warning("Falling back to project directory for cache: %s", exc)
-            settings_dir = get_project_root()
-        safe_name = re.sub(r"[^A-Za-z0-9_.-]", "_", library_name or "default")
-        if safe_name.lower() in ("default", ""):
-            legacy_path = os.path.join(settings_dir, self._DEFAULT_FILENAME)
-            if os.path.exists(legacy_path):
-                return legacy_path
-        return os.path.join(settings_dir, "model_cache", f"{safe_name}.sqlite")
+        env_override = os.environ.get("LORA_MANAGER_CACHE_DB")
+        return resolve_cache_path_with_migration(
+            CacheType.MODEL,
+            library_name=library_name,
+            env_override=env_override,
+        )
 
     def _initialize_schema(self) -> None:
         with self._db_lock:

py/services/persistent_recipe_cache.py

@@ -0,0 +1,484 @@
+"""SQLite-based persistent cache for recipe metadata.
+
+This module provides fast recipe cache persistence using SQLite, enabling
+quick startup by loading from cache instead of walking directories and
+parsing JSON files.
+"""
+
+from __future__ import annotations
+
+import json
+import logging
+import os
+import sqlite3
+import threading
+from dataclasses import dataclass
+from typing import Dict, List, Optional, Set, Tuple
+
+from ..utils.cache_paths import CacheType, resolve_cache_path_with_migration
+
+logger = logging.getLogger(__name__)
+
+
+@dataclass
+class PersistedRecipeData:
+    """Lightweight structure returned by the persistent recipe cache."""
+
+    raw_data: List[Dict]
+    file_stats: Dict[str, Tuple[float, int]]  # json_path -> (mtime, size)
+
+
+class PersistentRecipeCache:
+    """Persist recipe metadata in SQLite for fast startup."""
+
+    _DEFAULT_FILENAME = "recipe_cache.sqlite"
+    _RECIPE_COLUMNS: Tuple[str, ...] = (
+        "recipe_id",
+        "file_path",
+        "json_path",
+        "title",
+        "folder",
+        "base_model",
+        "fingerprint",
+        "created_date",
+        "modified",
+        "file_mtime",
+        "file_size",
+        "favorite",
+        "repair_version",
+        "preview_nsfw_level",
+        "loras_json",
+        "checkpoint_json",
+        "gen_params_json",
+        "tags_json",
+    )
+    _instances: Dict[str, "PersistentRecipeCache"] = {}
+    _instance_lock = threading.Lock()
+
+    def __init__(self, library_name: str = "default", db_path: Optional[str] = None) -> None:
+        self._library_name = library_name or "default"
+        self._db_path = db_path or self._resolve_default_path(self._library_name)
+        self._db_lock = threading.Lock()
+        self._schema_initialized = False
+        try:
+            directory = os.path.dirname(self._db_path)
+            if directory:
+                os.makedirs(directory, exist_ok=True)
+        except Exception as exc:
+            logger.warning("Could not create recipe cache directory %s: %s", directory, exc)
+        if self.is_enabled():
+            self._initialize_schema()
+
+    @classmethod
+    def get_default(cls, library_name: Optional[str] = None) -> "PersistentRecipeCache":
+        name = library_name or "default"
+        with cls._instance_lock:
+            if name not in cls._instances:
+                cls._instances[name] = cls(name)
+            return cls._instances[name]
+
+    @classmethod
+    def clear_instances(cls) -> None:
+        """Clear all cached instances (useful for library switching)."""
+        with cls._instance_lock:
+            cls._instances.clear()
+
+    def is_enabled(self) -> bool:
+        return os.environ.get("LORA_MANAGER_DISABLE_PERSISTENT_CACHE", "0") != "1"
+
+    def get_database_path(self) -> str:
+        """Expose the resolved SQLite database path."""
+        return self._db_path
+
+    def load_cache(self) -> Optional[PersistedRecipeData]:
+        """Load all cached recipes from SQLite.
+
+        Returns:
+            PersistedRecipeData with raw_data and file_stats if cache exists,
+            None if cache is empty or unavailable.
+        """
+        if not self.is_enabled():
+            return None
+        if not self._schema_initialized:
+            self._initialize_schema()
+        if not self._schema_initialized:
+            return None
+
+        try:
+            with self._db_lock:
+                conn = self._connect(readonly=True)
+                try:
+                    # Load all recipes
+                    columns_sql = ", ".join(self._RECIPE_COLUMNS)
+                    rows = conn.execute(f"SELECT {columns_sql} FROM recipes").fetchall()
+
+                    if not rows:
+                        return None
+
+                finally:
+                    conn.close()
+        except FileNotFoundError:
+            return None
+        except Exception as exc:
+            logger.warning("Failed to load persisted recipe cache: %s", exc)
+            return None
+
+        raw_data: List[Dict] = []
+        file_stats: Dict[str, Tuple[float, int]] = {}
+
+        for row in rows:
+            recipe = self._row_to_recipe(row)
+            raw_data.append(recipe)
+
+            json_path = row["json_path"]
+            if json_path:
+                file_stats[json_path] = (
+                    row["file_mtime"] or 0.0,
+                    row["file_size"] or 0,
+                )
+
+        return PersistedRecipeData(raw_data=raw_data, file_stats=file_stats)
+
+    def save_cache(self, recipes: List[Dict], json_paths: Optional[Dict[str, str]] = None) -> None:
+        """Save all recipes to SQLite cache.
+
+        Args:
+            recipes: List of recipe dictionaries to persist.
+            json_paths: Optional mapping of recipe_id -> json_path for file stats.
+        """
+        if not self.is_enabled():
+            return
+        if not self._schema_initialized:
+            self._initialize_schema()
+        if not self._schema_initialized:
+            return
+
+        try:
+            with self._db_lock:
+                conn = self._connect()
+                try:
+                    conn.execute("PRAGMA foreign_keys = ON")
+                    conn.execute("BEGIN")
+
+                    # Clear existing data
+                    conn.execute("DELETE FROM recipes")
+
+                    # Prepare and insert all rows
+                    recipe_rows = []
+                    for recipe in recipes:
+                        recipe_id = str(recipe.get("id", ""))
+                        if not recipe_id:
+                            continue
+
+                        json_path = ""
+                        if json_paths:
+                            json_path = json_paths.get(recipe_id, "")
+
+                        row = self._prepare_recipe_row(recipe, json_path)
+                        recipe_rows.append(row)
+
+                    if recipe_rows:
+                        placeholders = ", ".join(["?"] * len(self._RECIPE_COLUMNS))
+                        columns = ", ".join(self._RECIPE_COLUMNS)
+                        conn.executemany(
+                            f"INSERT INTO recipes ({columns}) VALUES ({placeholders})",
+                            recipe_rows,
+                        )
+
+                    conn.commit()
+                    logger.debug("Persisted %d recipes to cache", len(recipe_rows))
+                finally:
+                    conn.close()
+        except Exception as exc:
+            logger.warning("Failed to persist recipe cache: %s", exc)
+
+    def get_file_stats(self) -> Dict[str, Tuple[float, int]]:
+        """Return stored file stats for all cached recipes.
+
+        Returns:
+            Dictionary mapping json_path -> (mtime, size).
+        """
+        if not self.is_enabled() or not self._schema_initialized:
+            return {}
+
+        try:
+            with self._db_lock:
+                conn = self._connect(readonly=True)
+                try:
+                    rows = conn.execute(
+                        "SELECT json_path, file_mtime, file_size FROM recipes WHERE json_path IS NOT NULL"
+                    ).fetchall()
+                    return {
+                        row["json_path"]: (row["file_mtime"] or 0.0, row["file_size"] or 0)
+                        for row in rows
+                        if row["json_path"]
+                    }
+                finally:
+                    conn.close()
+        except Exception:
+            return {}
+
+    def update_recipe(self, recipe: Dict, json_path: Optional[str] = None) -> None:
+        """Update or insert a single recipe in the cache.
+
+        Args:
+            recipe: The recipe dictionary to persist.
+            json_path: Optional path to the recipe JSON file.
+        """
+        if not self.is_enabled() or not self._schema_initialized:
+            return
+
+        recipe_id = str(recipe.get("id", ""))
+        if not recipe_id:
+            return
+
+        try:
+            with self._db_lock:
+                conn = self._connect()
+                try:
+                    row = self._prepare_recipe_row(recipe, json_path or "")
+                    placeholders = ", ".join(["?"] * len(self._RECIPE_COLUMNS))
+                    columns = ", ".join(self._RECIPE_COLUMNS)
+                    conn.execute(
+                        f"INSERT OR REPLACE INTO recipes ({columns}) VALUES ({placeholders})",
+                        row,
+                    )
+                    conn.commit()
+                finally:
+                    conn.close()
+        except Exception as exc:
+            logger.debug("Failed to update recipe %s in cache: %s", recipe_id, exc)
+
+    def remove_recipe(self, recipe_id: str) -> None:
+        """Remove a recipe from the cache by ID.
+
+        Args:
+            recipe_id: The ID of the recipe to remove.
+        """
+        if not self.is_enabled() or not self._schema_initialized:
+            return
+
+        if not recipe_id:
+            return
+
+        try:
+            with self._db_lock:
+                conn = self._connect()
+                try:
+                    conn.execute("DELETE FROM recipes WHERE recipe_id = ?", (str(recipe_id),))
+                    conn.commit()
+                finally:
+                    conn.close()
+        except Exception as exc:
+            logger.debug("Failed to remove recipe %s from cache: %s", recipe_id, exc)
+
+    def get_indexed_recipe_ids(self) -> Set[str]:
+        """Return all recipe IDs in the cache.
+
+        Returns:
+            Set of recipe ID strings.
+        """
+        if not self.is_enabled() or not self._schema_initialized:
+            return set()
+
+        try:
+            with self._db_lock:
+                conn = self._connect(readonly=True)
+                try:
+                    rows = conn.execute("SELECT recipe_id FROM recipes").fetchall()
+                    return {row["recipe_id"] for row in rows if row["recipe_id"]}
+                finally:
+                    conn.close()
+        except Exception:
+            return set()
+
+    def get_recipe_count(self) -> int:
+        """Return the number of recipes in the cache."""
+        if not self.is_enabled() or not self._schema_initialized:
+            return 0
+
+        try:
+            with self._db_lock:
+                conn = self._connect(readonly=True)
+                try:
+                    result = conn.execute("SELECT COUNT(*) FROM recipes").fetchone()
+                    return result[0] if result else 0
+                finally:
+                    conn.close()
+        except Exception:
+            return 0
+
+    # Internal helpers
+
+    def _resolve_default_path(self, library_name: str) -> str:
+        env_override = os.environ.get("LORA_MANAGER_RECIPE_CACHE_DB")
+        return resolve_cache_path_with_migration(
+            CacheType.RECIPE,
+            library_name=library_name,
+            env_override=env_override,
+        )
+
+    def _initialize_schema(self) -> None:
+        with self._db_lock:
+            if self._schema_initialized:
+                return
+            try:
+                with self._connect() as conn:
+                    conn.execute("PRAGMA journal_mode=WAL")
+                    conn.execute("PRAGMA foreign_keys = ON")
+                    conn.executescript(
+                        """
+                        CREATE TABLE IF NOT EXISTS recipes (
+                            recipe_id TEXT PRIMARY KEY,
+                            file_path TEXT,
+                            json_path TEXT,
+                            title TEXT,
+                            folder TEXT,
+                            base_model TEXT,
+                            fingerprint TEXT,
+                            created_date REAL,
+                            modified REAL,
+                            file_mtime REAL,
+                            file_size INTEGER,
+                            favorite INTEGER DEFAULT 0,
+                            repair_version INTEGER DEFAULT 0,
+                            preview_nsfw_level INTEGER DEFAULT 0,
+                            loras_json TEXT,
+                            checkpoint_json TEXT,
+                            gen_params_json TEXT,
+                            tags_json TEXT
+                        );
+
+                        CREATE INDEX IF NOT EXISTS idx_recipes_json_path ON recipes(json_path);
+                        CREATE INDEX IF NOT EXISTS idx_recipes_fingerprint ON recipes(fingerprint);
+
+                        CREATE TABLE IF NOT EXISTS cache_metadata (
+                            key TEXT PRIMARY KEY,
+                            value TEXT
+                        );
+                        """
+                    )
+                    conn.commit()
+                self._schema_initialized = True
+            except Exception as exc:
+                logger.warning("Failed to initialize persistent recipe cache schema: %s", exc)
+
+    def _connect(self, readonly: bool = False) -> sqlite3.Connection:
+        uri = False
+        path = self._db_path
+        if readonly:
+            if not os.path.exists(path):
+                raise FileNotFoundError(path)
+            path = f"file:{path}?mode=ro"
+            uri = True
+        conn = sqlite3.connect(path, check_same_thread=False, uri=uri, detect_types=sqlite3.PARSE_DECLTYPES)
+        conn.row_factory = sqlite3.Row
+        return conn
+
+    def _prepare_recipe_row(self, recipe: Dict, json_path: str) -> Tuple:
+        """Convert a recipe dict to a row tuple for SQLite insertion."""
+        loras = recipe.get("loras")
+        loras_json = json.dumps(loras) if loras else None
+
+        checkpoint = recipe.get("checkpoint")
+        checkpoint_json = json.dumps(checkpoint) if checkpoint else None
+
+        gen_params = recipe.get("gen_params")
+        gen_params_json = json.dumps(gen_params) if gen_params else None
+
+        tags = recipe.get("tags")
+        tags_json = json.dumps(tags) if tags else None
+
+        # Get file stats if json_path exists
+        file_mtime = 0.0
+        file_size = 0
+        if json_path and os.path.exists(json_path):
+            try:
+                stat = os.stat(json_path)
+                file_mtime = stat.st_mtime
+                file_size = stat.st_size
+            except OSError:
+                pass
+
+        return (
+            str(recipe.get("id", "")),
+            recipe.get("file_path"),
+            json_path,
+            recipe.get("title"),
+            recipe.get("folder"),
+            recipe.get("base_model"),
+            recipe.get("fingerprint"),
+            float(recipe.get("created_date") or 0.0),
+            float(recipe.get("modified") or 0.0),
+            file_mtime,
+            file_size,
+            1 if recipe.get("favorite") else 0,
+            int(recipe.get("repair_version") or 0),
+            int(recipe.get("preview_nsfw_level") or 0),
+            loras_json,
+            checkpoint_json,
+            gen_params_json,
+            tags_json,
+        )
+
+    def _row_to_recipe(self, row: sqlite3.Row) -> Dict:
+        """Convert a SQLite row to a recipe dictionary."""
+        loras = []
+        if row["loras_json"]:
+            try:
+                loras = json.loads(row["loras_json"])
+            except json.JSONDecodeError:
+                pass
+
+        checkpoint = None
+        if row["checkpoint_json"]:
+            try:
+                checkpoint = json.loads(row["checkpoint_json"])
+            except json.JSONDecodeError:
+                pass
+
+        gen_params = {}
+        if row["gen_params_json"]:
+            try:
+                gen_params = json.loads(row["gen_params_json"])
+            except json.JSONDecodeError:
+                pass
+
+        tags = []
+        if row["tags_json"]:
+            try:
+                tags = json.loads(row["tags_json"])
+            except json.JSONDecodeError:
+                pass
+
+        recipe = {
+            "id": row["recipe_id"],
+            "file_path": row["file_path"] or "",
+            "title": row["title"] or "",
+            "folder": row["folder"] or "",
+            "base_model": row["base_model"] or "",
+            "fingerprint": row["fingerprint"] or "",
+            "created_date": row["created_date"] or 0.0,
+            "modified": row["modified"] or 0.0,
+            "favorite": bool(row["favorite"]),
+            "repair_version": row["repair_version"] or 0,
+            "preview_nsfw_level": row["preview_nsfw_level"] or 0,
+            "loras": loras,
+            "gen_params": gen_params,
+        }
+
+        if tags:
+            recipe["tags"] = tags
+
+        if checkpoint:
+            recipe["checkpoint"] = checkpoint
+
+        return recipe
+
+
+def get_persistent_recipe_cache() -> PersistentRecipeCache:
+    """Get the default persistent recipe cache instance for the active library."""
+    from .settings_manager import get_settings_manager
+
+    library_name = get_settings_manager().get_active_library_name()
+    return PersistentRecipeCache.get_default(library_name)

py/services/recipe_fts_index.py

@@ -15,7 +15,7 @@ import threading
 import time
 from typing import Any, Dict, List, Optional, Set
 
-from ..utils.settings_paths import get_settings_dir
+from ..utils.cache_paths import CacheType, resolve_cache_path_with_migration
 
 logger = logging.getLogger(__name__)
 
@@ -67,17 +67,11 @@ class RecipeFTSIndex:
 
     def _resolve_default_path(self) -> str:
         """Resolve the default database path."""
-        override = os.environ.get("LORA_MANAGER_RECIPE_FTS_DB")
-        if override:
-            return override
-
-        try:
-            settings_dir = get_settings_dir(create=True)
-        except Exception as exc:
-            logger.warning("Falling back to current directory for FTS index: %s", exc)
-            settings_dir = "."
-
-        return os.path.join(settings_dir, self._DEFAULT_FILENAME)
+        env_override = os.environ.get("LORA_MANAGER_RECIPE_FTS_DB")
+        return resolve_cache_path_with_migration(
+            CacheType.RECIPE_FTS,
+            env_override=env_override,
+        )
 
     def get_database_path(self) -> str:
         """Return the resolved database path."""
@@ -403,6 +397,78 @@ class RecipeFTSIndex:
         except Exception:
             return 0
 
+    def get_indexed_recipe_ids(self) -> Set[str]:
+        """Return all recipe IDs currently in the index.
+
+        Returns:
+            Set of recipe ID strings.
+        """
+        if not self._schema_initialized:
+            self.initialize()
+
+        if not self._schema_initialized:
+            return set()
+
+        try:
+            with self._lock:
+                conn = self._connect(readonly=True)
+                try:
+                    cursor = conn.execute("SELECT recipe_id FROM recipe_fts")
+                    return {row[0] for row in cursor.fetchall() if row[0]}
+                finally:
+                    conn.close()
+        except FileNotFoundError:
+            return set()
+        except Exception as exc:
+            logger.debug("Failed to get indexed recipe IDs: %s", exc)
+            return set()
+
+    def validate_index(self, recipe_count: int, recipe_ids: Set[str]) -> bool:
+        """Check if the FTS index matches the expected recipes.
+
+        This method validates whether the existing FTS index can be reused
+        without a full rebuild. It checks:
+        1. The index has been initialized
+        2. The count matches
+        3. The recipe IDs match
+
+        Args:
+            recipe_count: Expected number of recipes.
+            recipe_ids: Expected set of recipe IDs.
+
+        Returns:
+            True if the index is valid and can be reused, False otherwise.
+        """
+        if not self._schema_initialized:
+            self.initialize()
+
+        if not self._schema_initialized:
+            return False
+
+        try:
+            indexed_count = self.get_indexed_count()
+            if indexed_count != recipe_count:
+                logger.debug(
+                    "FTS index count mismatch: indexed=%d, expected=%d",
+                    indexed_count, recipe_count
+                )
+                return False
+
+            indexed_ids = self.get_indexed_recipe_ids()
+            if indexed_ids != recipe_ids:
+                missing = recipe_ids - indexed_ids
+                extra = indexed_ids - recipe_ids
+                if missing:
+                    logger.debug("FTS index missing %d recipe IDs", len(missing))
+                if extra:
+                    logger.debug("FTS index has %d extra recipe IDs", len(extra))
+                return False
+
+            return True
+        except Exception as exc:
+            logger.debug("FTS index validation failed: %s", exc)
+            return False
+
     # Internal helpers
 
     def _connect(self, readonly: bool = False) -> sqlite3.Connection:
@@ -509,21 +575,20 @@ class RecipeFTSIndex:
         if not fields:
             return term_expr
 
-        # Build field-restricted query with OR between fields
+        # Build field-restricted query where ALL words must match within at least one field
         field_clauses = []
         for field in fields:
             if field in self.FIELD_MAP:
                 cols = self.FIELD_MAP[field]
                 for col in cols:
-                    # FTS5 column filter syntax: column:term
-                    # Need to handle multiple terms properly
-                    for term in prefix_terms:
-                        field_clauses.append(f'{col}:{term}')
+                    # Create clause where ALL terms must match in this column (implicit AND)
+                    col_terms = [f'{col}:{term}' for term in prefix_terms]
+                    field_clauses.append('(' + ' '.join(col_terms) + ')')
 
         if not field_clauses:
             return term_expr
 
-        # Combine field clauses with OR
+        # Any field matching all terms is acceptable (OR between field clauses)
         return ' OR '.join(field_clauses)
 
     def _escape_fts_query(self, text: str) -> str:

py/services/recipe_scanner.py

@@ -9,6 +9,7 @@ from typing import Any, Callable, Dict, Iterable, List, Optional, Set, Tuple
 from ..config import config
 from .recipe_cache import RecipeCache
 from .recipe_fts_index import RecipeFTSIndex
+from .persistent_recipe_cache import PersistentRecipeCache, get_persistent_recipe_cache, PersistedRecipeData
 from .service_registry import ServiceRegistry
 from .lora_scanner import LoraScanner
 from .metadata_service import get_default_metadata_provider
@@ -78,6 +79,9 @@ class RecipeScanner:
             # FTS index for fast search
             self._fts_index: Optional[RecipeFTSIndex] = None
             self._fts_index_task: Optional[asyncio.Task] = None
+            # Persistent cache for fast startup
+            self._persistent_cache: Optional[PersistentRecipeCache] = None
+            self._json_path_map: Dict[str, str] = {}  # recipe_id -> json_path
             if lora_scanner:
                 self._lora_scanner = lora_scanner
             if checkpoint_scanner:
@@ -109,6 +113,11 @@ class RecipeScanner:
             self._fts_index.clear()
         self._fts_index = None
 
+        # Reset persistent cache instance for new library
+        self._persistent_cache = None
+        self._json_path_map = {}
+        PersistentRecipeCache.clear_instances()
+
         self._cache = None
         self._initialization_task = None
         self._is_initializing = False
@@ -321,12 +330,17 @@ class RecipeScanner:
             with open(recipe_json_path, 'w', encoding='utf-8') as f:
                 json.dump(recipe, f, indent=4, ensure_ascii=False)
 
-            # 4. Update EXIF if image exists
+            # 4. Update persistent SQLite cache
+            if self._persistent_cache:
+                self._persistent_cache.update_recipe(recipe, recipe_json_path)
+                self._json_path_map[str(recipe_id)] = recipe_json_path
+
+            # 5. Update EXIF if image exists
             image_path = recipe.get('file_path')
             if image_path and os.path.exists(image_path):
                 from ..utils.exif_utils import ExifUtils
                 ExifUtils.append_recipe_metadata(image_path, recipe)
-            
+
             return True
         except Exception as e:
             logger.error(f"Error persisting recipe {recipe_id}: {e}")
@@ -408,117 +422,281 @@ class RecipeScanner:
             logger.error(f"Recipe Scanner: Error initializing cache in background: {e}")
     
     def _initialize_recipe_cache_sync(self):
-        """Synchronous version of recipe cache initialization for thread pool execution"""
+        """Synchronous version of recipe cache initialization for thread pool execution.
+
+        Uses persistent cache for fast startup when available:
+        1. Try to load from persistent SQLite cache
+        2. Reconcile with filesystem (check mtime/size for changes)
+        3. Fall back to full directory scan if cache miss or reconciliation fails
+        4. Persist results for next startup
+        """
         try:
+            # Ensure cache exists to avoid None reference errors
+            if self._cache is None:
+                self._cache = RecipeCache(
+                    raw_data=[],
+                    sorted_by_name=[],
+                    sorted_by_date=[],
+                    folders=[],
+                    folder_tree={},
+                )
+
             # Create a new event loop for this thread
             loop = asyncio.new_event_loop()
             asyncio.set_event_loop(loop)
-            
-            # Create a synchronous method to bypass the async lock
-            def sync_initialize_cache():
-                # We need to implement scan_all_recipes logic synchronously here
-                # instead of calling the async method to avoid event loop issues
-                recipes = []
-                recipes_dir = self.recipes_dir
-                
-                if not recipes_dir or not os.path.exists(recipes_dir):
-                    logger.warning(f"Recipes directory not found: {recipes_dir}")
-                    return recipes
-                
-                # Get all recipe JSON files in the recipes directory
-                recipe_files = []
-                for root, _, files in os.walk(recipes_dir):
-                    recipe_count = sum(1 for f in files if f.lower().endswith('.recipe.json'))
-                    if recipe_count > 0:
-                        for file in files:
-                            if file.lower().endswith('.recipe.json'):
-                                recipe_files.append(os.path.join(root, file))
-                
-                # Process each recipe file
-                for recipe_path in recipe_files:
-                    try:
-                        with open(recipe_path, 'r', encoding='utf-8') as f:
-                            recipe_data = json.load(f)
-                        
-                        # Validate recipe data
-                        if not recipe_data or not isinstance(recipe_data, dict):
-                            logger.warning(f"Invalid recipe data in {recipe_path}")
-                            continue
-                        
-                        # Ensure required fields exist
-                        required_fields = ['id', 'file_path', 'title']
-                        if not all(field in recipe_data for field in required_fields):
-                            logger.warning(f"Missing required fields in {recipe_path}")
-                            continue
-                        
-                        # Ensure the image file exists and prioritize local siblings
-                        image_path = recipe_data.get('file_path')
-                        if image_path:
-                            recipe_dir = os.path.dirname(recipe_path)
-                            image_filename = os.path.basename(image_path)
-                            local_sibling_path = os.path.normpath(os.path.join(recipe_dir, image_filename))
-                            
-                            # If local sibling exists and stored path is different, prefer local
-                            if os.path.exists(local_sibling_path) and os.path.normpath(image_path) != local_sibling_path:
-                                recipe_data['file_path'] = local_sibling_path
-                                # Persist the repair
-                                try:
-                                    with open(recipe_path, 'w', encoding='utf-8') as f:
-                                        json.dump(recipe_data, f, indent=4, ensure_ascii=False)
-                                    logger.info(f"Updated recipe image path to local sibling: {local_sibling_path}")
-                                except Exception as e:
-                                    logger.warning(f"Failed to persist repair for {recipe_path}: {e}")
-                            elif not os.path.exists(image_path):
-                                logger.warning(f"Recipe image not found and no local sibling: {image_path}")
-                        
-                        # Ensure loras array exists
-                        if 'loras' not in recipe_data:
-                            recipe_data['loras'] = []
-                        
-                        # Ensure gen_params exists
-                        if 'gen_params' not in recipe_data:
-                            recipe_data['gen_params'] = {}
-                        
-                        # Add to list without async operations
-                        recipes.append(recipe_data)
-                    except Exception as e:
-                        logger.error(f"Error loading recipe file {recipe_path}: {e}")
-                        import traceback
-                        traceback.print_exc(file=sys.stderr)
-                
-                # Update cache with the collected data
-                self._cache.raw_data = recipes
-                self._update_folder_metadata(self._cache)
-                
-                # Create a simplified resort function that doesn't use await
-                if hasattr(self._cache, "resort"):
-                    try:
-                        # Sort by name
-                        self._cache.sorted_by_name = natsorted(
-                            self._cache.raw_data,
-                            key=lambda x: x.get('title', '').lower()
-                        )
-                        
-                        # Sort by date (modified or created)
-                        self._cache.sorted_by_date = sorted(
-                            self._cache.raw_data,
-                            key=lambda x: x.get('modified', x.get('created_date', 0)),
-                            reverse=True
-                        )
-                    except Exception as e:
-                        logger.error(f"Error sorting recipe cache: {e}")
-                
+
+            # Initialize persistent cache
+            if self._persistent_cache is None:
+                self._persistent_cache = get_persistent_recipe_cache()
+
+            recipes_dir = self.recipes_dir
+            if not recipes_dir or not os.path.exists(recipes_dir):
+                logger.warning(f"Recipes directory not found: {recipes_dir}")
                 return self._cache
-            
-            # Run our sync initialization that avoids lock conflicts
-            return sync_initialize_cache()
+
+            # Try to load from persistent cache first
+            persisted = self._persistent_cache.load_cache()
+            if persisted:
+                recipes, changed, json_paths = self._reconcile_recipe_cache(persisted, recipes_dir)
+                self._json_path_map = json_paths
+
+                if not changed:
+                    # Fast path: use cached data directly
+                    logger.info("Recipe cache hit: loaded %d recipes from persistent cache", len(recipes))
+                    self._cache.raw_data = recipes
+                    self._update_folder_metadata(self._cache)
+                    self._sort_cache_sync()
+                    return self._cache
+                else:
+                    # Partial update: some files changed
+                    logger.info("Recipe cache partial hit: reconciled %d recipes with filesystem", len(recipes))
+                    self._cache.raw_data = recipes
+                    self._update_folder_metadata(self._cache)
+                    self._sort_cache_sync()
+                    # Persist updated cache
+                    self._persistent_cache.save_cache(recipes, json_paths)
+                    return self._cache
+
+            # Fall back to full directory scan
+            logger.info("Recipe cache miss: performing full directory scan")
+            recipes, json_paths = self._full_directory_scan_sync(recipes_dir)
+            self._json_path_map = json_paths
+
+            # Update cache with the collected data
+            self._cache.raw_data = recipes
+            self._update_folder_metadata(self._cache)
+            self._sort_cache_sync()
+
+            # Persist for next startup
+            self._persistent_cache.save_cache(recipes, json_paths)
+
+            return self._cache
         except Exception as e:
             logger.error(f"Error in thread-based recipe cache initialization: {e}")
+            import traceback
+            traceback.print_exc(file=sys.stderr)
             return self._cache if hasattr(self, '_cache') else None
         finally:
             # Clean up the event loop
             loop.close()
 
+    def _reconcile_recipe_cache(
+        self,
+        persisted: PersistedRecipeData,
+        recipes_dir: str,
+    ) -> Tuple[List[Dict], bool, Dict[str, str]]:
+        """Reconcile persisted cache with current filesystem state.
+
+        Args:
+            persisted: The persisted recipe data from SQLite cache.
+            recipes_dir: Path to the recipes directory.
+
+        Returns:
+            Tuple of (recipes list, changed flag, json_paths dict).
+        """
+        recipes: List[Dict] = []
+        json_paths: Dict[str, str] = {}
+        changed = False
+
+        # Build set of current recipe files
+        current_files: Dict[str, Tuple[float, int]] = {}
+        for root, _, files in os.walk(recipes_dir):
+            for file in files:
+                if file.lower().endswith('.recipe.json'):
+                    file_path = os.path.join(root, file)
+                    try:
+                        stat = os.stat(file_path)
+                        current_files[file_path] = (stat.st_mtime, stat.st_size)
+                    except OSError:
+                        continue
+
+        # Build recipe_id -> recipe lookup (O(n) instead of O(n²))
+        recipe_by_id: Dict[str, Dict] = {
+            str(r.get('id', '')): r for r in persisted.raw_data if r.get('id')
+        }
+
+        # Build json_path -> recipe lookup from file_stats (O(m))
+        persisted_by_path: Dict[str, Dict] = {}
+        for json_path in persisted.file_stats.keys():
+            basename = os.path.basename(json_path)
+            if basename.lower().endswith('.recipe.json'):
+                recipe_id = basename[:-len('.recipe.json')]
+                if recipe_id in recipe_by_id:
+                    persisted_by_path[json_path] = recipe_by_id[recipe_id]
+
+        # Process current files
+        for file_path, (current_mtime, current_size) in current_files.items():
+            cached_stats = persisted.file_stats.get(file_path)
+
+            # Extract recipe_id from current file for fallback lookup
+            basename = os.path.basename(file_path)
+            recipe_id_from_file = basename[:-len('.recipe.json')] if basename.lower().endswith('.recipe.json') else None
+
+            if cached_stats:
+                cached_mtime, cached_size = cached_stats
+                # Check if file is unchanged
+                if abs(current_mtime - cached_mtime) < 1.0 and current_size == cached_size:
+                    # Try direct path lookup first
+                    cached_recipe = persisted_by_path.get(file_path)
+                    # Fallback to recipe_id lookup if path lookup fails
+                    if not cached_recipe and recipe_id_from_file:
+                        cached_recipe = recipe_by_id.get(recipe_id_from_file)
+                    if cached_recipe:
+                        recipe_id = str(cached_recipe.get('id', ''))
+                        # Track folder from file path
+                        cached_recipe['folder'] = cached_recipe.get('folder') or self._calculate_folder(file_path)
+                        recipes.append(cached_recipe)
+                        json_paths[recipe_id] = file_path
+                        continue
+
+            # File is new or changed - need to re-read
+            changed = True
+            recipe_data = self._load_recipe_file_sync(file_path)
+            if recipe_data:
+                recipe_id = str(recipe_data.get('id', ''))
+                recipes.append(recipe_data)
+                json_paths[recipe_id] = file_path
+
+        # Check for deleted files
+        for json_path in persisted.file_stats.keys():
+            if json_path not in current_files:
+                changed = True
+                logger.debug("Recipe file deleted: %s", json_path)
+
+        return recipes, changed, json_paths
+
+    def _full_directory_scan_sync(self, recipes_dir: str) -> Tuple[List[Dict], Dict[str, str]]:
+        """Perform a full synchronous directory scan for recipes.
+
+        Args:
+            recipes_dir: Path to the recipes directory.
+
+        Returns:
+            Tuple of (recipes list, json_paths dict).
+        """
+        recipes: List[Dict] = []
+        json_paths: Dict[str, str] = {}
+
+        # Get all recipe JSON files
+        recipe_files = []
+        for root, _, files in os.walk(recipes_dir):
+            for file in files:
+                if file.lower().endswith('.recipe.json'):
+                    recipe_files.append(os.path.join(root, file))
+
+        # Process each recipe file
+        for recipe_path in recipe_files:
+            recipe_data = self._load_recipe_file_sync(recipe_path)
+            if recipe_data:
+                recipe_id = str(recipe_data.get('id', ''))
+                recipes.append(recipe_data)
+                json_paths[recipe_id] = recipe_path
+
+        return recipes, json_paths
+
+    def _load_recipe_file_sync(self, recipe_path: str) -> Optional[Dict]:
+        """Load a single recipe file synchronously.
+
+        Args:
+            recipe_path: Path to the recipe JSON file.
+
+        Returns:
+            Recipe dictionary if valid, None otherwise.
+        """
+        try:
+            with open(recipe_path, 'r', encoding='utf-8') as f:
+                recipe_data = json.load(f)
+
+            # Validate recipe data
+            if not recipe_data or not isinstance(recipe_data, dict):
+                logger.warning(f"Invalid recipe data in {recipe_path}")
+                return None
+
+            # Ensure required fields exist
+            required_fields = ['id', 'file_path', 'title']
+            if not all(field in recipe_data for field in required_fields):
+                logger.warning(f"Missing required fields in {recipe_path}")
+                return None
+
+            # Ensure the image file exists and prioritize local siblings
+            image_path = recipe_data.get('file_path')
+            path_updated = False
+            if image_path:
+                recipe_dir = os.path.dirname(recipe_path)
+                image_filename = os.path.basename(image_path)
+                local_sibling_path = os.path.normpath(os.path.join(recipe_dir, image_filename))
+
+                # If local sibling exists and stored path is different, prefer local
+                if os.path.exists(local_sibling_path) and os.path.normpath(image_path) != local_sibling_path:
+                    recipe_data['file_path'] = local_sibling_path
+                    path_updated = True
+                    logger.info(f"Updated recipe image path to local sibling: {local_sibling_path}")
+                elif not os.path.exists(image_path):
+                    logger.warning(f"Recipe image not found and no local sibling: {image_path}")
+
+            if path_updated:
+                try:
+                    with open(recipe_path, 'w', encoding='utf-8') as f:
+                        json.dump(recipe_data, f, indent=4, ensure_ascii=False)
+                except Exception as e:
+                    logger.warning(f"Failed to persist repair for {recipe_path}: {e}")
+
+            # Track folder placement relative to recipes directory
+            recipe_data['folder'] = recipe_data.get('folder') or self._calculate_folder(recipe_path)
+
+            # Ensure loras array exists
+            if 'loras' not in recipe_data:
+                recipe_data['loras'] = []
+
+            # Ensure gen_params exists
+            if 'gen_params' not in recipe_data:
+                recipe_data['gen_params'] = {}
+
+            return recipe_data
+        except Exception as e:
+            logger.error(f"Error loading recipe file {recipe_path}: {e}")
+            import traceback
+            traceback.print_exc(file=sys.stderr)
+            return None
+
+    def _sort_cache_sync(self) -> None:
+        """Sort cache data synchronously."""
+        try:
+            # Sort by name
+            self._cache.sorted_by_name = natsorted(
+                self._cache.raw_data,
+                key=lambda x: x.get('title', '').lower()
+            )
+
+            # Sort by date (modified or created)
+            self._cache.sorted_by_date = sorted(
+                self._cache.raw_data,
+                key=lambda x: (x.get('modified', x.get('created_date', 0)), x.get('file_path', '')),
+                reverse=True
+            )
+        except Exception as e:
+            logger.error(f"Error sorting recipe cache: {e}")
+
     async def _wait_for_lora_scanner(self) -> None:
         """Ensure the LoRA scanner has initialized before recipe enrichment."""
 
@@ -570,7 +748,10 @@ class RecipeScanner:
         self._post_scan_task = loop.create_task(_run_enrichment(), name="recipe_cache_enrichment")
 
     def _schedule_fts_index_build(self) -> None:
-        """Build FTS index in background without blocking."""
+        """Build FTS index in background without blocking.
+
+        Validates existing index first and reuses it if valid.
+        """
 
         if self._fts_index_task and not self._fts_index_task.done():
             return  # Already running
@@ -587,7 +768,25 @@ class RecipeScanner:
             try:
                 self._fts_index = RecipeFTSIndex()
 
-                # Run in thread pool (SQLite is blocking)
+                # Check if existing index is valid
+                recipe_ids = {str(r.get('id', '')) for r in self._cache.raw_data if r.get('id')}
+                recipe_count = len(self._cache.raw_data)
+
+                # Run validation in thread pool
+                is_valid = await loop.run_in_executor(
+                    None,
+                    self._fts_index.validate_index,
+                    recipe_count,
+                    recipe_ids
+                )
+
+                if is_valid:
+                    logger.info("FTS index validated, reusing existing index with %d recipes", recipe_count)
+                    self._fts_index._ready.set()
+                    return
+
+                # Only rebuild if validation fails
+                logger.info("FTS index invalid or outdated, rebuilding...")
                 await loop.run_in_executor(
                     None,
                     self._fts_index.build_index,
@@ -632,7 +831,12 @@ class RecipeScanner:
             fields = None
 
         try:
-            return self._fts_index.search(search, fields)
+            result = self._fts_index.search(search, fields)
+            # Return None if empty to trigger fuzzy fallback
+            # Empty FTS results may indicate query syntax issues or need for fuzzy matching
+            if not result:
+                return None
+            return result
         except Exception as exc:
             logger.debug("FTS search failed, falling back to fuzzy search: %s", exc)
             return None
@@ -870,6 +1074,12 @@ class RecipeScanner:
         # Update FTS index
         self._update_fts_index_for_recipe(recipe_data, 'add')
 
+        # Persist to SQLite cache
+        if self._persistent_cache:
+            recipe_id = str(recipe_data.get('id', ''))
+            json_path = self._json_path_map.get(recipe_id, '')
+            self._persistent_cache.update_recipe(recipe_data, json_path)
+
     async def remove_recipe(self, recipe_id: str) -> bool:
         """Remove a recipe from the cache by ID."""
 
@@ -886,6 +1096,12 @@ class RecipeScanner:
 
         # Update FTS index
         self._update_fts_index_for_recipe(recipe_id, 'remove')
+
+        # Remove from SQLite cache
+        if self._persistent_cache:
+            self._persistent_cache.remove_recipe(recipe_id)
+            self._json_path_map.pop(recipe_id, None)
+
         return True
 
     async def bulk_remove(self, recipe_ids: Iterable[str]) -> int:
@@ -895,9 +1111,13 @@ class RecipeScanner:
         removed = await cache.bulk_remove(recipe_ids, resort=False)
         if removed:
             self._schedule_resort()
-            # Update FTS index for each removed recipe
-            for recipe_id in (str(r.get('id', '')) for r in removed):
+            # Update FTS index and persistent cache for each removed recipe
+            for recipe in removed:
+                recipe_id = str(recipe.get('id', ''))
                 self._update_fts_index_for_recipe(recipe_id, 'remove')
+                if self._persistent_cache:
+                    self._persistent_cache.remove_recipe(recipe_id)
+                    self._json_path_map.pop(recipe_id, None)
         return len(removed)
 
     async def scan_all_recipes(self) -> List[Dict]:
@@ -1690,11 +1910,11 @@ class RecipeScanner:
 
     async def update_recipe_metadata(self, recipe_id: str, metadata: dict) -> bool:
         """Update recipe metadata (like title and tags) in both file system and cache
-        
+
         Args:
             recipe_id: The ID of the recipe to update
             metadata: Dictionary containing metadata fields to update (title, tags, etc.)
-            
+
         Returns:
             bool: True if successful, False otherwise
         """
@@ -1702,16 +1922,16 @@ class RecipeScanner:
         recipe_json_path = await self.get_recipe_json_path(recipe_id)
         if not recipe_json_path or not os.path.exists(recipe_json_path):
             return False
-            
+
         try:
             # Load existing recipe data
             with open(recipe_json_path, 'r', encoding='utf-8') as f:
                 recipe_data = json.load(f)
-                
+
             # Update fields
             for key, value in metadata.items():
                 recipe_data[key] = value
-                
+
             # Save updated recipe
             with open(recipe_json_path, 'w', encoding='utf-8') as f:
                 json.dump(recipe_data, f, indent=4, ensure_ascii=False)
@@ -1724,6 +1944,11 @@ class RecipeScanner:
             # Update FTS index
             self._update_fts_index_for_recipe(recipe_data, 'update')
 
+            # Update persistent SQLite cache
+            if self._persistent_cache:
+                self._persistent_cache.update_recipe(recipe_data, recipe_json_path)
+                self._json_path_map[recipe_id] = recipe_json_path
+
             # If the recipe has an image, update its EXIF metadata
             from ..utils.exif_utils import ExifUtils
             image_path = recipe_data.get('file_path')
@@ -1795,6 +2020,11 @@ class RecipeScanner:
         # Update FTS index
         self._update_fts_index_for_recipe(recipe_data, 'update')
 
+        # Update persistent SQLite cache
+        if self._persistent_cache:
+            self._persistent_cache.update_recipe(recipe_data, recipe_json_path)
+            self._json_path_map[recipe_id] = recipe_json_path
+
         updated_lora = dict(lora_entry)
         if target_lora is not None:
             preview_url = target_lora.get('preview_url')
@@ -1918,26 +2148,31 @@ class RecipeScanner:
         if not recipes_to_update:
             return 0, 0
             
-        # Persist changes to disk
+        # Persist changes to disk and SQLite cache
         async with self._mutation_lock:
             for recipe in recipes_to_update:
-                recipe_id = recipe.get('id')
+                recipe_id = str(recipe.get('id', ''))
                 if not recipe_id:
                     continue
-                    
+
                 recipe_path = os.path.join(self.recipes_dir, f"{recipe_id}.recipe.json")
                 try:
                     self._write_recipe_file(recipe_path, recipe)
                     file_updated_count += 1
                     logger.info(f"Updated file_name in recipe {recipe_path}: -> {new_file_name}")
+
+                    # Update persistent SQLite cache
+                    if self._persistent_cache:
+                        self._persistent_cache.update_recipe(recipe, recipe_path)
+                        self._json_path_map[recipe_id] = recipe_path
                 except Exception as e:
                     logger.error(f"Error updating recipe file {recipe_path}: {e}")
-        
+
         # We don't necessarily need to resort because LoRA file_name isn't a sort key,
         # but we might want to schedule a resort if we're paranoid or if searching relies on sorted state.
         # Given it's a rename of a dependency, search results might change if searching by LoRA name.
         self._schedule_resort()
-            
+
         return file_updated_count, cache_updated_count
 
     async def find_recipes_by_fingerprint(self, fingerprint: str) -> list:
@@ -1996,3 +2231,26 @@ class RecipeScanner:
         duplicate_groups = {k: v for k, v in fingerprint_groups.items() if len(v) > 1}
         
         return duplicate_groups
+    
+    async def find_duplicate_recipes_by_source(self) -> dict:
+        """Find all recipe duplicates based on source_path (Civitai image URLs)
+        
+        Returns:
+            Dictionary where keys are source URLs and values are lists of recipe IDs
+        """
+        cache = await self.get_cached_data()
+        
+        url_groups = {}
+        for recipe in cache.raw_data:
+            source_url = recipe.get('source_path', '').strip()
+            if not source_url:
+                continue
+            
+            if source_url not in url_groups:
+                url_groups[source_url] = []
+                
+            url_groups[source_url].append(recipe.get('id'))
+        
+        duplicate_groups = {k: v for k, v in url_groups.items() if len(v) > 1}
+        
+        return duplicate_groups

py/services/settings_manager.py

@@ -35,6 +35,8 @@ DEFAULT_SETTINGS: Dict[str, Any] = {
     "hash_chunk_size_mb": DEFAULT_HASH_CHUNK_SIZE_MB,
     "language": "en",
     "show_only_sfw": False,
+    "onboarding_completed": False,
+    "dismissed_banners": [],
     "enable_metadata_archive_db": False,
     "proxy_enabled": False,
     "proxy_host": "",
@@ -61,7 +63,7 @@ DEFAULT_SETTINGS: Dict[str, Any] = {
     "compact_mode": False,
     "priority_tags": DEFAULT_PRIORITY_TAG_CONFIG.copy(),
     "model_name_display": "model_name",
-    "model_card_footer_action": "example_images",
+    "model_card_footer_action": "replace_preview",
     "update_flag_strategy": "same_base",
     "auto_organize_exclusions": [],
 }

py/services/tag_fts_index.py

@@ -0,0 +1,680 @@
+"""SQLite FTS5-based full-text search index for tags.
+
+This module provides fast tag search using SQLite's FTS5 extension,
+enabling sub-100ms search times for 221k+ Danbooru/e621 tags.
+
+Supports alias search: when a user searches for an alias (e.g., "miku"),
+the system returns the canonical tag (e.g., "hatsune_miku") and indicates
+which alias was matched.
+"""
+
+from __future__ import annotations
+
+import csv
+import logging
+import os
+import re
+import sqlite3
+import threading
+import time
+from pathlib import Path
+from typing import Dict, List, Optional, Set
+
+from ..utils.cache_paths import CacheType, resolve_cache_path_with_migration
+
+logger = logging.getLogger(__name__)
+
+# Schema version for tracking migrations
+SCHEMA_VERSION = 2  # Version 2: Added aliases support
+
+
+# Category definitions for Danbooru and e621
+CATEGORY_NAMES = {
+    # Danbooru categories
+    0: "general",
+    1: "artist",
+    3: "copyright",
+    4: "character",
+    5: "meta",
+    # e621 categories
+    7: "general",
+    8: "artist",
+    10: "copyright",
+    11: "character",
+    12: "species",
+    14: "meta",
+    15: "lore",
+}
+
+# Map category names to their IDs (for filtering)
+CATEGORY_NAME_TO_IDS = {
+    "general": [0, 7],
+    "artist": [1, 8],
+    "copyright": [3, 10],
+    "character": [4, 11],
+    "meta": [5, 14],
+    "species": [12],
+    "lore": [15],
+}
+
+
+class TagFTSIndex:
+    """SQLite FTS5-based full-text search index for tags.
+
+    Provides fast prefix-based search across the Danbooru/e621 tag database.
+    Supports category-based filtering and returns enriched results with
+    post counts and category information.
+    """
+
+    _DEFAULT_FILENAME = "tag_fts.sqlite"
+    _CSV_FILENAME = "danbooru_e621_merged.csv"
+
+    def __init__(self, db_path: Optional[str] = None, csv_path: Optional[str] = None) -> None:
+        """Initialize the FTS index.
+
+        Args:
+            db_path: Optional path to the SQLite database file.
+                     If not provided, uses the default location in settings directory.
+            csv_path: Optional path to the CSV file containing tag data.
+                      If not provided, looks in the refs/ directory.
+        """
+        self._db_path = db_path or self._resolve_default_db_path()
+        self._csv_path = csv_path or self._resolve_default_csv_path()
+        self._lock = threading.Lock()
+        self._ready = threading.Event()
+        self._indexing_in_progress = False
+        self._schema_initialized = False
+        self._warned_not_ready = False
+
+        # Ensure directory exists
+        try:
+            directory = os.path.dirname(self._db_path)
+            if directory:
+                os.makedirs(directory, exist_ok=True)
+        except Exception as exc:
+            logger.warning("Could not create FTS index directory %s: %s", directory, exc)
+
+    def _resolve_default_db_path(self) -> str:
+        """Resolve the default database path."""
+        env_override = os.environ.get("LORA_MANAGER_TAG_FTS_DB")
+        return resolve_cache_path_with_migration(
+            CacheType.TAG_FTS,
+            env_override=env_override,
+        )
+
+    def _resolve_default_csv_path(self) -> str:
+        """Resolve the default CSV file path."""
+        # Look for the CSV in the refs/ directory relative to the package
+        package_dir = Path(__file__).parent.parent.parent
+        csv_path = package_dir / "refs" / self._CSV_FILENAME
+        return str(csv_path)
+
+    def get_database_path(self) -> str:
+        """Return the resolved database path."""
+        return self._db_path
+
+    def get_csv_path(self) -> str:
+        """Return the resolved CSV path."""
+        return self._csv_path
+
+    def is_ready(self) -> bool:
+        """Check if the FTS index is ready for queries."""
+        return self._ready.is_set()
+
+    def is_indexing(self) -> bool:
+        """Check if indexing is currently in progress."""
+        return self._indexing_in_progress
+
+    def initialize(self) -> None:
+        """Initialize the database schema."""
+        if self._schema_initialized:
+            return
+
+        with self._lock:
+            if self._schema_initialized:
+                return
+
+            try:
+                conn = self._connect()
+                try:
+                    conn.execute("PRAGMA journal_mode=WAL")
+
+                    # Check if we need to migrate from old schema
+                    needs_rebuild = self._check_and_migrate_schema(conn)
+
+                    conn.executescript("""
+                        -- FTS5 virtual table for full-text search
+                        -- searchable_text contains "tag_name alias1 alias2 ..." for alias matching
+                        CREATE VIRTUAL TABLE IF NOT EXISTS tag_fts USING fts5(
+                            searchable_text,
+                            tokenize='unicode61 remove_diacritics 2'
+                        );
+
+                        -- Tags table with metadata and aliases
+                        CREATE TABLE IF NOT EXISTS tags (
+                            rowid INTEGER PRIMARY KEY,
+                            tag_name TEXT UNIQUE NOT NULL,
+                            category INTEGER NOT NULL DEFAULT 0,
+                            post_count INTEGER NOT NULL DEFAULT 0,
+                            aliases TEXT DEFAULT ''
+                        );
+
+                        -- Indexes for efficient filtering
+                        CREATE INDEX IF NOT EXISTS idx_tags_category ON tags(category);
+                        CREATE INDEX IF NOT EXISTS idx_tags_post_count ON tags(post_count DESC);
+
+                        -- Index version tracking
+                        CREATE TABLE IF NOT EXISTS fts_metadata (
+                            key TEXT PRIMARY KEY,
+                            value TEXT
+                        );
+                    """)
+
+                    # Set schema version
+                    conn.execute(
+                        "INSERT OR REPLACE INTO fts_metadata (key, value) VALUES (?, ?)",
+                        ("schema_version", str(SCHEMA_VERSION))
+                    )
+                    conn.commit()
+
+                    self._schema_initialized = True
+                    self._needs_rebuild = needs_rebuild
+                    logger.debug("Tag FTS index schema initialized at %s", self._db_path)
+                finally:
+                    conn.close()
+            except Exception as exc:
+                logger.error("Failed to initialize tag FTS schema: %s", exc)
+
+    def _check_and_migrate_schema(self, conn: sqlite3.Connection) -> bool:
+        """Check schema version and migrate if necessary.
+
+        Returns:
+            True if the index needs to be rebuilt, False otherwise.
+        """
+        try:
+            # Check if fts_metadata table exists
+            cursor = conn.execute(
+                "SELECT name FROM sqlite_master WHERE type='table' AND name='fts_metadata'"
+            )
+            if not cursor.fetchone():
+                return False  # Fresh database, no migration needed
+
+            # Check schema version
+            cursor = conn.execute(
+                "SELECT value FROM fts_metadata WHERE key='schema_version'"
+            )
+            row = cursor.fetchone()
+            if not row:
+                # Old schema without version, needs rebuild
+                logger.info("Migrating tag FTS index to schema version %d (adding alias support)", SCHEMA_VERSION)
+                self._drop_old_tables(conn)
+                return True
+
+            current_version = int(row[0])
+            if current_version < SCHEMA_VERSION:
+                logger.info("Migrating tag FTS index from version %d to %d", current_version, SCHEMA_VERSION)
+                self._drop_old_tables(conn)
+                return True
+
+            return False
+        except Exception as exc:
+            logger.warning("Error checking schema version: %s", exc)
+            return False
+
+    def _drop_old_tables(self, conn: sqlite3.Connection) -> None:
+        """Drop old tables for schema migration."""
+        try:
+            conn.executescript("""
+                DROP TABLE IF EXISTS tag_fts;
+                DROP TABLE IF EXISTS tags;
+            """)
+            conn.commit()
+        except Exception as exc:
+            logger.warning("Error dropping old tables: %s", exc)
+
+    def build_index(self) -> None:
+        """Build the FTS index from the CSV file.
+
+        This method parses the danbooru_e621_merged.csv file and creates
+        the FTS index for fast searching. The CSV format is:
+        tag_name,category,post_count,aliases
+
+        Where aliases is a comma-separated string (e.g., "miku,vocaloid_miku,39").
+        """
+        if self._indexing_in_progress:
+            logger.warning("Tag FTS indexing already in progress, skipping")
+            return
+
+        if not os.path.exists(self._csv_path):
+            logger.warning("CSV file not found at %s, cannot build tag index", self._csv_path)
+            return
+
+        self._indexing_in_progress = True
+        self._ready.clear()
+        start_time = time.time()
+
+        try:
+            self.initialize()
+            if not self._schema_initialized:
+                logger.error("Cannot build tag FTS index: schema not initialized")
+                return
+
+            with self._lock:
+                conn = self._connect()
+                try:
+                    conn.execute("BEGIN")
+
+                    # Clear existing data
+                    conn.execute("DELETE FROM tag_fts")
+                    conn.execute("DELETE FROM tags")
+
+                    # Parse CSV and insert in batches
+                    batch_size = 500
+                    rows = []
+                    total_inserted = 0
+                    tags_with_aliases = 0
+
+                    with open(self._csv_path, "r", encoding="utf-8") as f:
+                        reader = csv.reader(f)
+                        for row in reader:
+                            if len(row) < 3:
+                                continue
+
+                            tag_name = row[0].strip()
+                            if not tag_name:
+                                continue
+
+                            try:
+                                category = int(row[1])
+                            except (ValueError, IndexError):
+                                category = 0
+
+                            try:
+                                post_count = int(row[2])
+                            except (ValueError, IndexError):
+                                post_count = 0
+
+                            # Parse aliases from column 4 (if present)
+                            aliases = row[3].strip() if len(row) >= 4 else ""
+                            if aliases:
+                                tags_with_aliases += 1
+
+                            rows.append((tag_name, category, post_count, aliases))
+
+                            if len(rows) >= batch_size:
+                                self._insert_batch(conn, rows)
+                                total_inserted += len(rows)
+                                rows = []
+
+                    # Insert remaining rows
+                    if rows:
+                        self._insert_batch(conn, rows)
+                        total_inserted += len(rows)
+
+                    # Update metadata
+                    conn.execute(
+                        "INSERT OR REPLACE INTO fts_metadata (key, value) VALUES (?, ?)",
+                        ("last_build_time", str(time.time()))
+                    )
+                    conn.execute(
+                        "INSERT OR REPLACE INTO fts_metadata (key, value) VALUES (?, ?)",
+                        ("tag_count", str(total_inserted))
+                    )
+                    conn.execute(
+                        "INSERT OR REPLACE INTO fts_metadata (key, value) VALUES (?, ?)",
+                        ("schema_version", str(SCHEMA_VERSION))
+                    )
+
+                    conn.commit()
+                    elapsed = time.time() - start_time
+                    logger.info(
+                        "Tag FTS index built: %d tags indexed (%d with aliases) in %.2fs",
+                        total_inserted, tags_with_aliases, elapsed
+                    )
+                finally:
+                    conn.close()
+
+            self._ready.set()
+
+        except Exception as exc:
+            logger.error("Failed to build tag FTS index: %s", exc, exc_info=True)
+        finally:
+            self._indexing_in_progress = False
+
+    def _insert_batch(self, conn: sqlite3.Connection, rows: List[tuple]) -> None:
+        """Insert a batch of rows into the database.
+
+        Each row is a tuple of (tag_name, category, post_count, aliases).
+        The FTS searchable_text is built as "tag_name alias1 alias2 ..." for alias matching.
+        """
+        # Insert into tags table (with aliases)
+        conn.executemany(
+            "INSERT OR IGNORE INTO tags (tag_name, category, post_count, aliases) VALUES (?, ?, ?, ?)",
+            rows
+        )
+
+        # Build a map of tag_name -> aliases for FTS insertion
+        aliases_map = {row[0]: row[3] for row in rows}
+
+        # Get rowids and insert into FTS table with explicit rowid
+        # to ensure tags.rowid matches tag_fts.rowid for JOINs
+        tag_names = [row[0] for row in rows]
+        placeholders = ",".join("?" * len(tag_names))
+        cursor = conn.execute(
+            f"SELECT rowid, tag_name FROM tags WHERE tag_name IN ({placeholders})",
+            tag_names
+        )
+
+        # Build FTS rows with (rowid, searchable_text) = (tags.rowid, "tag_name alias1 alias2 ...")
+        fts_rows = []
+        for rowid, tag_name in cursor.fetchall():
+            aliases = aliases_map.get(tag_name, "")
+            if aliases:
+                # Replace commas with spaces to create searchable text
+                # Strip "/" prefix from aliases as it's an FTS5 special character
+                alias_parts = []
+                for alias in aliases.split(","):
+                    alias = alias.strip()
+                    if alias.startswith("/"):
+                        alias = alias[1:]  # Remove leading slash
+                    if alias:
+                        alias_parts.append(alias)
+                searchable_text = f"{tag_name} {' '.join(alias_parts)}" if alias_parts else tag_name
+            else:
+                searchable_text = tag_name
+            fts_rows.append((rowid, searchable_text))
+
+        if fts_rows:
+            conn.executemany("INSERT INTO tag_fts (rowid, searchable_text) VALUES (?, ?)", fts_rows)
+
+    def ensure_ready(self) -> bool:
+        """Ensure the index is ready, building if necessary.
+
+        Returns:
+            True if the index is ready, False otherwise.
+        """
+        if self.is_ready():
+            return True
+
+        # Check if index already exists and has data
+        self.initialize()
+        if self._schema_initialized:
+            # Check if schema migration requires rebuild
+            if getattr(self, "_needs_rebuild", False):
+                logger.info("Schema migration requires index rebuild")
+                self._needs_rebuild = False
+                self.build_index()
+                return self.is_ready()
+
+            count = self.get_indexed_count()
+            if count > 0:
+                self._ready.set()
+                logger.debug("Tag FTS index already populated with %d tags", count)
+                return True
+
+        # Build the index
+        self.build_index()
+        return self.is_ready()
+
+    def search(
+        self,
+        query: str,
+        categories: Optional[List[int]] = None,
+        limit: int = 20
+    ) -> List[Dict]:
+        """Search tags using FTS5 with prefix matching.
+
+        Supports alias search: if the query matches an alias rather than
+        the tag_name, the result will include a "matched_alias" field.
+
+        Args:
+            query: The search query string.
+            categories: Optional list of category IDs to filter by.
+            limit: Maximum number of results to return.
+
+        Returns:
+            List of dictionaries with tag_name, category, post_count,
+            and optionally matched_alias.
+        """
+        # Ensure index is ready (lazy initialization)
+        if not self.ensure_ready():
+            if not self._warned_not_ready:
+                logger.debug("Tag FTS index not ready, returning empty results")
+                self._warned_not_ready = True
+            return []
+
+        if not query or not query.strip():
+            return []
+
+        fts_query = self._build_fts_query(query)
+        if not fts_query:
+            return []
+
+        try:
+            with self._lock:
+                conn = self._connect(readonly=True)
+                try:
+                    # Build the SQL query - now also fetch aliases for matched_alias detection
+                    # Use subquery for category filter to ensure FTS is evaluated first
+                    if categories:
+                        placeholders = ",".join("?" * len(categories))
+                        sql = f"""
+                            SELECT t.tag_name, t.category, t.post_count, t.aliases
+                            FROM tags t
+                            WHERE t.rowid IN (
+                                SELECT rowid FROM tag_fts WHERE searchable_text MATCH ?
+                            )
+                            AND t.category IN ({placeholders})
+                            ORDER BY t.post_count DESC
+                            LIMIT ?
+                        """
+                        params = [fts_query] + categories + [limit]
+                    else:
+                        sql = """
+                            SELECT t.tag_name, t.category, t.post_count, t.aliases
+                            FROM tag_fts f
+                            JOIN tags t ON f.rowid = t.rowid
+                            WHERE f.searchable_text MATCH ?
+                            ORDER BY t.post_count DESC
+                            LIMIT ?
+                        """
+                        params = [fts_query, limit]
+
+                    cursor = conn.execute(sql, params)
+                    results = []
+                    for row in cursor.fetchall():
+                        result = {
+                            "tag_name": row[0],
+                            "category": row[1],
+                            "post_count": row[2],
+                        }
+
+                        # Check if search matched an alias rather than the tag_name
+                        matched_alias = self._find_matched_alias(query, row[0], row[3])
+                        if matched_alias:
+                            result["matched_alias"] = matched_alias
+
+                        results.append(result)
+                    return results
+                finally:
+                    conn.close()
+        except Exception as exc:
+            logger.debug("Tag FTS search error for query '%s': %s", query, exc)
+            return []
+
+    def _find_matched_alias(self, query: str, tag_name: str, aliases_str: str) -> Optional[str]:
+        """Find which alias matched the query, if any.
+
+        Args:
+            query: The original search query.
+            tag_name: The canonical tag name.
+            aliases_str: Comma-separated string of aliases.
+
+        Returns:
+            The matched alias string, or None if the query matched the tag_name directly.
+        """
+        query_lower = query.lower().strip()
+        if not query_lower:
+            return None
+
+        # Strip leading "/" from query if present (FTS index strips these)
+        query_normalized = query_lower.lstrip("/")
+
+        # Check if query matches tag_name prefix (direct match, no alias needed)
+        if tag_name.lower().startswith(query_normalized):
+            return None
+
+        # Check aliases first - if query matches an alias or a word within an alias, return it
+        if aliases_str:
+            for alias in aliases_str.split(","):
+                alias = alias.strip()
+                if not alias:
+                    continue
+                # Normalize alias for comparison (strip leading slash)
+                alias_normalized = alias.lower().lstrip("/")
+
+                # Check if alias starts with query
+                if alias_normalized.startswith(query_normalized):
+                    return alias  # Return original alias (with "/" if present)
+
+                # Check if any word within the alias starts with query
+                # (mirrors FTS5 tokenization which splits on underscores)
+                alias_words = alias_normalized.replace("_", " ").split()
+                for word in alias_words:
+                    if word.startswith(query_normalized):
+                        return alias
+
+        # If no alias matched, check if query matches a word in tag_name
+        # (handles cases like "long_hair" matching "long" - no alias indicator needed)
+        tag_words = tag_name.lower().replace("_", " ").split()
+        for word in tag_words:
+            if word.startswith(query_normalized):
+                return None
+
+        # Query matched via FTS but not tag_name words or aliases
+        # This shouldn't normally happen, but return None for safety
+        return None
+
+    def get_indexed_count(self) -> int:
+        """Return the number of tags currently indexed."""
+        if not self._schema_initialized:
+            return 0
+
+        try:
+            with self._lock:
+                conn = self._connect(readonly=True)
+                try:
+                    cursor = conn.execute("SELECT COUNT(*) FROM tags")
+                    result = cursor.fetchone()
+                    return result[0] if result else 0
+                finally:
+                    conn.close()
+        except Exception:
+            return 0
+
+    def clear(self) -> bool:
+        """Clear all data from the FTS index.
+
+        Returns:
+            True if successful, False otherwise.
+        """
+        try:
+            with self._lock:
+                conn = self._connect()
+                try:
+                    conn.execute("DELETE FROM tag_fts")
+                    conn.execute("DELETE FROM tags")
+                    conn.commit()
+                    self._ready.clear()
+                    return True
+                finally:
+                    conn.close()
+        except Exception as exc:
+            logger.error("Failed to clear tag FTS index: %s", exc)
+            return False
+
+    # Internal helpers
+
+    def _connect(self, readonly: bool = False) -> sqlite3.Connection:
+        """Create a database connection."""
+        uri = False
+        path = self._db_path
+        if readonly:
+            if not os.path.exists(path):
+                raise FileNotFoundError(path)
+            path = f"file:{path}?mode=ro"
+            uri = True
+        conn = sqlite3.connect(path, check_same_thread=False, uri=uri)
+        conn.row_factory = sqlite3.Row
+        return conn
+
+    def _build_fts_query(self, query: str) -> str:
+        """Build an FTS5 query string with prefix matching.
+
+        Args:
+            query: The user's search query.
+
+        Returns:
+            FTS5 query string.
+        """
+        # Split query into words and clean them
+        words = query.lower().split()
+        if not words:
+            return ""
+
+        # Escape and add prefix wildcard to each word
+        prefix_terms = []
+        for word in words:
+            escaped = self._escape_fts_query(word)
+            if escaped:
+                # Add prefix wildcard for substring-like matching
+                prefix_terms.append(f"{escaped}*")
+
+        if not prefix_terms:
+            return ""
+
+        # Combine terms with implicit AND (all words must match)
+        return " ".join(prefix_terms)
+
+    def _escape_fts_query(self, text: str) -> str:
+        """Escape special FTS5 characters.
+
+        FTS5 special characters: " ( ) * : ^ - /
+        We keep * for prefix matching but escape others.
+        """
+        if not text:
+            return ""
+
+        # Replace FTS5 special characters with space
+        # Note: "/" is special in FTS5 (column filter syntax), so we strip it
+        special = ['"', "(", ")", "*", ":", "^", "-", "{", "}", "[", "]", "/"]
+        result = text
+        for char in special:
+            result = result.replace(char, " ")
+
+        # Collapse multiple spaces and strip
+        result = re.sub(r"\s+", " ", result).strip()
+        return result
+
+
+# Singleton instance
+_tag_fts_index: Optional[TagFTSIndex] = None
+_tag_fts_lock = threading.Lock()
+
+
+def get_tag_fts_index() -> TagFTSIndex:
+    """Get the singleton TagFTSIndex instance."""
+    global _tag_fts_index
+    if _tag_fts_index is None:
+        with _tag_fts_lock:
+            if _tag_fts_index is None:
+                _tag_fts_index = TagFTSIndex()
+    return _tag_fts_index
+
+
+__all__ = [
+    "TagFTSIndex",
+    "get_tag_fts_index",
+    "CATEGORY_NAMES",
+    "CATEGORY_NAME_TO_IDS",
+]

py/services/use_cases/bulk_metadata_refresh_use_case.py

@@ -48,9 +48,14 @@ class BulkMetadataRefreshUseCase:
             for model in cache.raw_data
             if model.get("sha256")
             and (not model.get("civitai") or not model["civitai"].get("id"))
-            and (
-                (enable_metadata_archive_db and not model.get("db_checked", False))
-                or (not enable_metadata_archive_db and model.get("from_civitai") is True)
+            and not (
+                # Skip models confirmed not on CivitAI when no need to retry
+                model.get("from_civitai") is False
+                and model.get("civitai_deleted") is True
+                and (
+                    not enable_metadata_archive_db
+                    or model.get("db_checked", False)
+                )
             )
         ]
 

py/services/websocket_manager.py

@@ -255,6 +255,42 @@ class WebSocketManager:
             self._download_progress.pop(download_id, None)
             logger.debug(f"Cleaned up old download progress for {download_id}")
             
+    async def broadcast_cache_health_warning(self, report: 'HealthReport', page_type: str = None):
+        """
+        Broadcast cache health warning to frontend.
+
+        Args:
+            report: HealthReport instance from CacheHealthMonitor
+            page_type: The page type (loras, checkpoints, embeddings)
+        """
+        from .cache_health_monitor import CacheHealthStatus
+
+        # Only broadcast if there are issues
+        if report.status == CacheHealthStatus.HEALTHY:
+            return
+
+        payload = {
+            'type': 'cache_health_warning',
+            'status': report.status.value,
+            'message': report.message,
+            'pageType': page_type,
+            'details': {
+                'total': report.total_entries,
+                'valid': report.valid_entries,
+                'invalid': report.invalid_entries,
+                'repaired': report.repaired_entries,
+                'corruption_rate': f"{report.corruption_rate:.1%}",
+                'invalid_paths': report.invalid_paths[:5],  # Limit to first 5
+            }
+        }
+
+        logger.info(
+            f"Broadcasting cache health warning: {report.status.value} "
+            f"({report.invalid_entries} invalid entries)"
+        )
+
+        await self.broadcast(payload)
+
     def get_connected_clients_count(self) -> int:
         """Get number of connected clients"""
         return len(self._websockets)

py/utils/cache_paths.py

@@ -0,0 +1,421 @@
+"""Centralized cache path resolution with automatic migration support.
+
+This module provides a unified interface for resolving cache file paths,
+with automatic migration from legacy locations to the new organized
+cache directory structure.
+
+Target structure:
+    {settings_dir}/
+    └── cache/
+        ├── symlink/
+        │   └── symlink_map.json
+        ├── model/
+        │   └── {library_name}.sqlite
+        ├── recipe/
+        │   └── {library_name}.sqlite
+        └── fts/
+            ├── recipe_fts.sqlite
+            └── tag_fts.sqlite
+"""
+
+from __future__ import annotations
+
+import logging
+import os
+import re
+import shutil
+from enum import Enum
+from typing import List, Optional
+
+from .settings_paths import get_project_root, get_settings_dir
+
+logger = logging.getLogger(__name__)
+
+
+class CacheType(Enum):
+    """Types of cache files managed by the cache path resolver."""
+
+    MODEL = "model"
+    RECIPE = "recipe"
+    RECIPE_FTS = "recipe_fts"
+    TAG_FTS = "tag_fts"
+    SYMLINK = "symlink"
+
+
+# Subdirectory structure for each cache type
+_CACHE_SUBDIRS = {
+    CacheType.MODEL: "model",
+    CacheType.RECIPE: "recipe",
+    CacheType.RECIPE_FTS: "fts",
+    CacheType.TAG_FTS: "fts",
+    CacheType.SYMLINK: "symlink",
+}
+
+# Filename patterns for each cache type
+_CACHE_FILENAMES = {
+    CacheType.MODEL: "{library_name}.sqlite",
+    CacheType.RECIPE: "{library_name}.sqlite",
+    CacheType.RECIPE_FTS: "recipe_fts.sqlite",
+    CacheType.TAG_FTS: "tag_fts.sqlite",
+    CacheType.SYMLINK: "symlink_map.json",
+}
+
+
+def get_cache_base_dir(create: bool = True) -> str:
+    """Return the base cache directory path.
+
+    Args:
+        create: Whether to create the directory if it does not exist.
+
+    Returns:
+        The absolute path to the cache base directory ({settings_dir}/cache/).
+    """
+    settings_dir = get_settings_dir(create=create)
+    cache_dir = os.path.join(settings_dir, "cache")
+    if create:
+        os.makedirs(cache_dir, exist_ok=True)
+    return cache_dir
+
+
+def _sanitize_library_name(library_name: Optional[str]) -> str:
+    """Sanitize a library name for use in filenames.
+
+    Args:
+        library_name: The library name to sanitize.
+
+    Returns:
+        A sanitized version safe for use in filenames.
+    """
+    name = library_name or "default"
+    return re.sub(r"[^A-Za-z0-9_.-]", "_", name)
+
+
+def get_cache_file_path(
+    cache_type: CacheType,
+    library_name: Optional[str] = None,
+    create_dir: bool = True,
+) -> str:
+    """Get the canonical path for a cache file.
+
+    Args:
+        cache_type: The type of cache file.
+        library_name: The library name (only used for MODEL and RECIPE types).
+        create_dir: Whether to create the parent directory if it does not exist.
+
+    Returns:
+        The absolute path to the cache file in its canonical location.
+    """
+    cache_base = get_cache_base_dir(create=create_dir)
+    subdir = _CACHE_SUBDIRS[cache_type]
+    cache_dir = os.path.join(cache_base, subdir)
+
+    if create_dir:
+        os.makedirs(cache_dir, exist_ok=True)
+
+    filename_template = _CACHE_FILENAMES[cache_type]
+    safe_name = _sanitize_library_name(library_name)
+    filename = filename_template.format(library_name=safe_name)
+
+    return os.path.join(cache_dir, filename)
+
+
+def get_legacy_cache_paths(
+    cache_type: CacheType,
+    library_name: Optional[str] = None,
+) -> List[str]:
+    """Get a list of legacy cache file paths to check for migration.
+
+    The paths are returned in order of priority (most recent first).
+
+    Args:
+        cache_type: The type of cache file.
+        library_name: The library name (only used for MODEL and RECIPE types).
+
+    Returns:
+        A list of potential legacy paths to check, in order of preference.
+    """
+    try:
+        settings_dir = get_settings_dir(create=False)
+    except Exception:
+        settings_dir = get_project_root()
+
+    safe_name = _sanitize_library_name(library_name)
+    legacy_paths: List[str] = []
+
+    if cache_type == CacheType.MODEL:
+        # Legacy per-library path: {settings_dir}/model_cache/{library}.sqlite
+        legacy_paths.append(
+            os.path.join(settings_dir, "model_cache", f"{safe_name}.sqlite")
+        )
+        # Legacy root-level single cache (for "default" library only)
+        if safe_name.lower() in ("default", ""):
+            legacy_paths.append(os.path.join(settings_dir, "model_cache.sqlite"))
+
+    elif cache_type == CacheType.RECIPE:
+        # Legacy per-library path: {settings_dir}/recipe_cache/{library}.sqlite
+        legacy_paths.append(
+            os.path.join(settings_dir, "recipe_cache", f"{safe_name}.sqlite")
+        )
+        # Legacy root-level single cache (for "default" library only)
+        if safe_name.lower() in ("default", ""):
+            legacy_paths.append(os.path.join(settings_dir, "recipe_cache.sqlite"))
+
+    elif cache_type == CacheType.RECIPE_FTS:
+        # Legacy root-level path
+        legacy_paths.append(os.path.join(settings_dir, "recipe_fts.sqlite"))
+
+    elif cache_type == CacheType.TAG_FTS:
+        # Legacy root-level path
+        legacy_paths.append(os.path.join(settings_dir, "tag_fts.sqlite"))
+
+    elif cache_type == CacheType.SYMLINK:
+        # Current location in cache/ but without subdirectory
+        legacy_paths.append(
+            os.path.join(settings_dir, "cache", "symlink_map.json")
+        )
+
+    return legacy_paths
+
+
+def _cleanup_legacy_file_after_migration(
+    legacy_path: str,
+    canonical_path: str,
+) -> bool:
+    """Safely remove a legacy file after successful migration.
+
+    Args:
+        legacy_path: The legacy file path to remove.
+        canonical_path: The canonical path where the file was copied to.
+
+    Returns:
+        True if cleanup succeeded, False otherwise.
+    """
+    try:
+        if not os.path.exists(canonical_path):
+            logger.warning(
+                "Skipping cleanup of %s: canonical file not found at %s",
+                legacy_path,
+                canonical_path,
+            )
+            return False
+
+        legacy_size = os.path.getsize(legacy_path)
+        canonical_size = os.path.getsize(canonical_path)
+        if legacy_size != canonical_size:
+            logger.warning(
+                "Skipping cleanup of %s: file size mismatch (legacy=%d, canonical=%d)",
+                legacy_path,
+                legacy_size,
+                canonical_size,
+            )
+            return False
+
+        os.remove(legacy_path)
+        logger.info("Cleaned up legacy cache file: %s", legacy_path)
+
+        _cleanup_empty_legacy_directories(legacy_path)
+
+        return True
+
+    except Exception as exc:
+        logger.warning(
+            "Failed to cleanup legacy cache file %s: %s",
+            legacy_path,
+            exc,
+        )
+        return False
+
+
+def _cleanup_empty_legacy_directories(legacy_path: str) -> None:
+    """Remove empty parent directories of a legacy file.
+
+    This function only removes directories if they are empty,
+    using os.rmdir() which fails on non-empty directories.
+
+    Args:
+        legacy_path: The legacy file path whose parent directories should be cleaned.
+    """
+    try:
+        parent_dir = os.path.dirname(legacy_path)
+
+        legacy_dir_names = ("model_cache", "recipe_cache")
+
+        current = parent_dir
+        while current:
+            base_name = os.path.basename(current)
+
+            if base_name in legacy_dir_names:
+                if os.path.isdir(current) and not os.listdir(current):
+                    try:
+                        os.rmdir(current)
+                        logger.info("Removed empty legacy directory: %s", current)
+                    except Exception:
+                        pass
+
+            parent = os.path.dirname(current)
+            if parent == current:
+                break
+            current = parent
+
+    except Exception as exc:
+        logger.debug("Failed to cleanup empty legacy directories: %s", exc)
+
+
+def resolve_cache_path_with_migration(
+    cache_type: CacheType,
+    library_name: Optional[str] = None,
+    env_override: Optional[str] = None,
+) -> str:
+    """Resolve the cache file path, migrating from legacy locations if needed.
+
+    This function performs lazy migration: on first access, it checks if the
+    file exists at the canonical location. If not, it looks for legacy files
+    and copies them to the new location. After successful migration, the
+    legacy file is automatically removed.
+
+    Args:
+        cache_type: The type of cache file.
+        library_name: The library name (only used for MODEL and RECIPE types).
+        env_override: Optional environment variable value that overrides all
+            path resolution. When set, returns this path directly without
+            any migration.
+
+    Returns:
+        The resolved path to use for the cache file.
+    """
+    # Environment override bypasses all migration logic
+    if env_override:
+        return env_override
+
+    canonical_path = get_cache_file_path(cache_type, library_name, create_dir=True)
+
+    # If file already exists at canonical location, use it
+    if os.path.exists(canonical_path):
+        return canonical_path
+
+    # Check legacy paths for migration
+    legacy_paths = get_legacy_cache_paths(cache_type, library_name)
+
+    for legacy_path in legacy_paths:
+        if os.path.exists(legacy_path):
+            try:
+                shutil.copy2(legacy_path, canonical_path)
+                logger.info(
+                    "Migrated %s cache from %s to %s",
+                    cache_type.value,
+                    legacy_path,
+                    canonical_path,
+                )
+
+                _cleanup_legacy_file_after_migration(legacy_path, canonical_path)
+
+                return canonical_path
+            except Exception as exc:
+                logger.warning(
+                    "Failed to migrate %s cache from %s: %s",
+                    cache_type.value,
+                    legacy_path,
+                    exc,
+                )
+
+    # No legacy file found; return canonical path (will be created fresh)
+    return canonical_path
+
+
+def get_legacy_cache_files_for_cleanup() -> List[str]:
+    """Get a list of legacy cache files that can be removed after migration.
+
+    This function returns files that exist in legacy locations and have
+    corresponding files in the new canonical locations.
+
+    Returns:
+        A list of legacy file paths that are safe to remove.
+    """
+    files_to_remove: List[str] = []
+
+    try:
+        settings_dir = get_settings_dir(create=False)
+    except Exception:
+        return files_to_remove
+
+    # Check each cache type for migrated legacy files
+    for cache_type in CacheType:
+        # For MODEL and RECIPE, we need to check each library
+        if cache_type in (CacheType.MODEL, CacheType.RECIPE):
+            # Check default library
+            _check_legacy_for_cleanup(cache_type, "default", files_to_remove)
+            # Check for any per-library caches in legacy directories
+            legacy_dir_name = "model_cache" if cache_type == CacheType.MODEL else "recipe_cache"
+            legacy_dir = os.path.join(settings_dir, legacy_dir_name)
+            if os.path.isdir(legacy_dir):
+                try:
+                    for filename in os.listdir(legacy_dir):
+                        if filename.endswith(".sqlite"):
+                            library_name = filename[:-7]  # Remove .sqlite
+                            _check_legacy_for_cleanup(cache_type, library_name, files_to_remove)
+                except Exception:
+                    pass
+        else:
+            _check_legacy_for_cleanup(cache_type, None, files_to_remove)
+
+    return files_to_remove
+
+
+def _check_legacy_for_cleanup(
+    cache_type: CacheType,
+    library_name: Optional[str],
+    files_to_remove: List[str],
+) -> None:
+    """Check if a legacy cache file can be removed after migration.
+
+    Args:
+        cache_type: The type of cache file.
+        library_name: The library name (only used for MODEL and RECIPE types).
+        files_to_remove: List to append removable files to.
+    """
+    canonical_path = get_cache_file_path(cache_type, library_name, create_dir=False)
+    if not os.path.exists(canonical_path):
+        return
+
+    legacy_paths = get_legacy_cache_paths(cache_type, library_name)
+    for legacy_path in legacy_paths:
+        if os.path.exists(legacy_path) and legacy_path not in files_to_remove:
+            files_to_remove.append(legacy_path)
+
+
+def cleanup_legacy_cache_files(dry_run: bool = True) -> List[str]:
+    """Remove legacy cache files that have been migrated.
+
+    Args:
+        dry_run: If True, only return the list of files that would be removed
+            without actually removing them.
+
+    Returns:
+        A list of files that were (or would be) removed.
+    """
+    files = get_legacy_cache_files_for_cleanup()
+
+    if dry_run or not files:
+        return files
+
+    removed: List[str] = []
+    for file_path in files:
+        try:
+            os.remove(file_path)
+            removed.append(file_path)
+            logger.info("Removed legacy cache file: %s", file_path)
+        except Exception as exc:
+            logger.warning("Failed to remove legacy cache file %s: %s", file_path, exc)
+
+    # Try to remove empty legacy directories
+    try:
+        settings_dir = get_settings_dir(create=False)
+        for legacy_dir_name in ("model_cache", "recipe_cache"):
+            legacy_dir = os.path.join(settings_dir, legacy_dir_name)
+            if os.path.isdir(legacy_dir) and not os.listdir(legacy_dir):
+                os.rmdir(legacy_dir)
+                logger.info("Removed empty legacy directory: %s", legacy_dir)
+    except Exception:
+        pass
+
+    return removed

py/utils/constants.py

@@ -45,8 +45,13 @@ SUPPORTED_MEDIA_EXTENSIONS = {
     "videos": [".mp4", ".webm"],
 }
 
-# Valid Lora types
-VALID_LORA_TYPES = ["lora", "locon", "dora"]
+# Valid sub-types for each scanner type
+VALID_LORA_SUB_TYPES = ["lora", "locon", "dora"]
+VALID_CHECKPOINT_SUB_TYPES = ["checkpoint", "diffusion_model"]
+VALID_EMBEDDING_SUB_TYPES = ["embedding"]
+
+# Backward compatibility alias
+VALID_LORA_TYPES = VALID_LORA_SUB_TYPES
 
 # Supported Civitai model types for user model queries (case-insensitive)
 CIVITAI_USER_MODEL_TYPES = [

py/utils/example_images_download_manager.py

@@ -216,6 +216,11 @@ class DownloadManager:
                     self._progress["failed_models"] = set()
 
                 self._is_downloading = True
+                snapshot = self._progress.snapshot()
+
+                # Create the download task without awaiting it
+                # This ensures the HTTP response is returned immediately
+                # while the actual processing happens in the background
                 self._download_task = asyncio.create_task(
                     self._download_all_example_images(
                         output_dir,
@@ -227,7 +232,10 @@ class DownloadManager:
                     )
                 )
 
-                snapshot = self._progress.snapshot()
+                # Add a callback to handle task completion/errors
+                self._download_task.add_done_callback(
+                    lambda t: self._handle_download_task_done(t, output_dir)
+                )
             except ExampleImagesDownloadError:
                 # Re-raise our own exception types without wrapping
                 self._is_downloading = False
@@ -241,10 +249,25 @@ class DownloadManager:
                 )
                 raise ExampleImagesDownloadError(str(e)) from e
 
-        await self._broadcast_progress(status="running")
+        # Broadcast progress in the background without blocking the response
+        # This ensures the HTTP response is returned immediately
+        asyncio.create_task(self._broadcast_progress(status="running"))
 
         return {"success": True, "message": "Download started", "status": snapshot}
 
+    def _handle_download_task_done(self, task: asyncio.Task, output_dir: str) -> None:
+        """Handle download task completion, including saving progress on error."""
+        try:
+            # This will re-raise any exception from the task
+            task.result()
+        except Exception as e:
+            logger.error(f"Download task failed with error: {e}", exc_info=True)
+            # Ensure progress is saved even on failure
+            try:
+                self._save_progress(output_dir)
+            except Exception as save_error:
+                logger.error(f"Failed to save progress after task failure: {save_error}")
+
     async def get_status(self, request):
         """Get the current status of example images download."""
 
@@ -254,6 +277,130 @@ class DownloadManager:
             "status": self._progress.snapshot(),
         }
 
+    async def check_pending_models(self, model_types: list[str]) -> dict:
+        """Quickly check how many models need example images downloaded.
+        
+        This is a lightweight check that avoids the overhead of starting
+        a full download task when no work is needed.
+        
+        Returns:
+            dict with keys:
+                - total_models: Total number of models across specified types
+                - pending_count: Number of models needing example images
+                - processed_count: Number of already processed models
+                - failed_count: Number of models marked as failed
+                - needs_download: True if there are pending models to process
+        """
+        from ..services.service_registry import ServiceRegistry
+
+        if self._is_downloading:
+            return {
+                "success": True,
+                "is_downloading": True,
+                "total_models": 0,
+                "pending_count": 0,
+                "processed_count": 0,
+                "failed_count": 0,
+                "needs_download": False,
+                "message": "Download already in progress",
+            }
+
+        try:
+            # Get scanners
+            scanners = []
+            if "lora" in model_types:
+                lora_scanner = await ServiceRegistry.get_lora_scanner()
+                scanners.append(("lora", lora_scanner))
+
+            if "checkpoint" in model_types:
+                checkpoint_scanner = await ServiceRegistry.get_checkpoint_scanner()
+                scanners.append(("checkpoint", checkpoint_scanner))
+
+            if "embedding" in model_types:
+                embedding_scanner = await ServiceRegistry.get_embedding_scanner()
+                scanners.append(("embedding", embedding_scanner))
+
+            # Load progress file to check processed models
+            settings_manager = get_settings_manager()
+            active_library = settings_manager.get_active_library_name()
+            output_dir = self._resolve_output_dir(active_library)
+            
+            processed_models: set[str] = set()
+            failed_models: set[str] = set()
+            
+            if output_dir:
+                progress_file = os.path.join(output_dir, ".download_progress.json")
+                if os.path.exists(progress_file):
+                    try:
+                        with open(progress_file, "r", encoding="utf-8") as f:
+                            saved_progress = json.load(f)
+                            processed_models = set(saved_progress.get("processed_models", []))
+                            failed_models = set(saved_progress.get("failed_models", []))
+                    except Exception:
+                        pass  # Ignore progress file errors for quick check
+
+            # Count models
+            total_models = 0
+            models_with_hash = 0
+            
+            for scanner_type, scanner in scanners:
+                cache = await scanner.get_cached_data()
+                if cache and cache.raw_data:
+                    for model in cache.raw_data:
+                        total_models += 1
+                        if model.get("sha256"):
+                            models_with_hash += 1
+
+            # Calculate pending count
+            # A model is pending if it has a hash and is not in processed_models
+            # We also exclude failed_models unless force mode would be used
+            pending_count = models_with_hash - len(processed_models.intersection(
+                {m.get("sha256", "").lower() for scanner_type, scanner in scanners 
+                 for m in (await scanner.get_cached_data()).raw_data if m.get("sha256")}
+            ))
+            
+            # More accurate pending count: check which models actually need processing
+            pending_hashes = set()
+            for scanner_type, scanner in scanners:
+                cache = await scanner.get_cached_data()
+                if cache and cache.raw_data:
+                    for model in cache.raw_data:
+                        raw_hash = model.get("sha256")
+                        if not raw_hash:
+                            continue
+                        model_hash = raw_hash.lower()
+                        if model_hash not in processed_models:
+                            # Check if model folder exists with files
+                            model_dir = ExampleImagePathResolver.get_model_folder(
+                                model_hash, active_library
+                            )
+                            if not _model_directory_has_files(model_dir):
+                                pending_hashes.add(model_hash)
+
+            pending_count = len(pending_hashes)
+
+            return {
+                "success": True,
+                "is_downloading": False,
+                "total_models": total_models,
+                "pending_count": pending_count,
+                "processed_count": len(processed_models),
+                "failed_count": len(failed_models),
+                "needs_download": pending_count > 0,
+            }
+
+        except Exception as e:
+            logger.error(f"Error checking pending models: {e}", exc_info=True)
+            return {
+                "success": False,
+                "error": str(e),
+                "total_models": 0,
+                "pending_count": 0,
+                "processed_count": 0,
+                "failed_count": 0,
+                "needs_download": False,
+            }
+
     async def pause_download(self, request):
         """Pause the example images download."""
 

py/utils/example_images_processor.py

@@ -43,8 +43,15 @@ class ExampleImagesProcessor:
         return media_url
     
     @staticmethod
-    def _get_file_extension_from_content_or_headers(content, headers, fallback_url=None):
-        """Determine file extension from content magic bytes or headers"""
+    def _get_file_extension_from_content_or_headers(content, headers, fallback_url=None, media_type_hint=None):
+        """Determine file extension from content magic bytes or headers
+        
+        Args:
+            content: File content bytes
+            headers: HTTP response headers
+            fallback_url: Original URL for extension extraction
+            media_type_hint: Optional media type hint from metadata (e.g., "video" or "image")
+        """
         # Check magic bytes for common formats
         if content:
             if content.startswith(b'\xFF\xD8\xFF'):
@@ -82,6 +89,10 @@ class ExampleImagesProcessor:
             if ext in SUPPORTED_MEDIA_EXTENSIONS['images'] or ext in SUPPORTED_MEDIA_EXTENSIONS['videos']:
                 return ext
         
+        # Use media type hint from metadata if available
+        if media_type_hint == "video":
+            return '.mp4'
+        
         # Default fallback
         return '.jpg'
 
@@ -136,7 +147,7 @@ class ExampleImagesProcessor:
                 if success:
                     # Determine file extension from content or headers
                     media_ext = ExampleImagesProcessor._get_file_extension_from_content_or_headers(
-                        content, headers, original_url
+                        content, headers, original_url, image.get("type")
                     )
                     
                     # Check if the detected file type is supported
@@ -219,7 +230,7 @@ class ExampleImagesProcessor:
                 if success:
                     # Determine file extension from content or headers
                     media_ext = ExampleImagesProcessor._get_file_extension_from_content_or_headers(
-                        content, headers, original_url
+                        content, headers, original_url, image.get("type")
                     )
                     
                     # Check if the detected file type is supported

py/utils/lora_metadata.py

@@ -17,7 +17,7 @@ async def extract_lora_metadata(file_path: str) -> Dict:
                 base_model = determine_base_model(metadata.get("ss_base_model_version"))
                 return {"base_model": base_model}
     except Exception as e:
-        print(f"Error reading metadata from {file_path}: {str(e)}")
+        logger.error(f"Error reading metadata from {file_path}: {str(e)}")
     return {"base_model": "Unknown"}
 
 async def extract_checkpoint_metadata(file_path: str) -> dict:

py/utils/metadata_manager.py

@@ -223,7 +223,7 @@ class MetadataManager:
                     preview_url=normalize_path(preview_url),
                     tags=[],
                     modelDescription="",
-                    model_type="checkpoint",
+                    sub_type="checkpoint",
                     from_civitai=True
                 )
             elif model_class.__name__ == "EmbeddingMetadata":
@@ -238,6 +238,7 @@ class MetadataManager:
                     preview_url=normalize_path(preview_url),
                     tags=[],
                     modelDescription="",
+                    sub_type="embedding",
                     from_civitai=True
                 )
             else:  # Default to LoraMetadata

py/utils/models.py

@@ -173,14 +173,14 @@ class LoraMetadata(BaseModelMetadata):
 @dataclass
 class CheckpointMetadata(BaseModelMetadata):
     """Represents the metadata structure for a Checkpoint model"""
-    model_type: str = "checkpoint"  # Model type (checkpoint, diffusion_model, etc.)
+    sub_type: str = "checkpoint"  # Model sub-type (checkpoint, diffusion_model, etc.)
 
     @classmethod
     def from_civitai_info(cls, version_info: Dict, file_info: Dict, save_path: str) -> 'CheckpointMetadata':
         """Create CheckpointMetadata instance from Civitai version info"""
         file_name = file_info['name']
         base_model = determine_base_model(version_info.get('baseModel', ''))
-        model_type = version_info.get('type', 'checkpoint')
+        sub_type = version_info.get('type', 'checkpoint')
         
         # Extract tags and description if available
         tags = []
@@ -203,7 +203,7 @@ class CheckpointMetadata(BaseModelMetadata):
             preview_nsfw_level=0,
             from_civitai=True,
             civitai=version_info,
-            model_type=model_type,
+            sub_type=sub_type,
             tags=tags,
             modelDescription=description
         )
@@ -211,14 +211,14 @@ class CheckpointMetadata(BaseModelMetadata):
 @dataclass
 class EmbeddingMetadata(BaseModelMetadata):
     """Represents the metadata structure for an Embedding model"""
-    model_type: str = "embedding"  # Model type (embedding, textual_inversion, etc.)
+    sub_type: str = "embedding"
 
     @classmethod
     def from_civitai_info(cls, version_info: Dict, file_info: Dict, save_path: str) -> 'EmbeddingMetadata':
         """Create EmbeddingMetadata instance from Civitai version info"""
         file_name = file_info['name']
         base_model = determine_base_model(version_info.get('baseModel', ''))
-        model_type = version_info.get('type', 'embedding')
+        sub_type = version_info.get('type', 'embedding')
         
         # Extract tags and description if available
         tags = []
@@ -241,7 +241,7 @@ class EmbeddingMetadata(BaseModelMetadata):
             preview_nsfw_level=0,
             from_civitai=True,
             civitai=version_info,
-            model_type=model_type,
+            sub_type=sub_type,
             tags=tags,
             modelDescription=description
         )

py/utils/utils.py

@@ -138,19 +138,15 @@ def calculate_recipe_fingerprint(loras):
     if not loras:
         return ""
     
-    # Filter valid entries and extract hash and strength
     valid_loras = []
     for lora in loras:
-        # Skip excluded loras
         if lora.get("exclude", False):
             continue
             
-        # Get the hash - use modelVersionId as fallback if hash is empty
         hash_value = lora.get("hash", "").lower()
-        if not hash_value and lora.get("isDeleted", False) and lora.get("modelVersionId"):
+        if not hash_value and lora.get("modelVersionId"):
             hash_value = str(lora.get("modelVersionId"))
             
-        # Skip entries without a valid hash
         if not hash_value:
             continue
             

pyproject.toml

@@ -1,7 +1,7 @@
 [project]
 name = "comfyui-lora-manager"
 description = "Revolutionize your workflow with the ultimate LoRA companion for ComfyUI!"
-version = "0.9.13"
+version = "0.9.15"
 license = {file = "LICENSE"}
 dependencies = [
     "aiohttp",

static/css/components/card.css

@@ -113,6 +113,12 @@
     max-width: 110px;
 }
 
+/* Compact mode: hide sub-type to save space */
+.compact-density .model-sub-type,
+.compact-density .model-separator {
+    display: none;
+}
+
 .compact-density .card-actions i {
     font-size: 0.95em;
     padding: 3px;
@@ -296,6 +302,19 @@
     min-height: 20px;
 }
 
+/* Gradient overlay on right side for icon readability */
+.card-header::after {
+    content: '';
+    position: absolute;
+    top: 0;
+    right: 0;
+    width: 100px;
+    height: 100%;
+    background: linear-gradient(to left, oklch(0% 0 0 / 0.4) 0%, transparent 100%);
+    pointer-events: none;
+    border-top-right-radius: var(--border-radius);
+}
+
 .card-header-info {
     display: flex;
     align-items: center;
@@ -426,15 +445,39 @@
     white-space: nowrap;
     overflow: hidden;
     text-overflow: ellipsis;
-    display: inline-block;
+    display: inline-flex;
+    align-items: center;
+    gap: 4px;
     color: white;
     text-shadow: 1px 1px 2px rgba(0, 0, 0, 0.5);
-    background: rgba(255, 255, 255, 0.2);
-    padding: 2px var(--space-1);
+    background: rgba(255, 255, 255, 0.12);
+    padding: 2px 6px;
     border-radius: var(--border-radius-xs);
-    backdrop-filter: blur(2px);
-    font-size: 0.85em;
+    backdrop-filter: blur(4px);
+    font-size: 0.8em;
     line-height: 1.2;
+    font-weight: 500;
+}
+
+/* Subtle separator between sub-type and base model */
+.model-separator {
+    width: 1px;
+    height: 0.6em;
+    background: rgba(255, 255, 255, 0.25);
+    flex-shrink: 0;
+}
+
+/* Sub-type abbreviation styling */
+.model-sub-type {
+    opacity: 0.9;
+    flex-shrink: 0;
+}
+
+/* Base model abbreviation styling */
+.model-base-type {
+    flex-shrink: 1;
+    overflow: hidden;
+    text-overflow: ellipsis;
 }
 
 /* Style for version name */
@@ -596,18 +639,22 @@
 }
 
 .model-update-badge {
-    display: inline-flex;
-    align-items: center;
-    gap: 6px;
-    padding: 2px 10px;
-    border-radius: var(--border-radius-xs);
+    width: 18px;
+    height: 18px;
+    padding: 0;
+    border-radius: 4px;
     background: var(--badge-update-bg);
     color: var(--badge-update-text);
     font-size: 0.7rem;
-    font-weight: 600;
-    letter-spacing: 0.04em;
-    text-transform: uppercase;
-    box-shadow: 0 4px 12px var(--badge-update-glow);
+    display: inline-flex;
+    align-items: center;
+    justify-content: center;
+    flex-shrink: 0;
+    box-shadow: 0 2px 6px var(--badge-update-glow);
     border: 1px solid color-mix(in oklab, var(--badge-update-bg) 55%, transparent);
-    white-space: nowrap;
+}
+
+.model-update-badge i {
+    margin-left: 1px;
+    line-height: 1;
 }

static/css/components/lora-modal/showcase.css

@@ -482,3 +482,75 @@
 [data-theme="dark"] .import-container {
   background: rgba(255, 255, 255, 0.03);
 }
+
+/* Setup Guidance State - When example images path is not configured */
+.import-container--needs-setup {
+  cursor: default;
+  border-style: solid;
+  border-color: var(--border-color);
+  background: var(--lora-surface);
+}
+
+.import-container--needs-setup:hover {
+  border-color: var(--border-color);
+  transform: none;
+}
+
+.import-setup-guidance {
+  display: flex;
+  flex-direction: column;
+  align-items: center;
+  gap: var(--space-2);
+  padding: var(--space-3) var(--space-2);
+  text-align: center;
+}
+
+.setup-icon {
+  width: 64px;
+  height: 64px;
+  border-radius: 50%;
+  background: oklch(var(--lora-accent-l) var(--lora-accent-c) var(--lora-accent-h) / 0.1);
+  display: flex;
+  align-items: center;
+  justify-content: center;
+  margin-bottom: var(--space-1);
+}
+
+.setup-icon i {
+  font-size: 1.75rem;
+  color: var(--lora-accent);
+}
+
+.import-setup-guidance h3 {
+  margin: 0;
+  font-size: 1.2rem;
+  font-weight: 600;
+  color: var(--text-color);
+}
+
+.setup-description {
+  margin: 0;
+  color: var(--text-color);
+  opacity: 0.9;
+  max-width: 320px;
+  line-height: 1.5;
+}
+
+.setup-usage {
+  margin: 0;
+  font-size: 0.85em;
+  color: var(--text-color);
+  opacity: 0.6;
+  font-style: italic;
+}
+
+.setup-settings-btn {
+  margin-top: var(--space-2);
+  background: var(--lora-accent) !important;
+  color: var(--lora-text) !important;
+}
+
+.setup-settings-btn:hover {
+  opacity: 0.9;
+  transform: translateY(-1px);
+}

static/css/components/modal/update-modal.css

@@ -1,4 +1,146 @@
-/* Update Modal specific styles */
+/* Update Modal Styles */
+.update-header {
+    display: flex;
+    align-items: center;
+    gap: var(--space-2);
+    margin-bottom: var(--space-3);
+    padding-bottom: var(--space-2);
+    border-bottom: 1px solid var(--lora-border);
+}
+
+.notification-tabs {
+    display: flex;
+    gap: var(--space-2);
+    margin-bottom: var(--space-3);
+}
+
+.notification-tab {
+    display: inline-flex;
+    align-items: center;
+    gap: var(--space-1);
+    padding: 0.5rem 0.75rem;
+    background: var(--lora-surface);
+    border: 1px solid var(--lora-border);
+    border-radius: var(--border-radius-sm);
+    color: var(--text-color);
+    cursor: pointer;
+    transition: background 0.2s ease, border-color 0.2s ease;
+    font-weight: 500;
+}
+
+.notification-tab:hover,
+.notification-tab.active {
+    background: var(--lora-accent-light, rgba(0, 148, 255, 0.12));
+    border-color: var(--lora-accent);
+    color: var(--lora-accent-text, var(--text-color));
+}
+
+.notification-tab-badge {
+    display: none;
+    min-width: 1.25rem;
+    height: 1.25rem;
+    padding: 0 0.4rem;
+    border-radius: 999px;
+    background: var(--lora-accent);
+    color: #fff;
+    font-size: 0.75rem;
+    font-weight: 600;
+    align-items: center;
+    justify-content: center;
+}
+
+.notification-tab-badge.is-dot {
+    min-width: 0.5rem;
+    width: 0.5rem;
+    height: 0.5rem;
+    padding: 0;
+    border-radius: 50%;
+}
+
+.notification-tab-badge.visible {
+    display: inline-flex;
+}
+
+.notification-panels {
+    display: flex;
+    flex-direction: column;
+    gap: var(--space-3);
+}
+
+.notification-panel {
+    display: none;
+}
+
+.notification-panel.active {
+    display: block;
+}
+
+.update-icon {
+    font-size: 1.8em;
+    color: var(--lora-accent);
+    animation: bounce 1.5s infinite;
+}
+
+@keyframes bounce {
+    0%, 100% {
+        transform: translateY(0);
+    }
+    50% {
+        transform: translateY(-5px);
+    }
+}
+
+.update-content {
+    display: flex;
+    flex-direction: column;
+    gap: var(--space-3);
+}
+
+.update-info {
+    display: flex;
+    justify-content: space-between;
+    align-items: center;
+    border-radius: var(--border-radius-sm);
+    padding: var(--space-3);
+}
+
+.update-info .version-info {
+    display: flex;
+    flex-direction: column;
+    gap: 8px;
+}
+
+.current-version, .new-version {
+    display: flex;
+    align-items: center;
+    gap: 10px;
+}
+
+.label {
+    font-size: 0.9em;
+    color: var(--text-color);
+    opacity: 0.8;
+}
+
+.version-number {
+    font-family: monospace;
+    font-weight: 600;
+}
+
+.new-version .version-number {
+    color: var(--lora-accent);
+}
+
+/* Add styling for git info display */
+.git-info {
+    font-size: 0.85em;
+    opacity: 0.7;
+    margin-top: 4px;
+    font-family: monospace;
+    color: var(--text-color);
+}
+
+/* Update actions container */
 .update-actions {
     display: flex;
     flex-direction: column;
@@ -7,17 +149,24 @@
     flex-wrap: nowrap;
 }
 
+/* GitHub link button styling */
 .update-link {
-    color: var(--lora-accent);
-    text-decoration: none;
     display: flex;
     align-items: center;
     gap: 8px;
-    font-size: 0.95em;
+    padding: 8px 16px;
+    background: var(--lora-surface);
+    border: 1px solid var(--lora-border);
+    border-radius: var(--border-radius-sm);
+    text-decoration: none;
+    color: var(--text-color);
+    transition: all 0.2s ease;
 }
 
 .update-link:hover {
-    text-decoration: underline;
+    background: var(--lora-accent);
+    color: white;
+    transform: translateY(-2px);
 }
 
 /* Update progress styles */
@@ -83,15 +232,96 @@
     background-color: var(--lora-error);
 }
 
-/* Add styles for markdown elements in changelog */
+/* Changelog section */
+.changelog-section {
+    background: rgba(0, 0, 0, 0.02);
+    border: 1px solid rgba(0, 0, 0, 0.08);
+    border-radius: var(--border-radius-sm);
+    padding: var(--space-3);
+}
+
+.changelog-section h3 {
+    margin-top: 0;
+    margin-bottom: var(--space-2);
+    color: var(--lora-accent);
+    font-size: 1.1em;
+}
+
+.changelog-content {
+    max-height: 550px;
+    overflow-y: auto;
+    padding-left: var(--space-3);
+}
+
+.changelog-item {
+    margin-bottom: var(--space-2);
+    padding-bottom: var(--space-2);
+    border-bottom: 1px solid var(--lora-border);
+}
+
+.changelog-item:last-child {
+    margin-bottom: 0;
+    padding-bottom: 0;
+    border-bottom: none;
+}
+
+/* Multiple releases styling */
+.changelog-content > .changelog-item + .changelog-item {
+    margin-top: var(--space-4);
+    padding-top: var(--space-4);
+    border-top: 1px solid var(--lora-border);
+}
+
+.changelog-item h4 {
+    display: flex;
+    flex-wrap: wrap;
+    align-items: center;
+    gap: var(--space-2);
+    margin-bottom: var(--space-2);
+    font-size: 1em;
+    margin-top: 0;
+    color: var(--text-color);
+}
+
+.changelog-item .latest-badge {
+    background-color: var(--lora-accent);
+    color: white;
+    padding: 2px 8px;
+    border-radius: 12px;
+    font-size: 0.75em;
+    font-weight: 500;
+    text-transform: uppercase;
+}
+
+.changelog-item .publish-date {
+    font-size: 0.85em;
+    color: var(--text-color);
+    opacity: 0.6;
+    font-weight: normal;
+}
+
+.changelog-item.latest {
+    background-color: rgba(66, 153, 225, 0.05);
+    border-radius: var(--border-radius-sm);
+    padding: var(--space-2);
+    border: 1px solid rgba(66, 153, 225, 0.2);
+}
+
+[data-theme="dark"] .changelog-item.latest {
+    background-color: rgba(66, 153, 225, 0.1);
+    border-color: rgba(66, 153, 225, 0.3);
+}
+
+/* Changelog markdown styles */
 .changelog-item ul {
+    margin: 0;
     padding-left: 20px;
-    margin-top: 8px;
 }
 
 .changelog-item li {
     margin-bottom: 6px;
     line-height: 1.4;
+    color: var(--text-color);
 }
 
 .changelog-item strong {
@@ -121,4 +351,191 @@
 
 .changelog-item a:hover {
     text-decoration: underline;
-}
+}
+
+/* Update preferences section */
+.update-preferences {
+    border-top: 1px solid var(--lora-border);
+    margin-top: var(--space-2);
+    padding-top: var(--space-2);
+    display: flex;
+    align-items: center;
+    justify-content: flex-start;
+}
+
+/* Override toggle switch styles for update preferences */
+.update-preferences .toggle-switch {
+    position: relative;
+    display: inline-flex;
+    align-items: center;
+    width: auto;
+    height: 24px;
+    cursor: pointer;
+}
+
+.update-preferences .toggle-slider {
+    position: relative;
+    display: inline-block;
+    width: 50px;
+    height: 24px;
+    flex-shrink: 0;
+    margin-right: 10px;
+}
+
+.update-preferences .toggle-label {
+    margin-left: 0;
+    white-space: nowrap;
+    line-height: 24px;
+}
+
+/* Banner history */
+.banner-history {
+    display: flex;
+    flex-direction: column;
+    gap: var(--space-2);
+}
+
+.banner-history h3 {
+    margin: 0;
+    font-size: 1.05rem;
+    color: var(--lora-accent);
+}
+
+.banner-history-empty {
+    margin: 0;
+    padding: var(--space-3);
+    background: var(--lora-surface);
+    border: 1px dashed var(--lora-border);
+    border-radius: var(--border-radius-sm);
+    text-align: center;
+    color: var(--text-muted, rgba(0, 0, 0, 0.6));
+}
+
+.banner-history-list {
+    list-style: none;
+    margin: 0;
+    padding: 0;
+    display: flex;
+    flex-direction: column;
+    gap: var(--space-2);
+}
+
+.banner-history-item {
+    border: 1px solid var(--lora-border);
+    border-radius: var(--border-radius-sm);
+    padding: var(--space-2);
+    background: var(--card-bg, #fff);
+    display: flex;
+    flex-direction: column;
+    gap: var(--space-1);
+}
+
+[data-theme="dark"] .banner-history-item {
+    background: rgba(255, 255, 255, 0.03);
+}
+
+.banner-history-title {
+    margin: 0;
+    font-size: 1rem;
+}
+
+.banner-history-description {
+    margin: 0;
+    color: var(--text-color);
+    opacity: 0.85;
+}
+
+.banner-history-meta {
+    display: flex;
+    gap: var(--space-2);
+    font-size: 0.85rem;
+    color: var(--text-muted, rgba(0, 0, 0, 0.6));
+    flex-wrap: wrap;
+}
+
+.banner-history-time {
+    display: inline-flex;
+    align-items: center;
+}
+
+.banner-history-status {
+    display: inline-flex;
+    align-items: center;
+    gap: 0.35rem;
+    font-weight: 600;
+    text-transform: uppercase;
+    letter-spacing: 0.05em;
+}
+
+.banner-history-status.active {
+    color: var(--lora-success);
+}
+
+.banner-history-status.dismissed {
+    color: var(--lora-error);
+}
+
+.banner-history-actions {
+    display: flex;
+    flex-wrap: wrap;
+    gap: var(--space-2);
+    margin-top: var(--space-1);
+}
+
+.banner-history-action {
+    display: inline-flex;
+    align-items: center;
+    gap: 0.35rem;
+    padding: 0.35rem 0.65rem;
+    border-radius: var(--border-radius-sm);
+    border: 1px solid var(--lora-border);
+    text-decoration: none;
+    font-size: 0.85rem;
+    transition: background 0.2s ease, color 0.2s ease, border-color 0.2s ease;
+}
+
+.banner-history-action i {
+    font-size: 0.9rem;
+}
+
+.banner-history-action.banner-history-action-primary {
+    background: var(--lora-accent);
+    border-color: var(--lora-accent);
+    color: #fff;
+}
+
+.banner-history-action.banner-history-action-secondary {
+    background: var(--lora-surface);
+    color: var(--text-color);
+}
+
+.banner-history-action.banner-history-action-tertiary {
+    background: transparent;
+    border-style: dashed;
+}
+
+.banner-history-action:hover {
+    background: var(--lora-accent-light, rgba(0, 148, 255, 0.12));
+    border-color: var(--lora-accent);
+    color: var(--lora-accent-text, var(--text-color));
+}
+
+@media (max-width: 480px) {
+    .update-info {
+        flex-direction: column;
+        gap: var(--space-2);
+    }
+
+    .version-info {
+        width: 100%;
+    }
+
+    .update-preferences {
+        flex-direction: row;
+        flex-wrap: wrap;
+    }
+
+    .update-preferences .toggle-label {
+        margin-top: 5px;
+    }
+}

static/css/components/search-filter.css

@@ -466,6 +466,213 @@
     border-color: var(--lora-accent);
 }
 
+/* Presets Section Styles */
+.presets-section {
+    border-bottom: 1px solid var(--border-color);
+    padding-bottom: 16px;
+    margin-bottom: 16px;
+}
+
+.presets-section h4 {
+    margin: 0 0 8px 0;
+}
+
+.filter-presets {
+    display: flex;
+    flex-wrap: wrap;
+    gap: 6px;
+}
+
+.filter-preset {
+    display: inline-flex;
+    align-items: center;
+    gap: 8px;
+    padding: 4px 4px 4px 10px;
+    border-radius: var(--border-radius-sm);
+    background-color: var(--lora-surface);
+    border: 1px solid var(--border-color);
+    transition: all 0.2s ease;
+    cursor: pointer;
+}
+
+.filter-preset:hover {
+    background-color: var(--lora-surface-hover);
+    border-color: var(--lora-accent);
+}
+
+.filter-preset.active {
+    background-color: var(--lora-accent);
+    border-color: var(--lora-accent);
+}
+
+.filter-preset.active .preset-name {
+    color: white;
+    font-weight: 600;
+}
+
+.filter-preset.active .preset-delete-btn {
+    color: white;
+    opacity: 0;
+}
+
+.filter-preset:hover.active .preset-delete-btn {
+    opacity: 0.8;
+}
+
+.filter-preset.active .preset-delete-btn:hover {
+    opacity: 1;
+    color: white;
+}
+
+.preset-name {
+    cursor: pointer;
+    font-size: 14px;
+    color: var(--text-color);
+    user-select: none;
+    display: flex;
+    align-items: center;
+    gap: 6px;
+    white-space: nowrap;
+    max-width: 120px; /* Prevent long names from breaking layout */
+    overflow: hidden;
+    text-overflow: ellipsis;
+}
+
+.preset-delete-btn {
+    background: none;
+    border: none;
+    color: var(--text-color);
+    opacity: 0; /* Hidden by default */
+    cursor: pointer;
+    padding: 4px;
+    display: flex;
+    align-items: center;
+    justify-content: center;
+    font-size: 11px;
+    transition: all 0.2s ease;
+    margin-left: auto;
+}
+
+.filter-preset:hover .preset-delete-btn {
+    opacity: 0.5; /* Show on hover */
+}
+
+.preset-delete-btn:hover {
+    opacity: 1;
+    color: var(--lora-error, #e74c3c);
+}
+
+.add-preset-btn {
+    background-color: transparent;
+    border: 1px dashed var(--border-color);
+    color: var(--text-color);
+    opacity: 0.85;
+    display: inline-flex;
+    align-items: center;
+    gap: 6px;
+    padding: 4px 10px;
+    font-size: 14px;
+    border-radius: var(--border-radius-sm);
+    cursor: pointer;
+    transition: all 0.25s ease;
+}
+
+/* Enabled state - visual cue that button is actionable */
+.add-preset-btn:not(.disabled) {
+    border-color: var(--lora-accent);
+    border-style: solid;
+    background-color: rgba(66, 153, 225, 0.08);
+}
+
+.add-preset-btn:hover:not(.disabled) {
+    opacity: 1;
+    background-color: rgba(66, 153, 225, 0.15);
+    color: var(--lora-accent);
+    transform: translateY(-1px);
+    box-shadow: 0 2px 6px rgba(66, 153, 225, 0.2);
+}
+
+/* Disabled state - clear "unavailable" visual language */
+.add-preset-btn.disabled {
+    opacity: 0.35;
+    cursor: not-allowed;
+    background-color: rgba(128, 128, 128, 0.05);
+    border-style: dashed;
+    border-color: var(--border-color);
+    color: var(--text-muted);
+}
+
+.add-preset-btn i {
+    font-size: 12px;
+    transition: transform 0.2s ease;
+}
+
+/* Inline preset naming input */
+.preset-inline-input-container {
+    display: inline-flex;
+    align-items: center;
+    gap: 4px;
+    padding: 2px;
+    background-color: var(--lora-surface);
+    border: 1px solid var(--lora-accent);
+    border-radius: var(--border-radius-sm);
+}
+
+.preset-inline-input {
+    width: 120px;
+    padding: 4px 8px;
+    border: none;
+    background: transparent;
+    color: var(--text-color);
+    font-size: 13px;
+    outline: none;
+}
+
+.preset-inline-input::placeholder {
+    color: var(--text-color);
+    opacity: 0.5;
+}
+
+.preset-inline-btn {
+    background: none;
+    border: none;
+    color: var(--text-color);
+    cursor: pointer;
+    padding: 4px 6px;
+    display: flex;
+    align-items: center;
+    justify-content: center;
+    font-size: 12px;
+    transition: color 0.2s ease;
+    opacity: 0.7;
+}
+
+.preset-inline-btn:hover {
+    opacity: 1;
+}
+
+.preset-inline-btn.save:hover {
+    color: var(--lora-accent);
+}
+
+.preset-inline-btn.cancel:hover {
+    color: var(--lora-error, #e74c3c);
+}
+
+/* Two-step delete confirmation */
+.preset-delete-btn.confirm {
+    color: var(--lora-accent);
+    opacity: 1;
+    animation: pulse-confirm 0.5s ease-in-out infinite alternate;
+}
+
+@keyframes pulse-confirm {
+    from { opacity: 0.7; }
+    to { opacity: 1; }
+}
+
+
+
 /* Mobile adjustments */
 @media (max-width: 768px) {
     .search-options-panel,

static/css/components/update-modal.css

@@ -1,400 +0,0 @@
-/* Update Modal Styles */
-.update-modal {
-    max-width: 600px;
-}
-
-.update-header {
-    display: flex;
-    align-items: center;
-    gap: var(--space-2);
-    margin-bottom: var(--space-3);
-    padding-bottom: var(--space-2);
-    border-bottom: 1px solid var(--lora-border);
-}
-
-.notification-tabs {
-    display: flex;
-    gap: var(--space-2);
-    margin-bottom: var(--space-3);
-}
-
-.notification-tab {
-    display: inline-flex;
-    align-items: center;
-    gap: var(--space-1);
-    padding: 0.5rem 0.75rem;
-    background: var(--lora-surface);
-    border: 1px solid var(--lora-border);
-    border-radius: var(--border-radius-sm);
-    color: var(--text-color);
-    cursor: pointer;
-    transition: background 0.2s ease, border-color 0.2s ease;
-    font-weight: 500;
-}
-
-.notification-tab:hover,
-.notification-tab.active {
-    background: var(--lora-accent-light, rgba(0, 148, 255, 0.12));
-    border-color: var(--lora-accent);
-    color: var(--lora-accent-text, var(--text-color));
-}
-
-.notification-tab-badge {
-    display: none;
-    min-width: 1.25rem;
-    height: 1.25rem;
-    padding: 0 0.4rem;
-    border-radius: 999px;
-    background: var(--lora-accent);
-    color: #fff;
-    font-size: 0.75rem;
-    font-weight: 600;
-    align-items: center;
-    justify-content: center;
-}
-
-.notification-tab-badge.is-dot {
-    min-width: 0.5rem;
-    width: 0.5rem;
-    height: 0.5rem;
-    padding: 0;
-    border-radius: 50%;
-}
-
-.notification-tab-badge.visible {
-    display: inline-flex;
-}
-
-.notification-panels {
-    display: flex;
-    flex-direction: column;
-    gap: var(--space-3);
-}
-
-.notification-panel {
-    display: none;
-}
-
-.notification-panel.active {
-    display: block;
-}
-
-.update-icon {
-    font-size: 1.8em;
-    color: var(--lora-accent);
-    animation: bounce 1.5s infinite;
-}
-
-@keyframes bounce {
-    0%, 100% {
-        transform: translateY(0);
-    }
-    50% {
-        transform: translateY(-5px);
-    }
-}
-
-.update-content {
-    display: flex;
-    flex-direction: column;
-    gap: var(--space-3);
-}
-
-.update-info {
-    display: flex;
-    justify-content: space-between;
-    align-items: center;
-    border-radius: var(--border-radius-sm);
-    padding: var(--space-3);
-}
-
-.update-info .version-info {
-    display: flex;
-    flex-direction: column;
-    gap: 8px;
-}
-
-.current-version, .new-version {
-    display: flex;
-    align-items: center;
-    gap: 10px;
-}
-
-.label {
-    font-size: 0.9em;
-    color: var(--text-color);
-    opacity: 0.8;
-}
-
-.version-number {
-    font-family: monospace;
-    font-weight: 600;
-}
-
-.new-version .version-number {
-    color: var(--lora-accent);
-}
-
-/* Add styling for git info display */
-.git-info {
-    font-size: 0.85em;
-    opacity: 0.7;
-    margin-top: 4px;
-    font-family: monospace;
-    color: var(--text-color);
-}
-
-.update-link {
-    display: flex;
-    align-items: center;
-    gap: 8px;
-    padding: 8px 16px;
-    background: var(--lora-surface);
-    border: 1px solid var(--lora-border);
-    border-radius: var(--border-radius-sm);
-    text-decoration: none;
-    color: var(--text-color);
-    transition: all 0.2s ease;
-}
-
-.update-link:hover {
-    background: var(--lora-accent);
-    color: white;
-    transform: translateY(-2px);
-}
-
-.changelog-section {
-    background: rgba(0, 0, 0, 0.02); /* 轻微的灰色背景 */
-    border: 1px solid rgba(0, 0, 0, 0.08); /* 更明显的边框 */
-    border-radius: var(--border-radius-sm);
-    padding: var(--space-3);
-}
-
-.changelog-section h3 {
-    margin-top: 0;
-    margin-bottom: var(--space-2);
-    color: var(--lora-accent);
-    font-size: 1.1em;
-}
-
-.changelog-content {
-    max-height: 300px; /* Increased height since we removed instructions */
-    overflow-y: auto;
-}
-
-.changelog-item {
-    margin-bottom: var(--space-2);
-    padding-bottom: var(--space-2);
-    border-bottom: 1px solid var(--lora-border);
-}
-
-.changelog-item:last-child {
-    margin-bottom: 0;
-    padding-bottom: 0;
-    border-bottom: none;
-}
-
-.changelog-item h4 {
-    margin-top: 0;
-    margin-bottom: 8px;
-    font-size: 1em;
-    color: var(--text-color);
-}
-
-.changelog-item ul {
-    margin: 0;
-    padding-left: 20px;
-}
-
-.changelog-item li {
-    margin-bottom: 4px;
-    color: var(--text-color);
-}
-
-@media (max-width: 480px) {
-    .update-info {
-        flex-direction: column;
-        gap: var(--space-2);
-    }
-    
-    .version-info {
-        width: 100%;
-    }
-}
-
-/* Update preferences section */
-.update-preferences {
-    border-top: 1px solid var(--lora-border);
-    margin-top: var(--space-2);
-    padding-top: var(--space-2);
-    display: flex;
-    align-items: center;
-    justify-content: flex-start;
-}
-
-.banner-history {
-    display: flex;
-    flex-direction: column;
-    gap: var(--space-2);
-}
-
-.banner-history h3 {
-    margin: 0;
-    font-size: 1.05rem;
-    color: var(--lora-accent);
-}
-
-.banner-history-empty {
-    margin: 0;
-    padding: var(--space-3);
-    background: var(--lora-surface);
-    border: 1px dashed var(--lora-border);
-    border-radius: var(--border-radius-sm);
-    text-align: center;
-    color: var(--text-muted, rgba(0, 0, 0, 0.6));
-}
-
-.banner-history-list {
-    list-style: none;
-    margin: 0;
-    padding: 0;
-    display: flex;
-    flex-direction: column;
-    gap: var(--space-2);
-}
-
-.banner-history-item {
-    border: 1px solid var(--lora-border);
-    border-radius: var(--border-radius-sm);
-    padding: var(--space-2);
-    background: var(--card-bg, #fff);
-    display: flex;
-    flex-direction: column;
-    gap: var(--space-1);
-}
-
-[data-theme="dark"] .banner-history-item {
-    background: rgba(255, 255, 255, 0.03);
-}
-
-.banner-history-title {
-    margin: 0;
-    font-size: 1rem;
-}
-
-.banner-history-description {
-    margin: 0;
-    color: var(--text-color);
-    opacity: 0.85;
-}
-
-.banner-history-meta {
-    display: flex;
-    gap: var(--space-2);
-    font-size: 0.85rem;
-    color: var(--text-muted, rgba(0, 0, 0, 0.6));
-    flex-wrap: wrap;
-}
-
-.banner-history-time {
-    display: inline-flex;
-    align-items: center;
-}
-
-.banner-history-status {
-    display: inline-flex;
-    align-items: center;
-    gap: 0.35rem;
-    font-weight: 600;
-    text-transform: uppercase;
-    letter-spacing: 0.05em;
-}
-
-.banner-history-status.active {
-    color: var(--lora-success);
-}
-
-.banner-history-status.dismissed {
-    color: var(--lora-error);
-}
-
-.banner-history-actions {
-    display: flex;
-    flex-wrap: wrap;
-    gap: var(--space-2);
-    margin-top: var(--space-1);
-}
-
-.banner-history-action {
-    display: inline-flex;
-    align-items: center;
-    gap: 0.35rem;
-    padding: 0.35rem 0.65rem;
-    border-radius: var(--border-radius-sm);
-    border: 1px solid var(--lora-border);
-    text-decoration: none;
-    font-size: 0.85rem;
-    transition: background 0.2s ease, color 0.2s ease, border-color 0.2s ease;
-}
-
-.banner-history-action i {
-    font-size: 0.9rem;
-}
-
-.banner-history-action.banner-history-action-primary {
-    background: var(--lora-accent);
-    border-color: var(--lora-accent);
-    color: #fff;
-}
-
-.banner-history-action.banner-history-action-secondary {
-    background: var(--lora-surface);
-    color: var(--text-color);
-}
-
-.banner-history-action.banner-history-action-tertiary {
-    background: transparent;
-    border-style: dashed;
-}
-
-.banner-history-action:hover {
-    background: var(--lora-accent-light, rgba(0, 148, 255, 0.12));
-    border-color: var(--lora-accent);
-    color: var(--lora-accent-text, var(--text-color));
-}
-
-/* Override toggle switch styles for update preferences */
-.update-preferences .toggle-switch {
-    position: relative;
-    display: inline-flex;
-    align-items: center;
-    width: auto;
-    height: 24px;
-    cursor: pointer;
-}
-
-.update-preferences .toggle-slider {
-    position: relative;
-    display: inline-block;
-    width: 50px;
-    height: 24px;
-    flex-shrink: 0;
-    margin-right: 10px;
-}
-
-.update-preferences .toggle-label {
-    margin-left: 0;
-    white-space: nowrap;
-    line-height: 24px;
-}
-
-@media (max-width: 480px) {
-    .update-preferences {
-        flex-direction: row;
-        flex-wrap: wrap;
-    }
-    
-    .update-preferences .toggle-label {
-        margin-top: 5px;
-    }
-}

static/css/style.css

@@ -20,7 +20,6 @@
 @import 'components/toast.css';
 @import 'components/loading.css';
 @import 'components/menu.css';
-@import 'components/update-modal.css';
 @import 'components/lora-modal/lora-modal.css';
 @import 'components/lora-modal/description.css';
 @import 'components/lora-modal/tag.css';

static/js/api/baseModelApi.js

@@ -59,6 +59,18 @@ export class BaseModelApiClient {
                 sort_by: pageState.sortBy
             }, pageState);
 
+            // If params is null, it means wildcard resolved to no matches - return empty results
+            if (params === null) {
+                return {
+                    items: [],
+                    totalItems: 0,
+                    totalPages: 0,
+                    currentPage: page,
+                    hasMore: false,
+                    folders: []
+                };
+            }
+
             const response = await fetch(`${this.apiConfig.endpoints.list}?${params}`);
             if (!response.ok) {
                 throw new Error(`Failed to fetch ${this.apiConfig.config.displayName}s: ${response.statusText}`);
@@ -868,6 +880,13 @@ export class BaseModelApiClient {
             }
 
             if (pageState.filters.baseModel && pageState.filters.baseModel.length > 0) {
+                // Check for empty wildcard marker - if present, no models should match
+                const EMPTY_WILDCARD_MARKER = '__EMPTY_WILDCARD_RESULT__';
+                if (pageState.filters.baseModel.length === 1 && 
+                    pageState.filters.baseModel[0] === EMPTY_WILDCARD_MARKER) {
+                    // Wildcard resolved to no matches - return empty results
+                    return null;  // Signal to return empty results
+                }
                 pageState.filters.baseModel.forEach(model => {
                     params.append('base_model', model);
                 });

static/js/api/recipeApi.js

@@ -103,6 +103,19 @@ export async function fetchRecipesPage(page = 1, pageSize = 100) {
 
             // Add base model filters
             if (pageState.filters?.baseModel && pageState.filters.baseModel.length) {
+                // Check for empty wildcard marker - if present, no models should match
+                const EMPTY_WILDCARD_MARKER = '__EMPTY_WILDCARD_RESULT__';
+                if (pageState.filters.baseModel.length === 1 && 
+                    pageState.filters.baseModel[0] === EMPTY_WILDCARD_MARKER) {
+                    // Wildcard resolved to no matches - return empty results
+                    return {
+                        items: [],
+                        totalItems: 0,
+                        totalPages: 0,
+                        currentPage: page,
+                        hasMore: false
+                    };
+                }
                 params.append('base_models', pageState.filters.baseModel.join(','));
             }
 

static/js/components/ContextMenu/CheckpointContextMenu.js

@@ -26,7 +26,7 @@ export class CheckpointContextMenu extends BaseContextMenu {
         // Update the "Move to other root" label based on current model type
         const moveOtherItem = this.menu.querySelector('[data-action="move-other"]');
         if (moveOtherItem) {
-            const currentType = card.dataset.model_type || 'checkpoint';
+            const currentType = card.dataset.sub_type || 'checkpoint';
             const otherType = currentType === 'checkpoint' ? 'diffusion_model' : 'checkpoint';
             const typeLabel = i18n.t(`checkpoints.modelTypes.${otherType}`);
             moveOtherItem.innerHTML = `<i class="fas fa-exchange-alt"></i> ${i18n.t('checkpoints.contextMenu.moveToOtherTypeFolder', { otherType: typeLabel })}`;
@@ -65,11 +65,11 @@ export class CheckpointContextMenu extends BaseContextMenu {
                 apiClient.refreshSingleModelMetadata(this.currentCard.dataset.filepath);
                 break;
             case 'move':
-                moveManager.showMoveModal(this.currentCard.dataset.filepath, this.currentCard.dataset.model_type);
+                moveManager.showMoveModal(this.currentCard.dataset.filepath, this.currentCard.dataset.sub_type);
                 break;
             case 'move-other':
                 {
-                    const currentType = this.currentCard.dataset.model_type || 'checkpoint';
+                    const currentType = this.currentCard.dataset.sub_type || 'checkpoint';
                     const otherType = currentType === 'checkpoint' ? 'diffusion_model' : 'checkpoint';
                     moveManager.showMoveModal(this.currentCard.dataset.filepath, otherType);
                 }

static/js/components/RecipeCard.js

@@ -26,6 +26,7 @@ class RecipeCard {
         card.dataset.nsfwLevel = this.recipe.preview_nsfw_level || 0;
         card.dataset.created = this.recipe.created_date;
         card.dataset.id = this.recipe.id || '';
+        card.dataset.folder = this.recipe.folder || '';
 
         // Get base model with fallback
         const baseModelLabel = (this.recipe.base_model || '').trim() || 'Unknown';

static/js/components/RecipeModal.js

@@ -1075,7 +1075,7 @@ class RecipeModal {
         const checkpointName = checkpoint.name || checkpoint.modelName || checkpoint.file_name || 'Checkpoint';
         const versionLabel = checkpoint.version || checkpoint.modelVersionName || '';
         const baseModel = checkpoint.baseModel || checkpoint.base_model || '';
-        const modelTypeRaw = (checkpoint.model_type || checkpoint.type || 'checkpoint').toLowerCase();
+        const modelTypeRaw = (checkpoint.sub_type || checkpoint.type || 'checkpoint').toLowerCase();
         const modelTypeLabel = modelTypeRaw === 'diffusion_model' ? 'Diffusion Model' : 'Checkpoint';
 
         const previewMedia = isPreviewVideo ? `
@@ -1172,7 +1172,7 @@ class RecipeModal {
             return;
         }
 
-        const modelType = (checkpoint.model_type || checkpoint.type || 'checkpoint').toLowerCase();
+        const modelType = (checkpoint.sub_type || checkpoint.type || 'checkpoint').toLowerCase();
         const isDiffusionModel = modelType === 'diffusion_model' || modelType === 'unet';
         const widgetName = isDiffusionModel ? 'unet_name' : 'ckpt_name';
 

static/js/components/initialization.js

@@ -198,6 +198,12 @@ class InitializationManager {
     handleProgressUpdate(data) {
         if (!data) return;
         console.log('Received progress update:', data);
+
+        // Handle cache health warning messages
+        if (data.type === 'cache_health_warning') {
+            this.handleCacheHealthWarning(data);
+            return;
+        }
         
         // Check if this update is for our page type
         if (data.pageType && data.pageType !== this.pageType) {
@@ -466,6 +472,29 @@ class InitializationManager {
         }
     }
 
+    /**
+     * Handle cache health warning messages from WebSocket
+     */
+    handleCacheHealthWarning(data) {
+        console.log('Cache health warning received:', data);
+
+        // Import bannerService dynamically to avoid circular dependencies
+        import('../managers/BannerService.js').then(({ bannerService }) => {
+            // Initialize banner service if not already done
+            if (!bannerService.initialized) {
+                bannerService.initialize().then(() => {
+                    bannerService.registerCacheHealthBanner(data);
+                }).catch(err => {
+                    console.error('Failed to initialize banner service:', err);
+                });
+            } else {
+                bannerService.registerCacheHealthBanner(data);
+            }
+        }).catch(err => {
+            console.error('Failed to load banner service:', err);
+        });
+    }
+
     /**
      * Clean up resources when the component is destroyed
      */

static/js/components/shared/ModelCard.js

@@ -4,7 +4,7 @@ import { showModelModal } from './ModelModal.js';
 import { toggleShowcase } from './showcase/ShowcaseView.js';
 import { bulkManager } from '../../managers/BulkManager.js';
 import { modalManager } from '../../managers/ModalManager.js';
-import { NSFW_LEVELS, getBaseModelAbbreviation } from '../../utils/constants.js';
+import { NSFW_LEVELS, getBaseModelAbbreviation, getSubTypeAbbreviation, MODEL_SUBTYPE_DISPLAY_NAMES } from '../../utils/constants.js';
 import { MODEL_TYPES } from '../../api/apiConfig.js';
 import { getModelApiClient } from '../../api/modelApiFactory.js';
 import { showDeleteModal } from '../../utils/modalUtils.js';
@@ -176,7 +176,7 @@ function handleSendToWorkflow(card, replaceMode, modelType) {
             return;
         }
 
-        const subtype = (card.dataset.model_type || 'checkpoint').toLowerCase();
+        const subtype = (card.dataset.sub_type || 'checkpoint').toLowerCase();
         const isDiffusionModel = subtype === 'diffusion_model';
         const widgetName = isDiffusionModel ? 'unet_name' : 'ckpt_name';
         const actionTypeText = translate(
@@ -453,9 +453,9 @@ export function createModelCard(model, modelType) {
         card.dataset.usage_tips = model.usage_tips;
     }
 
-    // checkpoint specific data
-    if (modelType === MODEL_TYPES.CHECKPOINT) {
-        card.dataset.model_type = model.model_type; // checkpoint or diffusion_model
+    // Set sub_type for all model types (lora/locon/dora, checkpoint/diffusion_model, embedding)
+    if (model.sub_type) {
+        card.dataset.sub_type = model.sub_type;
     }
 
     // Store metadata if available
@@ -580,6 +580,11 @@ export function createModelCard(model, modelType) {
     const baseModelLabel = model.base_model || 'Unknown';
     const baseModelAbbreviation = getBaseModelAbbreviation(baseModelLabel);
 
+    // Sub-type display (e.g., LoRA, LyCO, DoRA, CKPT, DM, EMB)
+    const subType = model.sub_type || '';
+    const subTypeAbbreviation = getSubTypeAbbreviation(subType);
+    const fullSubTypeName = MODEL_SUBTYPE_DISPLAY_NAMES[subType?.toLowerCase()] || subType || '';
+
     card.innerHTML = `
         <div class="card-preview ${shouldBlur ? 'blurred' : ''}">
             ${isVideo ?
@@ -592,12 +597,15 @@ export function createModelCard(model, modelType) {
                       <i class="fas fa-eye"></i>
                   </button>` : ''}
                 <div class="card-header-info">
-                    <span class="base-model-label ${shouldBlur ? 'with-toggle' : ''}" title="${baseModelLabel}">
-                        ${baseModelAbbreviation}
+                    <span class="base-model-label ${shouldBlur ? 'with-toggle' : ''}"
+                          title="${fullSubTypeName ? fullSubTypeName + ' | ' : ''}${baseModelLabel}">
+                        ${subTypeAbbreviation ? `<span class="model-sub-type">${subTypeAbbreviation}</span>` : ''}
+                        ${subTypeAbbreviation ? `<span class="model-separator"></span>` : ''}
+                        <span class="model-base-type">${baseModelAbbreviation}</span>
                     </span>
                     ${hasUpdateAvailable ? `
                         <span class="model-update-badge" title="${updateBadgeTooltip}">
-                            ${updateBadgeLabel}
+                            <i class="fas fa-arrow-up"></i>
                         </span>
                     ` : ''}
                 </div>

static/js/components/shared/TriggerWords.js

@@ -382,6 +382,19 @@ export function setupTriggerWordsEditMode() {
                 this.value = ''; // Clear input after adding
             }
         });
+
+        // Auto-commit on blur to prevent data loss when clicking save
+        triggerWordInput.addEventListener('blur', function () {
+            if (this.value.trim()) {
+                // Small delay to avoid conflict with save button click
+                setTimeout(() => {
+                    if (document.contains(this) && this.value.trim()) {
+                        addNewTriggerWord(this.value.trim());
+                        this.value = '';
+                    }
+                }, 150);
+            }
+        });
     }
 
     // Set up save button
@@ -619,6 +632,14 @@ async function saveTriggerWords() {
     const editBtn = document.querySelector('.edit-trigger-words-btn');
     const filePath = editBtn.dataset.filePath;
     const triggerWordsSection = editBtn.closest('.trigger-words');
+
+    // Auto-commit any pending input to prevent data loss
+    const input = triggerWordsSection.querySelector('.metadata-input');
+    if (input && input.value.trim()) {
+        addNewTriggerWord(input.value.trim());
+        input.value = '';
+    }
+
     const triggerWordTags = triggerWordsSection.querySelectorAll('.trigger-word-tag');
     const words = Array.from(triggerWordTags).map(tag => tag.dataset.word);