fix(recipe): fall back to original image for metadata extraction when optimized lacks embedded data (#968 )

When CivitAI API returns meta=null and the optimized CDN image has no embedded generation parameters (e.g. PNG tEXt chunks stripped by Cloudflare Images), download the original image as fallback to recover full recipe metadata (prompt, seed, LoRAs, etc.). Also fixes Chrome password manager popping up on recipe save by adding autocomplete="new-password" to the settings API key and proxy password fields.
chore(template): update template workflow
2026-06-10 04:49:24 -03:00 · 2026-06-10 15:06:56 +08:00 · 2026-06-10 15:01:48 +08:00 · 2026-06-09 15:21:45 +08:00 · 2026-06-09 14:37:02 +08:00 · 2026-06-09 14:27:53 +08:00
215 changed files with 11558 additions and 2319 deletions
--- a/.github/ISSUE_TEMPLATE/feature_request.md
+++ b/.github/ISSUE_TEMPLATE/feature_request.md
@@ -13,8 +13,5 @@ A clear and concise description of what the problem is. Ex. I'm always frustrate
 **Describe the solution you'd like**
 A clear and concise description of what you want to happen.
 **Describe alternatives you've considered**
 A clear and concise description of any alternative solutions or features you've considered.
 **Additional context**
 Add any other context or screenshots about the feature request here.
--- a/.gitignore
+++ b/.gitignore
@@ -15,7 +15,9 @@ model_cache/
 # agent
 .opencode/
 .claude/
 .sisyphus/
 .codex
 .omo
 # Vue widgets development cache (but keep build output)
 vue-widgets/node_modules/
--- a/README.md
+++ b/README.md
--- a/data/supporters.json
+++ b/data/supporters.json
--- a/docs/LM-Extension-Wiki.md
+++ b/docs/LM-Extension-Wiki.md
@@ -1,183 +0,0 @@
 ## Overview
 The **LoRA Manager Civitai Extension** is a Browser extension designed to work seamlessly with [LoRA Manager](https://github.com/willmiao/ComfyUI-Lora-Manager) to significantly enhance your browsing experience on [Civitai](https://civitai.com). With this extension, you can:
 ✅ Instantly see which models are already present in your local library  
 ✅ Download new models with a single click  
 ✅ Manage downloads efficiently with queue and parallel download support  
 ✅ Keep your downloaded models automatically organized according to your custom settings    
 ![Civitai Models page](https://github.com/willmiao/ComfyUI-Lora-Manager/blob/main/wiki-images/civitai-models-page.png) 
 **Update:** It now also supports browsing on [CivArchive](https://civarchive.com/) (formerly CivitaiArchive). 
 ![CivArchive Models page](https://github.com/willmiao/ComfyUI-Lora-Manager/blob/main/wiki-images/civarchive-models-page.png)  
 ---
 ## Why Supporter Access?
 LoRA Manager is built with love for the Stable Diffusion and ComfyUI communities. Your support makes it possible for me to keep improving and maintaining the tool full-time.
 Supporter-exclusive features help ensure the long-term sustainability of LoRA Manager, allowing continuous updates, new features, and better performance for everyone.
 Every contribution directly fuels development and keeps the core LoRA Manager free and open-source. In addition to monthly supporters, one-time donation supporters will also receive a license key, with the duration scaling according to the contribution amount. Thank you for helping keep this project alive and growing. ❤️
 ---
 ## Installation
 ### Supported Browsers & Installation Methods
 | Browser            | Installation Method                                                                 |
 |--------------------|-------------------------------------------------------------------------------------|
 | **Google Chrome**   | [Chrome Web Store link](https://chromewebstore.google.com/detail/capigligggeijgmocnaflanlbghnamgm?utm_source=item-share-cb) |
 | **Microsoft Edge**  | Install via Chrome Web Store (compatible)                                          |
 | **Brave Browser**   | Install via Chrome Web Store (compatible)                                          |
 | **Opera**           | Install via Chrome Web Store (compatible)                                          |
 | **Firefox**         | <div id="firefox-install" class="install-ok"><a href="https://github.com/willmiao/lm-civitai-extension-firefox/releases/latest/download/extension.xpi">📦 Install Firefox Extension (reviewed and verified by Mozilla)</a></div>             |
 For non-Chrome browsers (e.g., Microsoft Edge), you can typically install extensions from the Chrome Web Store by following these steps: open the extension’s Chrome Web Store page, click 'Get extension', then click 'Allow' when prompted to enable installations from other stores, and finally click 'Add extension' to complete the installation.
 ---
 ## Privacy & Security
 I understand concerns around browser extensions and privacy, and I want to be fully transparent about how the **LM Civitai Extension** works:
 - **Reviewed and Verified**  
  This extension has been **manually reviewed and approved by the Chrome Web Store**. The Firefox version uses the **exact same code** (only the packaging format differs) and has passed **Mozilla’s Add-on review**.
 - **Minimal Network Access**  
  The only external server this extension connects to is:  
  **`https://willmiao.shop`** — used solely for **license validation**.
  It does **not collect, transmit, or store any personal or usage data**.  
  No browsing history, no user IDs, no analytics, no hidden trackers.
 - **Local-Only Model Detection**  
  Model detection and LoRA Manager communication all happen **locally** within your browser, directly interacting with your local LoRA Manager backend.
 I value your trust and are committed to keeping your local setup private and secure. If you have any questions, feel free to reach out!
 ---
 ## How to Use
 After installing the extension, you'll automatically receive a **7-day trial** to explore all features.
 When the extension is correctly installed and your license is valid:
 - Open **Civitai**, and you'll see visual indicators added by the extension on model cards, showing:
  - ✅ Models already present in your local library  
  - ⬇️ A download button for models not in your library  
 Clicking the download button adds the corresponding model version to the download queue, waiting to be downloaded. You can set up to **5 models to download simultaneously**.
 ### Visual Indicators Appear On:
 - **Home Page** — Featured models  
 - **Models Page**  
 - **Creator Profiles** — If the creator has set their models to be visible  
 - **Recommended Resources** — On individual model pages  
 ### Version Buttons on Model Pages
 On a specific model page, visual indicators also appear on version buttons, showing which versions are already in your local library.
 **Starting from v0.4.8**, model pages use a dedicated download button for better compatibility. When switching to a specific version by clicking a version button:
 - The new **dedicated download button** directly triggers download via **LoRA Manager**
 - The **original download button** remains unchanged for standard browser downloads
 ![Civitai Model Page](https://github.com/willmiao/ComfyUI-Lora-Manager/blob/main/wiki-images/civitai-model-page.png)
 ### Hide Models Already in Library (Beta)
 **New in v0.4.8**: A new **Hide models already in library (Beta)** option makes it easier to focus on models you haven't added yet. It can be enabled from Settings, or toggled quickly using **Ctrl + Shift + H** (macOS: **Command + Shift + H**).
 ### Resources on Image Pages — now shows in-library indicators for image resources plus one-click recipe import
 - **One-Click Import Civitai Image as Recipe** — Import any Civitai image as a recipe with a single click in the Resources Used panel.
 - **Auto-Queue Missing Assets** — In Settings you can decide if LoRAs or checkpoints referenced by that image should automatically be added to your download queue.
 - **More Accurate Metadata** — Importing directly from the page is faster than copying inside LM and keeps on-site tags and other metadata perfectly aligned.
 ![Civitai Image Page](https://github.com/willmiao/ComfyUI-Lora-Manager/blob/main/wiki-images/civitai-image-page.jpg)
 [![alt](url)](https://github.com/user-attachments/assets/41fd4240-c949-4f83-bde7-8f3124c09494)
 ---
 ## Model Download Location & LoRA Manager Settings
 To use the **one-click download function**, you must first set:
 - Your **Default LoRAs Root**  
 - Your **Default Checkpoints Root**  
 These are set within LoRA Manager's settings.
 When everything is configured, downloaded model files will be placed in:  
 `<Default_Models_Root>/<Base_Model_of_the_Model>/<First_Tag_of_the_Model>`
 ### Update: Default Path Customization (2025-07-21)  
 A new setting to customize the default download path has been added in the nightly version. You can now personalize where models are saved when downloading via the LM Civitai Extension.
 ![Default Path Customization](https://github.com/willmiao/ComfyUI-Lora-Manager/blob/main/wiki-images/default-path-customization.png)
 The previous YAML path mapping file will be deprecated—settings will now be unified in settings.json to simplify configuration.
 ---
 ## Backend Port Configuration
 If your **ComfyUI** or **LoRA Manager** backend is running on a port **other than the default 8188**, you must configure the backend port in the extension's settings.
 After correctly setting and saving the port, you'll see in the extension's header area:  
 - A **Healthy** status with the tooltip:  `Connected to LoRA Manager on port xxxx`
 ---
 ## Advanced Usage
 ### Connecting to a Remote LoRA Manager
 If your LoRA Manager is running on another computer, you can still connect from your browser using port forwarding.
 > **Why can't you set a remote IP directly?**
 >
 > For privacy and security, the extension only requests access to `http://127.0.0.1/*`. Supporting remote IPs would require much broader permissions, which may be rejected by browser stores and could raise user concerns.
 **Solution: Port Forwarding with `socat`**
 On your browser computer, run:
 `socat TCP-LISTEN:8188,bind=127.0.0.1,fork TCP:REMOTE.IP.ADDRESS.HERE:8188`
 - Replace `REMOTE.IP.ADDRESS.HERE` with the IP of the machine running LoRA Manager.
 - Adjust the port if needed.
 This lets the extension connect to `127.0.0.1:8188` as usual, with traffic forwarded to your remote server.
 _Thanks to user **Temikus** for sharing this solution!_
 ---
 ## Roadmap
 The extension will evolve alongside **LoRA Manager** improvements. Planned features include:
 - [x] Support for **additional model types** (e.g., embeddings)
 - [x] One-click **Recipe Import**
 - [x] Display of in-library status for all resources in the **Resources Used** section of the image page
 - [x] One-click **Auto-organize Models**
 - [x] **Hide models already in library (Beta)** - Focus on models you haven't added yet
 **Stay tuned — and thank you for your support!**
 ---
--- a/example_workflows/Lora_Manager_Basic.json
+++ b/example_workflows/Lora_Manager_Basic.json
--- a/locales/de.json
+++ b/locales/de.json
@@ -15,7 +15,10 @@
            "settings": "Einstellungen",
            "help": "Hilfe",
            "add": "Hinzufügen",
-            "close": "Schließen"
+            "close": "Schließen",
            "menu": "Menü",
            "remove": "Entfernen",
            "change": "Ändern"
        },
        "status": {
            "loading": "Wird geladen...",
@@ -231,7 +234,10 @@
            "license": "Lizenz",
            "noCreditRequired": "Kein Credit erforderlich",
            "allowSellingGeneratedContent": "Verkauf erlaubt",
            "allowSellingGeneratedContentTooltip": "Verkauf generierter Bilder erlauben",
            "noCreditRequiredTooltip": "Modell ohne Nennung des Erstellers verwenden",
            "noTags": "Keine Tags",
            "autoTags": "Auto-Tags",
            "noBaseModelMatches": "Keine Basismodelle entsprechen der aktuellen Suche.",
            "clearAll": "Alle Filter löschen",
            "any": "Beliebig",
@@ -265,10 +271,10 @@
        },
        "downloadBackend": {
            "label": "Download-Backend",
-            "help": "Wähle aus, wie Modelldateien heruntergeladen werden. Python verwendet den eingebauten Downloader. aria2 verwendet den experimentellen externen Downloader-Prozess.",
+            "help": "Wähle aus, wie Modelldateien heruntergeladen werden. Python verwendet den eingebauten Downloader. aria2 verwendet den empfohlenen externen Downloader-Prozess.",
            "options": {
                "python": "Python (integriert)",
-                "aria2": "aria2 (experimentell)"
+                "aria2": "aria2 (empfohlen)"
            }
        },
        "aria2cPath": {
@@ -575,7 +581,13 @@
        },
        "misc": {
            "includeTriggerWords": "Trigger Words in LoRA-Syntax einschließen",
-            "includeTriggerWordsHelp": "Trainierte Trigger Words beim Kopieren der LoRA-Syntax in die Zwischenablage einschließen"
+            "includeTriggerWordsHelp": "Trainierte Trigger Words beim Kopieren der LoRA-Syntax in die Zwischenablage einschließen",
            "loraSyntaxFormat": "LoRA-Syntaxformat",
            "loraSyntaxFormatHelp": "LoRA-Syntaxformat. Der vollständige Pfad enthält den Unterordnerpfad (<lora:style/anime/x:1.0>) für verlustfreie Modellauflösung. Legacy verwendet nur den Dateinamen (<lora:x:1.0>) — A1111-Konvention, kann bei doppelten Dateinamen in verschiedenen Ordnern zu Mehrdeutigkeiten führen.",
            "loraSyntaxFormatOptions": {
                "full": "Vollständiger Pfad (Unterordner/Name)",
                "legacy": "Legacy A1111 (nur Name)"
            }
        },
        "metadataArchive": {
            "enableArchiveDb": "Metadaten-Archiv-Datenbank aktivieren",
@@ -639,8 +651,6 @@
            },
            "refresh": {
                "title": "Modelliste aktualisieren",
                "quick": "Änderungen synchronisieren",
                "quickTooltip": "Nach neuen oder fehlenden Modelldateien suchen, damit die Liste aktuell bleibt.",
                "full": "Cache neu aufbauen",
                "fullTooltip": "Alle Modelldetails aus Metadatendateien neu laden – nutzen, wenn die Bibliothek veraltet wirkt oder nach manuellen Änderungen."
            },
@@ -681,16 +691,29 @@
            "setContentRating": "Inhaltsbewertung für alle festlegen",
            "copyAll": "Alle Syntax kopieren",
            "refreshAll": "Alle Metadaten aktualisieren",
            "repairMetadata": "Metadaten der Auswahl reparieren",
            "checkUpdates": "Auswahl auf Updates prüfen",
            "moveAll": "Alle in Ordner verschieben",
            "autoOrganize": "Automatisch organisieren",
            "skipMetadataRefresh": "Metadaten-Aktualisierung für ausgewählte Modelle überspringen",
            "resumeMetadataRefresh": "Metadaten-Aktualisierung für ausgewählte Modelle fortsetzen",
            "setFavorite": "Als Favorit setzen",
            "setFavoriteCount": "Als Favorit setzen ({favorited}/{total})",
            "unfavorite": "Aus Favoriten entfernen",
            "deleteAll": "Ausgewählte löschen",
            "downloadMissingLoras": "Fehlende LoRAs herunterladen",
            "downloadExamples": "Beispielbilder herunterladen",
            "clear": "Auswahl löschen",
            "skipMetadataRefreshCount": "Überspringen（{count} Modelle）",
            "resumeMetadataRefreshCount": "Fortsetzen（{count} Modelle）",
            "sendToWorkflow": "An Workflow senden",
            "sections": {
                "workflow": "Workflow",
                "metadata": "Metadaten",
                "attributes": "Attribute",
                "organize": "Organisieren",
                "download": "Download"
            },
            "autoOrganizeProgress": {
                "initializing": "Automatische Organisation wird initialisiert...",
                "starting": "Automatische Organisation für {type} wird gestartet...",
@@ -803,8 +826,6 @@
            },
            "refresh": {
                "title": "Rezeptliste aktualisieren",
                "quick": "Änderungen synchronisieren",
                "quickTooltip": "Änderungen synchronisieren - schnelle Aktualisierung ohne Cache-Neubau",
                "full": "Cache neu aufbauen",
                "fullTooltip": "Cache neu aufbauen - vollständiger Rescan aller Rezeptdateien"
            },
@@ -944,6 +965,13 @@
        "empty": {
            "noFolders": "Keine Ordner gefunden",
            "dragHint": "Elemente hierher ziehen, um Ordner zu erstellen"
        },
        "folderUpdateCheck": {
            "label": "Auf Updates in diesem Ordner prüfen",
            "loading": "Prüfe {type}-Updates in diesem Ordner...",
            "success": "{count} Update(s) für {type}s in diesem Ordner gefunden",
            "none": "Alle {type}s in diesem Ordner sind aktuell",
            "error": "Fehler beim Prüfen des Ordners auf {type}-Updates: {message}"
        }
    },
    "statistics": {
@@ -988,9 +1016,9 @@
        "download": {
            "title": "Modell von URL herunterladen",
            "titleWithType": "{type} von URL herunterladen",
            "url": "Civitai URL",
            "civitaiUrl": "Civitai URL:",
            "placeholder": "https://civitai.com/models/...",
            "urlHint": "Geben Sie eine CivitAI- oder CivArchive-URL pro Zeile ein. Unterstützt mehrere URLs für den Batch-Download.",
            "locationPreview": "Download-Speicherort Vorschau",
            "useDefaultPath": "Standardpfad verwenden",
            "useDefaultPathTooltip": "Wenn aktiviert, werden Dateien automatisch mit konfigurierten Pfadvorlagen organisiert",
@@ -1012,6 +1040,11 @@
            "downloadedTooltip": "Zuvor heruntergeladen, aber derzeit nicht in Ihrer Bibliothek.",
            "alreadyInLibrary": "Bereits in Bibliothek",
            "autoOrganizedPath": "[Automatisch organisiert durch Pfadvorlage]",
            "fileSelection": {
                "title": "Dateiformat auswählen",
                "files": "Dateien",
                "select": "Datei auswählen"
            },
            "errors": {
                "invalidUrl": "Ungültiges Civitai URL-Format",
                "noVersions": "Keine Versionen für dieses Modell verfügbar"
@@ -1076,6 +1109,12 @@
            "countMessage": "Modelle werden dauerhaft gelöscht.",
            "action": "Alle löschen"
        },
        "bulkDeleteRecipes": {
            "title": "Mehrere Rezepte löschen",
            "message": "Sind Sie sicher, dass Sie alle ausgewählten Rezepte und ihre zugehörigen Dateien löschen möchten?",
            "countMessage": "Rezepte werden dauerhaft gelöscht.",
            "action": "Alle löschen"
        },
        "checkUpdates": {
            "title": "Alle {typePlural} auf Updates prüfen?",
            "message": "Damit werden alle {typePlural} in deiner Bibliothek auf Updates geprüft. Bei großen Sammlungen kann das etwas länger dauern.",
@@ -1156,6 +1195,7 @@
                "editModelName": "Modellname bearbeiten",
                "editFileName": "Dateiname bearbeiten",
                "editBaseModel": "Basis-Modell bearbeiten",
                "editVersionName": "Versionsname bearbeiten",
                "viewOnCivitai": "Auf Civitai anzeigen",
                "viewOnCivitaiText": "Auf Civitai anzeigen",
                "viewCreatorProfile": "Ersteller-Profil anzeigen",
@@ -1187,7 +1227,9 @@
            },
            "notes": {
                "saved": "Notizen erfolgreich gespeichert",
-                "saveFailed": "Fehler beim Speichern der Notizen"
+                "saveFailed": "Fehler beim Speichern der Notizen",
                "showMore": "Mehr anzeigen",
                "showLess": "Weniger anzeigen"
            },
            "usageTips": {
                "addPresetParameter": "Voreingestellten Parameter hinzufügen...",
@@ -1292,12 +1334,15 @@
                    "earlyAccess": "Früher Zugriff",
                    "earlyAccessTooltip": "Für diese Version ist derzeit Civitai Early Access erforderlich",
                    "ignored": "Ignoriert",
-                    "ignoredTooltip": "Für diese Version sind Update-Benachrichtigungen deaktiviert"
+                    "ignoredTooltip": "Für diese Version sind Update-Benachrichtigungen deaktiviert",
                    "onSiteOnly": "Nur On-Site",
                    "onSiteOnlyTooltip": "Diese Version ist nur für die On-Site-Generierung auf Civitai verfügbar"
                },
                "actions": {
                    "download": "Herunterladen",
                    "downloadTooltip": "Diese Version herunterladen",
                    "downloadEarlyAccessTooltip": "Diese Early-Access-Version von Civitai herunterladen",
                    "downloadNotAllowedTooltip": "Diese Version ist nur für die On-Site-Generierung auf Civitai verfügbar",
                    "delete": "Löschen",
                    "deleteTooltip": "Diese lokale Version löschen",
                    "ignore": "Ignorieren",
@@ -1627,6 +1672,10 @@
            "noRecipeId": "Keine Rezept-ID verfügbar",
            "sendToWorkflowFailed": "Fehler beim Senden des Rezepts an den Workflow: {message}",
            "copyFailed": "Fehler beim Kopieren der Rezept-Syntax: {message}",
            "createError": "Fehler beim Erstellen des Rezepts：{message}",
            "createFailed": "Fehler beim Erstellen des Rezepts：{error}",
            "createMissingData": "Erforderliche Daten zum Erstellen des Rezepts fehlen",
            "created": "Rezept erfolgreich erstellt",
            "noMissingLoras": "Keine fehlenden LoRAs zum Herunterladen",
            "missingLorasInfoFailed": "Fehler beim Abrufen der Informationen für fehlende LoRAs",
            "preparingForDownloadFailed": "Fehler beim Vorbereiten der LoRAs für den Download",
@@ -1665,6 +1714,9 @@
            "batchImportBrowseFailed": "Failed to browse directory: {message}",
            "batchImportDirectorySelected": "Directory selected: {path}",
            "noRecipesSelected": "Keine Rezepte ausgewählt",
            "repairBulkComplete": "Reparatur abgeschlossen: {repaired} repariert, {skipped} übersprungen (von {total})",
            "repairBulkSkipped": "Keine Reparatur für die {total} ausgewählten Rezepte erforderlich",
            "repairBulkFailed": "Reparatur der ausgewählten Rezepte fehlgeschlagen: {message}",
            "noMissingLorasInSelection": "Keine fehlenden LoRAs in ausgewählten Rezepten gefunden",
            "noLoraRootConfigured": "Kein LoRA-Stammverzeichnis konfiguriert. Bitte legen Sie ein Standard-LoRA-Stammverzeichnis in den Einstellungen fest."
        },
@@ -1695,6 +1747,11 @@
            "bulkContentRatingSet": "Inhaltsbewertung auf {level} für {count} Modell(e) gesetzt",
            "bulkContentRatingPartial": "Inhaltsbewertung auf {level} für {success} Modell(e) gesetzt, {failed} fehlgeschlagen",
            "bulkContentRatingFailed": "Inhaltsbewertung für ausgewählte Modelle konnte nicht aktualisiert werden",
            "bulkFavoriteUpdating": "Füge {count} Modell(e) zu Favoriten hinzu...",
            "bulkUnfavoriteUpdating": "Entferne {count} Modell(e) aus Favoriten...",
            "bulkFavoritePartialAdded": "{success} Modell(e) zu Favoriten hinzugefügt, {failed} fehlgeschlagen",
            "bulkFavoritePartialRemoved": "{success} Modell(e) aus Favoriten entfernt, {failed} fehlgeschlagen",
            "bulkFavoriteFailed": "Fehler beim Aktualisieren des Favoritenstatus",
            "bulkUpdatesChecking": "Ausgewählte {type}-Modelle werden auf Updates geprüft...",
            "bulkUpdatesSuccess": "Updates für {count} ausgewählte {type}-Modelle verfügbar",
            "bulkUpdatesNone": "Keine Updates für ausgewählte {type}-Modelle gefunden",
@@ -1897,9 +1954,32 @@
            "warning": "Handlungsbedarf",
            "error": "Aktion erforderlich"
        },
        "issues": {
            "civitai_api_key": {
                "title": "Civitai API Key"
            },
            "cache_health": {
                "title": "Model Cache Health"
            },
            "filename_conflicts": {
                "title": "Duplicate Filename Conflicts"
            },
            "ui_version": {
                "title": "UI Version"
            }
        },
        "actions": {
            "runAgain": "Erneut ausführen",
-            "exportBundle": "Paket exportieren"
+            "exportBundle": "Paket exportieren",
            "open-settings": "Open Settings",
            "open-settings-syntax-format": "Switch to Full Path Syntax",
            "repair-cache": "Rebuild Cache",
            "resolve-filename-conflicts": "Resolve Conflicts",
            "reload-page": "Reload UI"
        },
        "labels": {
            "conflicts": "Conflicts",
            "version": "Version"
        },
        "toast": {
            "loadFailed": "Diagnose konnte nicht geladen werden: {message}",
@@ -1911,6 +1991,15 @@
            "conflictsResolveFailed": "Auflösung der Dateinamenskonflikte fehlgeschlagen: {message}"
        }
    },
    "conflictConfirm": {
        "title": "Dateinamenskonflikte auflösen",
        "message": "Umbenennen durch Anhängen eines 4-stelligen Hashs an jeden doppelten Dateinamen.",
        "note": "Dieser Vorgang benennt Dateien auf der Festplatte um. Modellreferenzen in vorhandenen Workflows müssen möglicherweise aktualisiert werden, wenn Sie das A1111-Syntaxformat verwenden.",
        "detail": "Beispiel: <code>filename_v1.2</code> → <code>filename_v1.2-ab3c</code>",
        "impact": "Benennt <strong>{count}</strong> Datei(en) in <strong>{groups}</strong> Duplikatgruppe(n) um",
        "confirm": "Dateien umbenennen",
        "cancel": "Abbrechen"
    },
    "banners": {
        "versionMismatch": {
            "title": "Anwendungs-Update erkannt",
--- a/locales/en.json
+++ b/locales/en.json
@@ -15,7 +15,10 @@
            "settings": "Settings",
            "help": "Help",
            "add": "Add",
-            "close": "Close"
+            "close": "Close",
            "menu": "Menu",
            "remove": "Remove",
            "change": "Change"
        },
        "status": {
            "loading": "Loading...",
@@ -231,7 +234,10 @@
            "license": "License",
            "noCreditRequired": "No Credit Required",
            "allowSellingGeneratedContent": "Allow Selling",
            "allowSellingGeneratedContentTooltip": "Allow selling generated images",
            "noCreditRequiredTooltip": "Use the model without crediting the creator",
            "noTags": "No tags",
            "autoTags": "Auto Tags",
            "noBaseModelMatches": "No base models match the current search.",
            "clearAll": "Clear All Filters",
            "any": "Any",
@@ -265,10 +271,10 @@
        },
        "downloadBackend": {
            "label": "Download backend",
-            "help": "Choose how model files are downloaded. Python uses the built-in downloader. aria2 uses the experimental external downloader process.",
+            "help": "Choose how model files are downloaded. Python uses the built-in downloader. aria2 uses the recommended external downloader process.",
            "options": {
                "python": "Python (built-in)",
-                "aria2": "aria2 (experimental)"
+                "aria2": "aria2 (recommended)"
            }
        },
        "aria2cPath": {
@@ -575,7 +581,13 @@
        },
        "misc": {
            "includeTriggerWords": "Include Trigger Words in LoRA Syntax",
-            "includeTriggerWordsHelp": "Include trained trigger words when copying LoRA syntax to clipboard"
+            "includeTriggerWordsHelp": "Include trained trigger words when copying LoRA syntax to clipboard",
            "loraSyntaxFormat": "LoRA Syntax Format",
            "loraSyntaxFormatHelp": "LoRA syntax format. Full includes subfolder path (<lora:style/anime/x:1.0>) for lossless model resolution. Legacy uses filename only (<lora:x:1.0>) — A1111 convention, may be ambiguous with duplicate filenames across folders.",
            "loraSyntaxFormatOptions": {
                "full": "Full path (subfolder/name)",
                "legacy": "Legacy A1111 (name only)"
            }
        },
        "metadataArchive": {
            "enableArchiveDb": "Enable Metadata Archive Database",
@@ -639,8 +651,6 @@
            },
            "refresh": {
                "title": "Refresh model list",
                "quick": "Sync Changes",
                "quickTooltip": "Scan for new or missing model files so the list stays current.",
                "full": "Rebuild Cache",
                "fullTooltip": "Reload all model details from metadata files—use if the library looks out of date or after manual edits."
            },
@@ -681,16 +691,29 @@
            "setContentRating": "Set Content Rating for Selected",
            "copyAll": "Copy Selected Syntax",
            "refreshAll": "Refresh Selected Metadata",
            "repairMetadata": "Repair Metadata for Selected",
            "checkUpdates": "Check Updates for Selected",
            "moveAll": "Move Selected to Folder",
            "autoOrganize": "Auto-Organize Selected",
            "skipMetadataRefresh": "Skip Metadata Refresh for Selected",
            "resumeMetadataRefresh": "Resume Metadata Refresh for Selected",
            "setFavorite": "Set as Favorite",
            "setFavoriteCount": "Set as Favorite ({favorited}/{total})",
            "unfavorite": "Remove from Favorites",
            "deleteAll": "Delete Selected",
            "downloadMissingLoras": "Download Missing LoRAs",
            "downloadExamples": "Download Example Images",
            "clear": "Clear Selection",
            "skipMetadataRefreshCount": "Skip ({count} models)",
            "resumeMetadataRefreshCount": "Resume ({count} models)",
            "sendToWorkflow": "Send to Workflow",
            "sections": {
                "workflow": "Workflow",
                "metadata": "Metadata",
                "attributes": "Attributes",
                "organize": "Organize",
                "download": "Download"
            },
            "autoOrganizeProgress": {
                "initializing": "Initializing auto-organize...",
                "starting": "Starting auto-organize for {type}...",
@@ -803,8 +826,6 @@
            },
            "refresh": {
                "title": "Refresh recipe list",
                "quick": "Sync Changes",
                "quickTooltip": "Sync changes - quick refresh without rebuilding cache",
                "full": "Rebuild Cache",
                "fullTooltip": "Rebuild cache - full rescan of all recipe files"
            },
@@ -944,6 +965,13 @@
        "empty": {
            "noFolders": "No folders found",
            "dragHint": "Drag items here to create folders"
        },
        "folderUpdateCheck": {
            "label": "Check for updates in this folder",
            "loading": "Checking {type} updates for this folder...",
            "success": "Found {count} update(s) for {type}s in this folder",
            "none": "All {type}s in this folder are up to date",
            "error": "Failed to check folder for {type} updates: {message}"
        }
    },
    "statistics": {
@@ -988,9 +1016,9 @@
        "download": {
            "title": "Download Model from URL",
            "titleWithType": "Download {type} from URL",
-            "url": "Civitai URL",
+            "civitaiUrl": "Civitai URL(s):",
            "civitaiUrl": "Civitai URL:",
            "placeholder": "https://civitai.com/models/...",
            "urlHint": "Enter one CivitAI or CivArchive URL per line. Supports multiple URLs for batch download.",
            "locationPreview": "Download Location Preview",
            "useDefaultPath": "Use Default Path",
            "useDefaultPathTooltip": "When enabled, files are automatically organized using configured path templates",
@@ -1012,6 +1040,11 @@
            "downloadedTooltip": "Previously downloaded, but it is not currently in your library.",
            "alreadyInLibrary": "Already in Library",
            "autoOrganizedPath": "[Auto-organized by path template]",
            "fileSelection": {
                "title": "Select File Format",
                "files": "files",
                "select": "Select File"
            },
            "errors": {
                "invalidUrl": "Invalid Civitai URL format",
                "noVersions": "No versions available for this model"
@@ -1076,6 +1109,12 @@
            "countMessage": "models will be permanently deleted.",
            "action": "Delete All"
        },
        "bulkDeleteRecipes": {
            "title": "Delete Multiple Recipes",
            "message": "Are you sure you want to delete all selected recipes and their associated files?",
            "countMessage": "recipes will be permanently deleted.",
            "action": "Delete All"
        },
        "checkUpdates": {
            "title": "Check updates for all {typePlural}?",
            "message": "This checks every {typePlural} in your library for updates. Large collections may take a little longer.",
@@ -1156,6 +1195,7 @@
                "editModelName": "Edit model name",
                "editFileName": "Edit file name",
                "editBaseModel": "Edit base model",
                "editVersionName": "Edit version name",
                "viewOnCivitai": "View on Civitai",
                "viewOnCivitaiText": "View on Civitai",
                "viewCreatorProfile": "View Creator Profile",
@@ -1187,7 +1227,9 @@
            },
            "notes": {
                "saved": "Notes saved successfully",
-                "saveFailed": "Failed to save notes"
+                "saveFailed": "Failed to save notes",
                "showMore": "Show more",
                "showLess": "Show less"
            },
            "usageTips": {
                "addPresetParameter": "Add preset parameter...",
@@ -1292,12 +1334,15 @@
                    "earlyAccess": "Early Access",
                    "earlyAccessTooltip": "This version currently requires Civitai early access",
                    "ignored": "Ignored",
-                    "ignoredTooltip": "Update notifications are disabled for this version"
+                    "ignoredTooltip": "Update notifications are disabled for this version",
                    "onSiteOnly": "On-Site Only",
                    "onSiteOnlyTooltip": "This version is only available for on-site generation on Civitai"
                },
                "actions": {
                    "download": "Download",
                    "downloadTooltip": "Download this version",
                    "downloadEarlyAccessTooltip": "Download this early access version from Civitai",
                    "downloadNotAllowedTooltip": "This version is only available for on-site generation on Civitai",
                    "delete": "Delete",
                    "deleteTooltip": "Delete this local version",
                    "ignore": "Ignore",
@@ -1627,6 +1672,10 @@
            "noRecipeId": "No recipe ID available",
            "sendToWorkflowFailed": "Failed to send recipe to workflow: {message}",
            "copyFailed": "Error copying recipe syntax: {message}",
            "createError": "Error creating recipe: {message}",
            "createFailed": "Failed to create recipe: {error}",
            "createMissingData": "Missing required data to create recipe",
            "created": "Recipe created successfully",
            "noMissingLoras": "No missing LoRAs to download",
            "missingLorasInfoFailed": "Failed to get information for missing LoRAs",
            "preparingForDownloadFailed": "Error preparing LoRAs for download",
@@ -1665,6 +1714,9 @@
            "batchImportBrowseFailed": "Failed to browse directory: {message}",
            "batchImportDirectorySelected": "Directory selected: {path}",
            "noRecipesSelected": "No recipes selected",
            "repairBulkComplete": "Repair complete: {repaired} repaired, {skipped} skipped (of {total})",
            "repairBulkSkipped": "No repair needed for any of the {total} selected recipes",
            "repairBulkFailed": "Failed to repair selected recipes: {message}",
            "noMissingLorasInSelection": "No missing LoRAs found in selected recipes",
            "noLoraRootConfigured": "No LoRA root directory configured. Please set a default LoRA root in settings."
        },
@@ -1695,6 +1747,11 @@
            "bulkContentRatingSet": "Set content rating to {level} for {count} model(s)",
            "bulkContentRatingPartial": "Set content rating to {level} for {success} model(s), {failed} failed",
            "bulkContentRatingFailed": "Failed to update content rating for selected models",
            "bulkFavoriteUpdating": "Adding {count} model(s) to favorites...",
            "bulkUnfavoriteUpdating": "Removing {count} model(s) from favorites...",
            "bulkFavoritePartialAdded": "Added {success} model(s) to favorites, {failed} failed",
            "bulkFavoritePartialRemoved": "Removed {success} model(s) from favorites, {failed} failed",
            "bulkFavoriteFailed": "Failed to update favorite status for selected models",
            "bulkUpdatesChecking": "Checking selected {type}(s) for updates...",
            "bulkUpdatesSuccess": "Updates available for {count} selected {type}(s)",
            "bulkUpdatesNone": "No updates found for selected {type}(s)",
@@ -1897,9 +1954,32 @@
            "warning": "Needs Attention",
            "error": "Action Required"
        },
        "issues": {
            "civitai_api_key": {
                "title": "Civitai API Key"
            },
            "cache_health": {
                "title": "Model Cache Health"
            },
            "filename_conflicts": {
                "title": "Duplicate Filename Conflicts"
            },
            "ui_version": {
                "title": "UI Version"
            }
        },
        "actions": {
            "runAgain": "Run Again",
-            "exportBundle": "Export Bundle"
+            "exportBundle": "Export Bundle",
            "open-settings": "Open Settings",
            "open-settings-syntax-format": "Switch to Full Path Syntax",
            "repair-cache": "Rebuild Cache",
            "resolve-filename-conflicts": "Resolve Conflicts",
            "reload-page": "Reload UI"
        },
        "labels": {
            "conflicts": "Conflicts",
            "version": "Version"
        },
        "toast": {
            "loadFailed": "Failed to load diagnostics: {message}",
@@ -1911,6 +1991,15 @@
            "conflictsResolveFailed": "Failed to resolve filename conflicts: {message}"
        }
    },
    "conflictConfirm": {
        "title": "Resolve Filename Conflicts",
        "message": "Renaming by appending a 4-character hash to each duplicate filename.",
        "note": "This operation renames files on disk. Model references in existing workflows may need updating if you use the A1111 syntax format.",
        "detail": "Example: <code>filename_v1.2</code> → <code>filename_v1.2-ab3c</code>",
        "impact": "Will rename <strong>{count}</strong> file(s) across <strong>{groups}</strong> duplicate group(s).",
        "confirm": "Rename Files",
        "cancel": "Cancel"
    },
    "banners": {
        "versionMismatch": {
            "title": "Application Update Detected",
@@ -1940,4 +2029,4 @@
            "retry": "Retry"
        }
    }
-}
+}
--- a/locales/es.json
+++ b/locales/es.json
@@ -15,7 +15,10 @@
            "settings": "Configuración",
            "help": "Ayuda",
            "add": "Añadir",
-            "close": "Cerrar"
+            "close": "Cerrar",
            "menu": "Menú",
            "remove": "Eliminar",
            "change": "Cambiar"
        },
        "status": {
            "loading": "Cargando...",
@@ -231,7 +234,10 @@
            "license": "Licencia",
            "noCreditRequired": "Sin crédito requerido",
            "allowSellingGeneratedContent": "Venta permitida",
            "allowSellingGeneratedContentTooltip": "Permitir la venta de imágenes generadas",
            "noCreditRequiredTooltip": "Usar el modelo sin atribuir al creador",
            "noTags": "Sin etiquetas",
            "autoTags": "Etiquetas automáticas",
            "noBaseModelMatches": "Ningún modelo base coincide con la búsqueda actual.",
            "clearAll": "Limpiar todos los filtros",
            "any": "Cualquiera",
@@ -265,10 +271,10 @@
        },
        "downloadBackend": {
            "label": "Backend de descarga",
-            "help": "Elige cómo se descargan los archivos del modelo. Python usa el descargador integrado. aria2 usa el proceso externo experimental de descarga.",
+            "help": "Elige cómo se descargan los archivos del modelo. Python usa el descargador integrado. aria2 usa el proceso externo recomendado de descarga.",
            "options": {
                "python": "Python (integrado)",
-                "aria2": "aria2 (experimental)"
+                "aria2": "aria2 (recomendado)"
            }
        },
        "aria2cPath": {
@@ -575,7 +581,13 @@
        },
        "misc": {
            "includeTriggerWords": "Incluir palabras clave en la sintaxis de LoRA",
-            "includeTriggerWordsHelp": "Incluir palabras clave entrenadas al copiar la sintaxis de LoRA al portapapeles"
+            "includeTriggerWordsHelp": "Incluir palabras clave entrenadas al copiar la sintaxis de LoRA al portapapeles",
            "loraSyntaxFormat": "Formato de sintaxis LoRA",
            "loraSyntaxFormatHelp": "Formato de sintaxis LoRA. El formato completo incluye la ruta de la subcarpeta (<lora:style/anime/x:1.0>) para una resolución de modelo sin pérdidas. El formato heredado usa solo el nombre del archivo (<lora:x:1.0>) — convención A1111, puede ser ambiguo con nombres de archivo duplicados entre carpetas.",
            "loraSyntaxFormatOptions": {
                "full": "Ruta completa (subcarpeta/nombre)",
                "legacy": "A1111 heredado (solo nombre)"
            }
        },
        "metadataArchive": {
            "enableArchiveDb": "Habilitar base de datos de archivo de metadatos",
@@ -639,8 +651,6 @@
            },
            "refresh": {
                "title": "Actualizar lista de modelos",
                "quick": "Sincronizar cambios",
                "quickTooltip": "Busca archivos de modelo nuevos o faltantes para mantener la lista al día.",
                "full": "Reconstruir caché",
                "fullTooltip": "Vuelve a cargar todos los detalles desde los archivos de metadatos; úsalo si la biblioteca parece desactualizada o tras ediciones manuales."
            },
@@ -681,16 +691,29 @@
            "setContentRating": "Establecer clasificación de contenido para todos",
            "copyAll": "Copiar toda la sintaxis",
            "refreshAll": "Actualizar todos los metadatos",
            "repairMetadata": "Reparar metadatos de la selección",
            "checkUpdates": "Comprobar actualizaciones para la selección",
            "moveAll": "Mover todos a carpeta",
            "autoOrganize": "Auto-organizar seleccionados",
            "skipMetadataRefresh": "Omitir actualización de metadatos para seleccionados",
            "resumeMetadataRefresh": "Reanudar actualización de metadatos para seleccionados",
            "setFavorite": "Marcar como favorito",
            "setFavoriteCount": "Marcar como favorito ({favorited}/{total})",
            "unfavorite": "Quitar de favoritos",
            "deleteAll": "Eliminar seleccionados",
            "downloadMissingLoras": "Descargar LoRAs faltantes",
            "downloadExamples": "Descargar imágenes de ejemplo",
            "clear": "Limpiar selección",
            "skipMetadataRefreshCount": "Omitir（{count} modelos）",
            "resumeMetadataRefreshCount": "Reanudar（{count} modelos）",
            "sendToWorkflow": "Enviar al workflow",
            "sections": {
                "workflow": "Workflow",
                "metadata": "Metadatos",
                "attributes": "Atributos",
                "organize": "Organizar",
                "download": "Descargar"
            },
            "autoOrganizeProgress": {
                "initializing": "Inicializando auto-organización...",
                "starting": "Iniciando auto-organización para {type}...",
@@ -803,8 +826,6 @@
            },
            "refresh": {
                "title": "Actualizar lista de recetas",
                "quick": "Sincronizar cambios",
                "quickTooltip": "Sincronizar cambios - actualización rápida sin reconstruir caché",
                "full": "Reconstruir caché",
                "fullTooltip": "Reconstruir caché - reescaneo completo de todos los archivos de recetas"
            },
@@ -944,6 +965,13 @@
        "empty": {
            "noFolders": "No se encontraron carpetas",
            "dragHint": "Arrastra elementos aquí para crear carpetas"
        },
        "folderUpdateCheck": {
            "label": "Buscar actualizaciones en esta carpeta",
            "loading": "Buscando actualizaciones de {type} en esta carpeta...",
            "success": "Se encontraron {count} actualización(es) para {type}s en esta carpeta",
            "none": "Todos los {type}s en esta carpeta están actualizados",
            "error": "Error al buscar actualizaciones de {type} en la carpeta: {message}"
        }
    },
    "statistics": {
@@ -988,9 +1016,9 @@
        "download": {
            "title": "Descargar modelo desde URL",
            "titleWithType": "Descargar {type} desde URL",
            "url": "URL de Civitai",
            "civitaiUrl": "URL de Civitai:",
            "placeholder": "https://civitai.com/models/...",
            "urlHint": "Ingrese una URL de CivitAI o CivArchive por línea. Admite múltiples URLs para descarga por lotes.",
            "locationPreview": "Vista previa de ubicación de descarga",
            "useDefaultPath": "Usar ruta predeterminada",
            "useDefaultPathTooltip": "Cuando está habilitado, los archivos se organizan automáticamente usando plantillas de rutas configuradas",
@@ -1012,6 +1040,11 @@
            "downloadedTooltip": "Descargado anteriormente, pero actualmente no está en tu biblioteca.",
            "alreadyInLibrary": "Ya en la biblioteca",
            "autoOrganizedPath": "[Auto-organizado por plantilla de ruta]",
            "fileSelection": {
                "title": "Seleccionar formato de archivo",
                "files": "archivos",
                "select": "Seleccionar archivo"
            },
            "errors": {
                "invalidUrl": "Formato de URL de Civitai inválido",
                "noVersions": "No hay versiones disponibles para este modelo"
@@ -1076,6 +1109,12 @@
            "countMessage": "modelos serán eliminados permanentemente.",
            "action": "Eliminar todo"
        },
        "bulkDeleteRecipes": {
            "title": "Eliminar múltiples recetas",
            "message": "¿Estás seguro de que quieres eliminar todas las recetas seleccionadas y sus archivos asociados?",
            "countMessage": "recetas serán eliminadas permanentemente.",
            "action": "Eliminar todo"
        },
        "checkUpdates": {
            "title": "¿Comprobar actualizaciones para todos los {typePlural}?",
            "message": "Esto comprobará las actualizaciones de todos los {typePlural} de tu biblioteca. En colecciones grandes puede tardar un poco más.",
@@ -1156,6 +1195,7 @@
                "editModelName": "Editar nombre del modelo",
                "editFileName": "Editar nombre de archivo",
                "editBaseModel": "Editar modelo base",
                "editVersionName": "Editar nombre de versión",
                "viewOnCivitai": "Ver en Civitai",
                "viewOnCivitaiText": "Ver en Civitai",
                "viewCreatorProfile": "Ver perfil del creador",
@@ -1187,7 +1227,9 @@
            },
            "notes": {
                "saved": "Notas guardadas exitosamente",
-                "saveFailed": "Error al guardar notas"
+                "saveFailed": "Error al guardar notas",
                "showMore": "Mostrar más",
                "showLess": "Mostrar menos"
            },
            "usageTips": {
                "addPresetParameter": "Añadir parámetro preestablecido...",
@@ -1292,12 +1334,15 @@
                    "earlyAccess": "Acceso temprano",
                    "earlyAccessTooltip": "Esta versión requiere actualmente acceso temprano de Civitai",
                    "ignored": "Ignorada",
-                    "ignoredTooltip": "Las notificaciones de actualización están desactivadas para esta versión"
+                    "ignoredTooltip": "Las notificaciones de actualización están desactivadas para esta versión",
                    "onSiteOnly": "Solo en Sitio",
                    "onSiteOnlyTooltip": "Esta versión solo está disponible para generación en el sitio de Civitai"
                },
                "actions": {
                    "download": "Descargar",
                    "downloadTooltip": "Descargar esta versión",
                    "downloadEarlyAccessTooltip": "Descargar esta versión de acceso temprano desde Civitai",
                    "downloadNotAllowedTooltip": "Esta versión solo está disponible para generación en el sitio de Civitai",
                    "delete": "Eliminar",
                    "deleteTooltip": "Eliminar esta versión local",
                    "ignore": "Ignorar",
@@ -1627,6 +1672,10 @@
            "noRecipeId": "No hay ID de receta disponible",
            "sendToWorkflowFailed": "Error al enviar la receta al flujo de trabajo: {message}",
            "copyFailed": "Error copiando sintaxis de receta: {message}",
            "createError": "Error al crear la receta：{message}",
            "createFailed": "Error al crear la receta：{error}",
            "createMissingData": "Faltan datos necesarios para crear la receta",
            "created": "Receta creada exitosamente",
            "noMissingLoras": "No hay LoRAs faltantes para descargar",
            "missingLorasInfoFailed": "Error al obtener información de LoRAs faltantes",
            "preparingForDownloadFailed": "Error preparando LoRAs para descarga",
@@ -1665,6 +1714,9 @@
            "batchImportBrowseFailed": "Failed to browse directory: {message}",
            "batchImportDirectorySelected": "Directory selected: {path}",
            "noRecipesSelected": "No se han seleccionado recetas",
            "repairBulkComplete": "Reparación completa: {repaired} reparadas, {skipped} omitidas (de {total})",
            "repairBulkSkipped": "No se necesita reparación para ninguna de las {total} recetas seleccionadas",
            "repairBulkFailed": "Error al reparar las recetas seleccionadas: {message}",
            "noMissingLorasInSelection": "No se encontraron LoRAs faltantes en las recetas seleccionadas",
            "noLoraRootConfigured": "No se ha configurado el directorio raíz de LoRA. Por favor, establezca un directorio raíz de LoRA predeterminado en la configuración."
        },
@@ -1695,6 +1747,11 @@
            "bulkContentRatingSet": "Clasificación de contenido establecida en {level} para {count} modelo(s)",
            "bulkContentRatingPartial": "Clasificación de contenido establecida en {level} para {success} modelo(s), {failed} fallaron",
            "bulkContentRatingFailed": "No se pudo actualizar la clasificación de contenido para los modelos seleccionados",
            "bulkFavoriteUpdating": "Añadiendo {count} modelo(s) a favoritos...",
            "bulkUnfavoriteUpdating": "Eliminando {count} modelo(s) de favoritos...",
            "bulkFavoritePartialAdded": "{success} modelo(s) añadido(s) a favoritos, {failed} fallido(s)",
            "bulkFavoritePartialRemoved": "{success} modelo(s) eliminado(s) de favoritos, {failed} fallido(s)",
            "bulkFavoriteFailed": "Error al actualizar el estado de favorito",
            "bulkUpdatesChecking": "Comprobando actualizaciones para {type} seleccionados...",
            "bulkUpdatesSuccess": "Actualizaciones disponibles para {count} {type} seleccionados",
            "bulkUpdatesNone": "No se encontraron actualizaciones para los {type} seleccionados",
@@ -1897,9 +1954,32 @@
            "warning": "Requiere atención",
            "error": "Se requiere acción"
        },
        "issues": {
            "civitai_api_key": {
                "title": "Civitai API Key"
            },
            "cache_health": {
                "title": "Model Cache Health"
            },
            "filename_conflicts": {
                "title": "Duplicate Filename Conflicts"
            },
            "ui_version": {
                "title": "UI Version"
            }
        },
        "actions": {
            "runAgain": "Ejecutar de nuevo",
-            "exportBundle": "Exportar paquete"
+            "exportBundle": "Exportar paquete",
            "open-settings": "Open Settings",
            "open-settings-syntax-format": "Switch to Full Path Syntax",
            "repair-cache": "Rebuild Cache",
            "resolve-filename-conflicts": "Resolve Conflicts",
            "reload-page": "Reload UI"
        },
        "labels": {
            "conflicts": "Conflicts",
            "version": "Version"
        },
        "toast": {
            "loadFailed": "Error al cargar los diagnósticos: {message}",
@@ -1911,6 +1991,15 @@
            "conflictsResolveFailed": "Error al resolver conflictos de nombre de archivo: {message}"
        }
    },
    "conflictConfirm": {
        "title": "Resolver conflictos de nombres de archivo",
        "message": "Renombrar añadiendo un hash de 4 caracteres a cada nombre de archivo duplicado.",
        "note": "Esta operación renombra archivos en el disco. Es posible que las referencias a modelos en flujos de trabajo existentes deban actualizarse si usas el formato de sintaxis A1111.",
        "detail": "Ejemplo: <code>filename_v1.2</code> → <code>filename_v1.2-ab3c</code>",
        "impact": "Renombrará <strong>{count}</strong> archivo(s) en <strong>{groups}</strong> grupo(s) de duplicados",
        "confirm": "Renombrar archivos",
        "cancel": "Cancelar"
    },
    "banners": {
        "versionMismatch": {
            "title": "Actualización de la aplicación detectada",
--- a/locales/fr.json
+++ b/locales/fr.json
@@ -15,7 +15,10 @@
            "settings": "Paramètres",
            "help": "Aide",
            "add": "Ajouter",
-            "close": "Fermer"
+            "close": "Fermer",
            "menu": "Menu",
            "remove": "Supprimer",
            "change": "Modifier"
        },
        "status": {
            "loading": "Chargement...",
@@ -231,7 +234,10 @@
            "license": "Licence",
            "noCreditRequired": "Crédit non requis",
            "allowSellingGeneratedContent": "Vente autorisée",
            "allowSellingGeneratedContentTooltip": "Autoriser la vente d\"images générées",
            "noCreditRequiredTooltip": "Utiliser le modèle sans créditer le créateur",
            "noTags": "Aucun tag",
            "autoTags": "Auto-Tags",
            "noBaseModelMatches": "Aucun modèle de base ne correspond à la recherche actuelle.",
            "clearAll": "Effacer tous les filtres",
            "any": "N'importe quel",
@@ -265,10 +271,10 @@
        },
        "downloadBackend": {
            "label": "Moteur de téléchargement",
-            "help": "Choisissez comment les fichiers de modèles sont téléchargés. Python utilise le téléchargeur intégré. aria2 utilise le processus externe expérimental de téléchargement.",
+            "help": "Choisissez comment les fichiers de modèles sont téléchargés. Python utilise le téléchargeur intégré. aria2 utilise le processus externe recommandé de téléchargement.",
            "options": {
                "python": "Python (intégré)",
-                "aria2": "aria2 (expérimental)"
+                "aria2": "aria2 (recommandé)"
            }
        },
        "aria2cPath": {
@@ -575,7 +581,13 @@
        },
        "misc": {
            "includeTriggerWords": "Inclure les mots-clés dans la syntaxe LoRA",
-            "includeTriggerWordsHelp": "Inclure les mots-clés d'entraînement lors de la copie de la syntaxe LoRA dans le presse-papiers"
+            "includeTriggerWordsHelp": "Inclure les mots-clés d'entraînement lors de la copie de la syntaxe LoRA dans le presse-papiers",
            "loraSyntaxFormat": "Format de syntaxe LoRA",
            "loraSyntaxFormatHelp": "Format de syntaxe LoRA. Le format complet inclut le chemin du sous-dossier (<lora:style/anime/x:1.0>) pour une résolution de modèle sans perte. Le format hérité utilise uniquement le nom du fichier (<lora:x:1.0>) — convention A1111, peut être ambiguë en cas de noms de fichiers en double dans différents dossiers.",
            "loraSyntaxFormatOptions": {
                "full": "Chemin complet (sous-dossier/nom)",
                "legacy": "A1111 hérité (nom uniquement)"
            }
        },
        "metadataArchive": {
            "enableArchiveDb": "Activer la base de données d'archive des métadonnées",
@@ -639,8 +651,6 @@
            },
            "refresh": {
                "title": "Actualiser la liste des modèles",
                "quick": "Synchroniser les changements",
                "quickTooltip": "Analyse les nouveaux fichiers de modèle ou les fichiers manquants pour garder la liste à jour.",
                "full": "Reconstruire le cache",
                "fullTooltip": "Recharge tous les détails des modèles depuis les fichiers metadata — à utiliser si la bibliothèque paraît obsolète ou après des modifications manuelles."
            },
@@ -681,16 +691,29 @@
            "setContentRating": "Définir la classification du contenu pour tous",
            "copyAll": "Copier toute la syntaxe",
            "refreshAll": "Actualiser toutes les métadonnées",
            "repairMetadata": "Réparer les métadonnées de la sélection",
            "checkUpdates": "Vérifier les mises à jour pour la sélection",
            "moveAll": "Déplacer tout vers un dossier",
            "autoOrganize": "Auto-organiser la sélection",
            "skipMetadataRefresh": "Ignorer l'actualisation des métadonnées pour la sélection",
            "resumeMetadataRefresh": "Reprendre l'actualisation des métadonnées pour la sélection",
            "setFavorite": "Définir comme favori",
            "setFavoriteCount": "Définir comme favori ({favorited}/{total})",
            "unfavorite": "Retirer des favoris",
            "deleteAll": "Supprimer la sélection",
            "downloadMissingLoras": "Télécharger les LoRAs manquants",
            "downloadExamples": "Télécharger les images d'exemple",
            "clear": "Effacer la sélection",
            "skipMetadataRefreshCount": "Ignorer（{count} modèles）",
            "resumeMetadataRefreshCount": "Reprendre（{count} modèles）",
            "sendToWorkflow": "Envoyer au workflow",
            "sections": {
                "workflow": "Workflow",
                "metadata": "Métadonnées",
                "attributes": "Attributs",
                "organize": "Organiser",
                "download": "Télécharger"
            },
            "autoOrganizeProgress": {
                "initializing": "Initialisation de l'auto-organisation...",
                "starting": "Démarrage de l'auto-organisation pour {type}...",
@@ -803,8 +826,6 @@
            },
            "refresh": {
                "title": "Actualiser la liste des recipes",
                "quick": "Synchroniser les changements",
                "quickTooltip": "Synchroniser les changements - actualisation rapide sans reconstruire le cache",
                "full": "Reconstruire le cache",
                "fullTooltip": "Reconstruire le cache - rescan complet de tous les fichiers de recipes"
            },
@@ -944,6 +965,13 @@
        "empty": {
            "noFolders": "Aucun dossier trouvé",
            "dragHint": "Faites glisser des éléments ici pour créer des dossiers"
        },
        "folderUpdateCheck": {
            "label": "Vérifier les mises à jour dans ce dossier",
            "loading": "Vérification des mises à jour {type} dans ce dossier...",
            "success": "{count} mise(s) à jour trouvée(s) pour les {type}s dans ce dossier",
            "none": "Tous les {type}s dans ce dossier sont à jour",
            "error": "Échec de la vérification des mises à jour {type} dans ce dossier : {message}"
        }
    },
    "statistics": {
@@ -988,9 +1016,9 @@
        "download": {
            "title": "Télécharger un modèle depuis une URL",
            "titleWithType": "Télécharger {type} depuis une URL",
            "url": "URL Civitai",
            "civitaiUrl": "URL Civitai :",
            "placeholder": "https://civitai.com/models/...",
            "urlHint": "Entrez une URL CivitAI ou CivArchive par ligne. Prend en charge plusieurs URLs pour le téléchargement par lot.",
            "locationPreview": "Aperçu de l'emplacement de téléchargement",
            "useDefaultPath": "Utiliser le chemin par défaut",
            "useDefaultPathTooltip": "Lorsque activé, les fichiers sont automatiquement organisés selon les modèles de chemin configurés",
@@ -1012,6 +1040,11 @@
            "downloadedTooltip": "Déjà téléchargé, mais il n'est actuellement pas dans votre bibliothèque.",
            "alreadyInLibrary": "Déjà dans la bibliothèque",
            "autoOrganizedPath": "[Auto-organisé par modèle de chemin]",
            "fileSelection": {
                "title": "Choisir le format de fichier",
                "files": "fichiers",
                "select": "Choisir le fichier"
            },
            "errors": {
                "invalidUrl": "Format d'URL Civitai invalide",
                "noVersions": "Aucune version disponible pour ce modèle"
@@ -1076,6 +1109,12 @@
            "countMessage": "modèles seront définitivement supprimés.",
            "action": "Tout supprimer"
        },
        "bulkDeleteRecipes": {
            "title": "Supprimer plusieurs recipes",
            "message": "Êtes-vous sûr de vouloir supprimer toutes les recipes sélectionnées et leurs fichiers associés ?",
            "countMessage": "recipes seront définitivement supprimées.",
            "action": "Tout supprimer"
        },
        "checkUpdates": {
            "title": "Vérifier les mises à jour pour tous les {typePlural} ?",
            "message": "Cette action vérifie les mises à jour pour tous les {typePlural} de votre bibliothèque. Les grandes collections peuvent prendre un peu plus de temps.",
@@ -1156,6 +1195,7 @@
                "editModelName": "Modifier le nom du modèle",
                "editFileName": "Modifier le nom de fichier",
                "editBaseModel": "Modifier le modèle de base",
                "editVersionName": "Modifier le nom de la version",
                "viewOnCivitai": "Voir sur Civitai",
                "viewOnCivitaiText": "Voir sur Civitai",
                "viewCreatorProfile": "Voir le profil du créateur",
@@ -1187,7 +1227,9 @@
            },
            "notes": {
                "saved": "Notes sauvegardées avec succès",
-                "saveFailed": "Échec de la sauvegarde des notes"
+                "saveFailed": "Échec de la sauvegarde des notes",
                "showMore": "Afficher plus",
                "showLess": "Afficher moins"
            },
            "usageTips": {
                "addPresetParameter": "Ajouter un paramètre prédéfini...",
@@ -1292,12 +1334,15 @@
                    "earlyAccess": "Accès anticipé",
                    "earlyAccessTooltip": "Cette version nécessite actuellement l'accès anticipé Civitai",
                    "ignored": "Ignorée",
-                    "ignoredTooltip": "Les notifications de mise à jour sont désactivées pour cette version"
+                    "ignoredTooltip": "Les notifications de mise à jour sont désactivées pour cette version",
                    "onSiteOnly": "Uniquement sur Site",
                    "onSiteOnlyTooltip": "Cette version n'est disponible que pour la génération sur le site Civitai"
                },
                "actions": {
                    "download": "Télécharger",
                    "downloadTooltip": "Télécharger cette version",
                    "downloadEarlyAccessTooltip": "Télécharger cette version en accès anticipé depuis Civitai",
                    "downloadNotAllowedTooltip": "Cette version n'est disponible que pour la génération sur le site Civitai",
                    "delete": "Supprimer",
                    "deleteTooltip": "Supprimer cette version locale",
                    "ignore": "Ignorer",
@@ -1627,6 +1672,10 @@
            "noRecipeId": "Aucun ID de recipe disponible",
            "sendToWorkflowFailed": "Échec de l'envoi de la recette vers le workflow : {message}",
            "copyFailed": "Erreur lors de la copie de la syntaxe de la recipe : {message}",
            "createError": "Erreur lors de la création du Recipe ：{message}",
            "createFailed": "Échec de la création du Recipe ：{error}",
            "createMissingData": "Données requises manquantes pour créer le Recipe",
            "created": "Recipe créé avec succès",
            "noMissingLoras": "Aucun LoRA manquant à télécharger",
            "missingLorasInfoFailed": "Échec de l'obtention des informations pour les LoRAs manquants",
            "preparingForDownloadFailed": "Erreur lors de la préparation des LoRAs pour le téléchargement",
@@ -1665,6 +1714,9 @@
            "batchImportBrowseFailed": "Failed to browse directory: {message}",
            "batchImportDirectorySelected": "Directory selected: {path}",
            "noRecipesSelected": "Aucune recette sélectionnée",
            "repairBulkComplete": "Réparation terminée : {repaired} réparée(s), {skipped} ignorée(s) (sur {total})",
            "repairBulkSkipped": "Aucune réparation nécessaire parmi les {total} recettes sélectionnées",
            "repairBulkFailed": "Échec de la réparation des recettes sélectionnées : {message}",
            "noMissingLorasInSelection": "Aucun LoRA manquant trouvé dans les recettes sélectionnées",
            "noLoraRootConfigured": "Aucun répertoire racine LoRA configuré. Veuillez définir un répertoire racine LoRA par défaut dans les paramètres."
        },
@@ -1695,6 +1747,11 @@
            "bulkContentRatingSet": "Classification du contenu définie sur {level} pour {count} modèle(s)",
            "bulkContentRatingPartial": "Classification du contenu définie sur {level} pour {success} modèle(s), {failed} échec(s)",
            "bulkContentRatingFailed": "Impossible de mettre à jour la classification du contenu pour les modèles sélectionnés",
            "bulkFavoriteUpdating": "Ajout de {count} modèle(s) aux favoris...",
            "bulkUnfavoriteUpdating": "Suppression de {count} modèle(s) des favoris...",
            "bulkFavoritePartialAdded": "{success} modèle(s) ajouté(s) aux favoris, {failed} échec(s)",
            "bulkFavoritePartialRemoved": "{success} modèle(s) retiré(s) des favoris, {failed} échec(s)",
            "bulkFavoriteFailed": "Échec de la mise à jour du statut de favori",
            "bulkUpdatesChecking": "Vérification des mises à jour pour les {type} sélectionnés...",
            "bulkUpdatesSuccess": "Mises à jour disponibles pour {count} {type} sélectionnés",
            "bulkUpdatesNone": "Aucune mise à jour trouvée pour les {type} sélectionnés",
@@ -1897,9 +1954,32 @@
            "warning": "Nécessite une attention",
            "error": "Action requise"
        },
        "issues": {
            "civitai_api_key": {
                "title": "Civitai API Key"
            },
            "cache_health": {
                "title": "Model Cache Health"
            },
            "filename_conflicts": {
                "title": "Duplicate Filename Conflicts"
            },
            "ui_version": {
                "title": "UI Version"
            }
        },
        "actions": {
            "runAgain": "Relancer",
-            "exportBundle": "Exporter le lot"
+            "exportBundle": "Exporter le lot",
            "open-settings": "Open Settings",
            "open-settings-syntax-format": "Switch to Full Path Syntax",
            "repair-cache": "Rebuild Cache",
            "resolve-filename-conflicts": "Resolve Conflicts",
            "reload-page": "Reload UI"
        },
        "labels": {
            "conflicts": "Conflicts",
            "version": "Version"
        },
        "toast": {
            "loadFailed": "Échec du chargement des diagnostics : {message}",
@@ -1911,6 +1991,15 @@
            "conflictsResolveFailed": "Échec de la résolution des conflits de nom de fichier : {message}"
        }
    },
    "conflictConfirm": {
        "title": "Résoudre les conflits de noms de fichiers",
        "message": "Renommer en ajoutant un hachage de 4 caractères à chaque nom de fichier en double.",
        "note": "Cette opération renomme les fichiers sur le disque. Les références de modèle dans les workflows existants peuvent nécessiter une mise à jour si vous utilisez le format de syntaxe A1111.",
        "detail": "Exemple : <code>filename_v1.2</code> → <code>filename_v1.2-ab3c</code>",
        "impact": "Renommera <strong>{count}</strong> fichier(s) dans <strong>{groups}</strong> groupe(s) de doublons",
        "confirm": "Renommer les fichiers",
        "cancel": "Annuler"
    },
    "banners": {
        "versionMismatch": {
            "title": "Mise à jour de l'application détectée",
--- a/locales/he.json
+++ b/locales/he.json
@@ -15,7 +15,10 @@
            "settings": "הגדרות",
            "help": "עזרה",
            "add": "הוספה",
-            "close": "סגור"
+            "close": "סגור",
            "menu": "תפריט",
            "remove": "הסר",
            "change": "שנה"
        },
        "status": {
            "loading": "טוען...",
@@ -231,7 +234,10 @@
            "license": "רישיון",
            "noCreditRequired": "ללא קרדיט נדרש",
            "allowSellingGeneratedContent": "אפשר מכירה",
            "allowSellingGeneratedContentTooltip": "אפשר מכירת תמונות שנוצרו",
            "noCreditRequiredTooltip": "שימוש במודל ללא מתן קרדיט ליוצר",
            "noTags": "ללא תגיות",
            "autoTags": "תגיות אוטומטיות",
            "noBaseModelMatches": "אין מודלי בסיס התואמים לחיפוש הנוכחי.",
            "clearAll": "נקה את כל המסננים",
            "any": "כלשהו",
@@ -265,10 +271,10 @@
        },
        "downloadBackend": {
            "label": "מנגנון הורדה",
-            "help": "בחר כיצד יורדים קבצי המודל. Python משתמש במוריד המובנה. aria2 משתמש בתהליך הורדה חיצוני ניסיוני.",
+            "help": "בחר כיצד יורדים קבצי המודל. Python משתמש במוריד המובנה. aria2 משתמש בתהליך הורדה חיצוני מומלץ.",
            "options": {
                "python": "Python (מובנה)",
-                "aria2": "aria2 (ניסיוני)"
+                "aria2": "aria2 (מומלץ)"
            }
        },
        "aria2cPath": {
@@ -575,7 +581,13 @@
        },
        "misc": {
            "includeTriggerWords": "כלול מילות טריגר בתחביר LoRA",
-            "includeTriggerWordsHelp": "כלול מילות טריגר מאומנות בעת העתקת תחביר LoRA ללוח"
+            "includeTriggerWordsHelp": "כלול מילות טריגר מאומנות בעת העתקת תחביר LoRA ללוח",
            "loraSyntaxFormat": "פורמט תחביר LoRA",
            "loraSyntaxFormatHelp": "פורמט תחביר LoRA. נתיב מלא כולל תת-תיקייה (<lora:style/anime/x:1.0>) לפתרון מודל ללא אובדן. גרסה ישנה משתמשת בשם קובץ בלבד (<lora:x:1.0>) — מוסכמת A1111, עלולה להיות לא חד משמעית עם שמות קבצים כפולים בתיקיות שונות.",
            "loraSyntaxFormatOptions": {
                "full": "נתיב מלא (תת-תיקייה/שם)",
                "legacy": "A1111 ישן (שם בלבד)"
            }
        },
        "metadataArchive": {
            "enableArchiveDb": "הפעל מסד נתונים של ארכיון מטא-דאטה",
@@ -639,8 +651,6 @@
            },
            "refresh": {
                "title": "רענן רשימת מודלים",
                "quick": "סנכרון שינויים",
                "quickTooltip": "סריקה לאיתור קבצי מודל חדשים או חסרים כדי לשמור את הרשימה מעודכנת.",
                "full": "בניית מטמון מחדש",
                "fullTooltip": "טוען מחדש את כל פרטי המודלים מקבצי המטא-דאטה – לשימוש אם הספרייה נראית לא מעודכנת או לאחר עריכות ידניות."
            },
@@ -681,16 +691,29 @@
            "setContentRating": "הגדר דירוג תוכן לכל המודלים",
            "copyAll": "העתק את כל התחבירים",
            "refreshAll": "רענן את כל המטא-דאטה",
            "repairMetadata": "תקן מטא-דאטה עבור הנבחרים",
            "checkUpdates": "בדוק עדכונים לבחירה",
            "moveAll": "העבר הכל לתיקייה",
            "autoOrganize": "ארגן אוטומטית נבחרים",
            "skipMetadataRefresh": "דילוג על רענון מטא-נתונים לנבחרים",
            "resumeMetadataRefresh": "המשך רענון מטא-נתונים לנבחרים",
            "setFavorite": "הגדר כמועדף",
            "setFavoriteCount": "הגדר כמועדף ({favorited}/{total})",
            "unfavorite": "הסר ממועדפים",
            "deleteAll": "מחק נבחרים",
            "downloadMissingLoras": "הורדת LoRAs חסרים",
            "downloadExamples": "הורד תמונות דוגמה",
            "clear": "נקה בחירה",
            "skipMetadataRefreshCount": "דילוג（{count} מודלים）",
            "resumeMetadataRefreshCount": "המשך（{count} מודלים）",
            "sendToWorkflow": "שלח ל-Workflow",
            "sections": {
                "workflow": "Workflow",
                "metadata": "מטא-נתונים",
                "attributes": "מאפיינים",
                "organize": "ארגן",
                "download": "הורדה"
            },
            "autoOrganizeProgress": {
                "initializing": "מאתחל ארגון אוטומטי...",
                "starting": "מתחיל ארגון אוטומטי עבור {type}...",
@@ -803,8 +826,6 @@
            },
            "refresh": {
                "title": "רענן רשימת מתכונים",
                "quick": "סנכרן שינויים",
                "quickTooltip": "סנכרן שינויים - רענון מהיר ללא בניית מטמון מחדש",
                "full": "בנה מטמון מחדש",
                "fullTooltip": "בנה מטמון מחדש - סריקה מחדש מלאה של כל קבצי המתכונים"
            },
@@ -944,6 +965,13 @@
        "empty": {
            "noFolders": "לא נמצאו תיקיות",
            "dragHint": "גרור פריטים לכאן כדי ליצור תיקיות"
        },
        "folderUpdateCheck": {
            "label": "בדוק עדכונים בתיקייה זו",
            "loading": "בודק עדכוני {type} בתיקייה זו...",
            "success": "נמצאו {count} עדכון/ים עבור {type}s בתיקייה זו",
            "none": "כל ה-{type}s בתיקייה זו מעודכנים",
            "error": "נכשל בבדיקת עדכוני {type} בתיקייה: {message}"
        }
    },
    "statistics": {
@@ -988,9 +1016,9 @@
        "download": {
            "title": "הורד מודל מכתובת URL",
            "titleWithType": "הורד {type} מכתובת URL",
            "url": "כתובת URL של Civitai",
            "civitaiUrl": "כתובת URL של Civitai:",
            "placeholder": "https://civitai.com/models/...",
            "urlHint": "יש להזין כתובת URL אחת של CivitAI או CivArchive בכל שורה. תומך במספר כתובות URL להורדה בבת אחת.",
            "locationPreview": "תצוגה מקדימה של מיקום ההורדה",
            "useDefaultPath": "השתמש בנתיב ברירת מחדל",
            "useDefaultPathTooltip": "כאשר מופעל, קבצים מאורגנים אוטומטית באמצעות תבניות נתיב מוגדרות",
@@ -1012,6 +1040,11 @@
            "downloadedTooltip": "הורד בעבר, אך הוא אינו נמצא כרגע בספרייה שלך.",
            "alreadyInLibrary": "כבר בספרייה",
            "autoOrganizedPath": "[מאורגן אוטומטית לפי תבנית נתיב]",
            "fileSelection": {
                "title": "בחר פורמט קובץ",
                "files": "קבצים",
                "select": "בחר קובץ"
            },
            "errors": {
                "invalidUrl": "פורמט URL של Civitai לא חוקי",
                "noVersions": "אין גרסאות זמינות למודל זה"
@@ -1076,6 +1109,12 @@
            "countMessage": "מודלים יימחקו לצמיתות.",
            "action": "מחק הכל"
        },
        "bulkDeleteRecipes": {
            "title": "מחק מספר מתכונים",
            "message": "האם אתה בטוח שברצונך למחוק את כל המתכונים שנבחרו ואת הקבצים הנלווים אליהם?",
            "countMessage": "מתכונים יימחקו לצמיתות.",
            "action": "מחק הכל"
        },
        "checkUpdates": {
            "title": "לבדוק עדכונים לכל ה-{typePlural}?",
            "message": "הפעולה תבדוק עדכונים עבור כל ה-{typePlural} בספרייה שלך. באוספים גדולים זה עלול לקחת מעט יותר זמן.",
@@ -1156,6 +1195,7 @@
                "editModelName": "ערוך שם מודל",
                "editFileName": "ערוך שם קובץ",
                "editBaseModel": "ערוך מודל בסיס",
                "editVersionName": "ערוך שם גרסה",
                "viewOnCivitai": "הצג ב-Civitai",
                "viewOnCivitaiText": "הצג ב-Civitai",
                "viewCreatorProfile": "הצג פרופיל יוצר",
@@ -1187,7 +1227,9 @@
            },
            "notes": {
                "saved": "הערות נשמרו בהצלחה",
-                "saveFailed": "שמירת ההערות נכשלה"
+                "saveFailed": "שמירת ההערות נכשלה",
                "showMore": "הצג עוד",
                "showLess": "הצג פחות"
            },
            "usageTips": {
                "addPresetParameter": "הוסף פרמטר קבוע מראש...",
@@ -1292,12 +1334,15 @@
                    "earlyAccess": "גישה מוקדמת",
                    "earlyAccessTooltip": "גרסה זו דורשת כרגע גישת Early Access של Civitai",
                    "ignored": "התעלם",
-                    "ignoredTooltip": "התראות העדכון מושבתות עבור גרסה זו"
+                    "ignoredTooltip": "התראות העדכון מושבתות עבור גרסה זו",
                    "onSiteOnly": "רק באתר",
                    "onSiteOnlyTooltip": "גרסה זו זמינה רק ליצירה באתר Civitai"
                },
                "actions": {
                    "download": "הורדה",
                    "downloadTooltip": "הורד את הגרסה הזו",
                    "downloadEarlyAccessTooltip": "הורד את גרסת ה-Early Access הזו מ-Civitai",
                    "downloadNotAllowedTooltip": "גרסה זו זמינה רק ליצירה באתר Civitai",
                    "delete": "מחיקה",
                    "deleteTooltip": "מחק את הגרסה המקומית הזו",
                    "ignore": "התעלם",
@@ -1627,6 +1672,10 @@
            "noRecipeId": "אין מזהה מתכון זמין",
            "sendToWorkflowFailed": "נכשל שליחת המתכון ל-workflow: {message}",
            "copyFailed": "שגיאה בהעתקת תחביר המתכון: {message}",
            "createError": "שגיאה ביצירת המתכון：{message}",
            "createFailed": "יצירת המתכון נכשלה：{error}",
            "createMissingData": "חסרים נתונים נדרשים ליצירת המתכון",
            "created": "המתכון נוצר בהצלחה",
            "noMissingLoras": "אין LoRAs חסרים להורדה",
            "missingLorasInfoFailed": "קבלת מידע עבור LoRAs חסרים נכשלה",
            "preparingForDownloadFailed": "שגיאה בהכנת LoRAs להורדה",
@@ -1665,6 +1714,9 @@
            "batchImportBrowseFailed": "Failed to browse directory: {message}",
            "batchImportDirectorySelected": "Directory selected: {path}",
            "noRecipesSelected": "לא נבחרו מתכונים",
            "repairBulkComplete": "התיקון הושלם: {repaired} תוקנו, {skipped} דולגו (מתוך {total})",
            "repairBulkSkipped": "אין צורך בתיקון עבור {total} המתכונים הנבחרים",
            "repairBulkFailed": "תיקון המתכונים הנבחרים נכשל: {message}",
            "noMissingLorasInSelection": "לא נמצאו LoRAs חסרים במתכונים שנבחרו",
            "noLoraRootConfigured": "תיקיית השורש של LoRA לא מוגדרת. אנא הגדר תיקיית שורש LoRA ברירת מחדל בהגדרות."
        },
@@ -1695,6 +1747,11 @@
            "bulkContentRatingSet": "דירוג התוכן הוגדר ל-{level} עבור {count} מודלים",
            "bulkContentRatingPartial": "דירוג התוכן הוגדר ל-{level} עבור {success} מודלים, {failed} נכשלו",
            "bulkContentRatingFailed": "עדכון דירוג התוכן עבור המודלים שנבחרו נכשל",
            "bulkFavoriteUpdating": "מוסיף {count} דגמים למועדפים...",
            "bulkUnfavoriteUpdating": "מסיר {count} דגמים ממועדפים...",
            "bulkFavoritePartialAdded": "{success} דגמים נוספו למועדפים, {failed} נכשלו",
            "bulkFavoritePartialRemoved": "{success} דגמים הוסרו ממועדפים, {failed} נכשלו",
            "bulkFavoriteFailed": "עדכון סטטוס מועדפים נכשל",
            "bulkUpdatesChecking": "בודק עדכונים עבור {type} שנבחרו...",
            "bulkUpdatesSuccess": "יש עדכונים עבור {count} {type} שנבחרו",
            "bulkUpdatesNone": "לא נמצאו עדכונים עבור {type} שנבחרו",
@@ -1897,9 +1954,32 @@
            "warning": "דורש תשומת לב",
            "error": "נדרשת פעולה"
        },
        "issues": {
            "civitai_api_key": {
                "title": "Civitai API Key"
            },
            "cache_health": {
                "title": "Model Cache Health"
            },
            "filename_conflicts": {
                "title": "Duplicate Filename Conflicts"
            },
            "ui_version": {
                "title": "UI Version"
            }
        },
        "actions": {
            "runAgain": "הפעל שוב",
-            "exportBundle": "ייצוא חבילה"
+            "exportBundle": "ייצוא חבילה",
            "open-settings": "Open Settings",
            "open-settings-syntax-format": "Switch to Full Path Syntax",
            "repair-cache": "Rebuild Cache",
            "resolve-filename-conflicts": "Resolve Conflicts",
            "reload-page": "Reload UI"
        },
        "labels": {
            "conflicts": "Conflicts",
            "version": "Version"
        },
        "toast": {
            "loadFailed": "טעינת האבחון נכשלה: {message}",
@@ -1911,6 +1991,15 @@
            "conflictsResolveFailed": "פתרון התנגשויות שמות קבצים נכשל: {message}"
        }
    },
    "conflictConfirm": {
        "title": "פתור התנגשויות בשמות קבצים",
        "message": "שינוי שם על ידי הוספת האש באורך 4 תווים לכל שם קובץ כפול.",
        "note": "פעולה זו משנה שמות של קבצים בדיסק. ייתכן שיהיה צורך לעדכן הפניות למודלים בזרימות עבודה קיימות אם אתה משתמש בפורמט התחביר A1111.",
        "detail": "דוגמה: <code>filename_v1.2</code> → <code>filename_v1.2-ab3c</code>",
        "impact": "ישנה שם של <strong>{count}</strong> קבצים ב-<strong>{groups}</strong> קבוצות כפולות",
        "confirm": "שנה שמות קבצים",
        "cancel": "ביטול"
    },
    "banners": {
        "versionMismatch": {
            "title": "זוהה עדכון יישום",
--- a/locales/ja.json
+++ b/locales/ja.json
@@ -15,7 +15,10 @@
            "settings": "設定",
            "help": "ヘルプ",
            "add": "追加",
-            "close": "閉じる"
+            "close": "閉じる",
            "menu": "メニュー",
            "remove": "削除",
            "change": "変更"
        },
        "status": {
            "loading": "読み込み中...",
@@ -231,7 +234,10 @@
            "license": "ライセンス",
            "noCreditRequired": "クレジット不要",
            "allowSellingGeneratedContent": "販売許可",
            "allowSellingGeneratedContentTooltip": "生成した画像の販売を許可",
            "noCreditRequiredTooltip": "クレジット表記なしでモデルを使用可能",
            "noTags": "タグなし",
            "autoTags": "自動タグ",
            "noBaseModelMatches": "現在の検索に一致するベースモデルはありません。",
            "clearAll": "すべてのフィルタをクリア",
            "any": "いずれか",
@@ -265,10 +271,10 @@
        },
        "downloadBackend": {
            "label": "ダウンロードバックエンド",
-            "help": "モデルファイルのダウンロード方法を選択します。Python は内蔵ダウンローダーを使用し、aria2 は実験的な外部ダウンローダープロセスを使用します。",
+            "help": "モデルファイルのダウンロード方法を選択します。Python は内蔵ダウンローダーを使用し、aria2 は推奨の外部ダウンローダープロセスを使用します。",
            "options": {
                "python": "Python（内蔵）",
-                "aria2": "aria2（実験的）"
+                "aria2": "aria2（推奨）"
            }
        },
        "aria2cPath": {
@@ -575,7 +581,13 @@
        },
        "misc": {
            "includeTriggerWords": "LoRA構文にトリガーワードを含める",
-            "includeTriggerWordsHelp": "LoRA構文をクリップボードにコピーする際、学習済みトリガーワードを含めます"
+            "includeTriggerWordsHelp": "LoRA構文をクリップボードにコピーする際、学習済みトリガーワードを含めます",
            "loraSyntaxFormat": "LoRA構文形式",
            "loraSyntaxFormatHelp": "LoRA構文形式。フルパスはサブフォルダパスを含み（<lora:style/anime/x:1.0>）、モデルをロスレスで解決します。レガシーはファイル名のみ（<lora:x:1.0>）— A1111規約ですが、フォルダ間でファイル名が重複する場合に曖昧になる可能性があります。",
            "loraSyntaxFormatOptions": {
                "full": "フルパス（サブフォルダ/名前）",
                "legacy": "レガシーA1111（名前のみ）"
            }
        },
        "metadataArchive": {
            "enableArchiveDb": "メタデータアーカイブデータベースを有効化",
@@ -639,8 +651,6 @@
            },
            "refresh": {
                "title": "モデルリストを更新",
                "quick": "変更を同期",
                "quickTooltip": "新しいモデルファイルや欠けているファイルをスキャンして一覧を最新に保ちます。",
                "full": "キャッシュを再構築",
                "fullTooltip": "メタデータファイルから全モデル情報を再読み込みします。リストが古いと感じるときや手動編集後に使用してください。"
            },
@@ -681,16 +691,29 @@
            "setContentRating": "すべてのモデルのコンテンツレーティングを設定",
            "copyAll": "すべての構文をコピー",
            "refreshAll": "すべてのメタデータを更新",
            "repairMetadata": "選択したレシピのメタデータを修復",
            "checkUpdates": "選択項目の更新を確認",
            "moveAll": "すべてをフォルダに移動",
            "autoOrganize": "自動整理を実行",
            "skipMetadataRefresh": "選択したモデルのメタデータ更新をスキップ",
            "resumeMetadataRefresh": "選択したモデルのメタデータ更新を再開",
            "setFavorite": "お気に入りに設定",
            "setFavoriteCount": "お気に入りに設定 ({favorited}/{total})",
            "unfavorite": "お気に入りから削除",
            "deleteAll": "選択したものを削除",
            "downloadMissingLoras": "不足している LoRA をダウンロード",
            "downloadExamples": "例画像をダウンロード",
            "clear": "選択をクリア",
            "skipMetadataRefreshCount": "スキップ（{count}モデル）",
            "resumeMetadataRefreshCount": "再開（{count}モデル）",
            "sendToWorkflow": "ワークフローに送信",
            "sections": {
                "workflow": "ワークフロー",
                "metadata": "メタデータ",
                "attributes": "属性",
                "organize": "整理",
                "download": "ダウンロード"
            },
            "autoOrganizeProgress": {
                "initializing": "自動整理を初期化中...",
                "starting": "{type}の自動整理を開始中...",
@@ -803,8 +826,6 @@
            },
            "refresh": {
                "title": "レシピリストを更新",
                "quick": "変更を同期",
                "quickTooltip": "変更を同期 - キャッシュを再構築せずにクイック更新",
                "full": "キャッシュを再構築",
                "fullTooltip": "キャッシュを再構築 - すべてのレシピファイルを完全に再スキャン"
            },
@@ -944,6 +965,13 @@
        "empty": {
            "noFolders": "フォルダが見つかりません",
            "dragHint": "ここへアイテムをドラッグしてフォルダを作成します"
        },
        "folderUpdateCheck": {
            "label": "このフォルダのアップデートを確認",
            "loading": "このフォルダの{type}アップデートを確認中...",
            "success": "このフォルダの{type}sに{count}件のアップデートが見つかりました",
            "none": "このフォルダのすべての{type}sは最新です",
            "error": "フォルダの{type}アップデート確認に失敗しました: {message}"
        }
    },
    "statistics": {
@@ -988,9 +1016,9 @@
        "download": {
            "title": "URLからモデルをダウンロード",
            "titleWithType": "URLから{type}をダウンロード",
            "url": "Civitai URL",
            "civitaiUrl": "Civitai URL：",
            "placeholder": "https://civitai.com/models/...",
            "urlHint": "1行に1つのCivitAIまたはCivArchive URLを入力してください。複数のURLを一括ダウンロードできます。",
            "locationPreview": "ダウンロード場所プレビュー",
            "useDefaultPath": "デフォルトパスを使用",
            "useDefaultPathTooltip": "有効にすると、設定されたパステンプレートを使用してファイルが自動的に整理されます",
@@ -1012,6 +1040,11 @@
            "downloadedTooltip": "以前にダウンロード済みですが、現在はライブラリにありません。",
            "alreadyInLibrary": "既にライブラリ内",
            "autoOrganizedPath": "[パステンプレートによる自動整理]",
            "fileSelection": {
                "title": "ファイル形式を選択",
                "files": "ファイル",
                "select": "ファイルを選択"
            },
            "errors": {
                "invalidUrl": "無効なCivitai URL形式",
                "noVersions": "このモデルの利用可能なバージョンがありません"
@@ -1076,6 +1109,12 @@
            "countMessage": "モデルが完全に削除されます。",
            "action": "すべて削除"
        },
        "bulkDeleteRecipes": {
            "title": "複数のレシピを削除",
            "message": "選択したすべてのレシピと関連ファイルを削除してもよろしいですか？",
            "countMessage": "レシピが完全に削除されます。",
            "action": "すべて削除"
        },
        "checkUpdates": {
            "title": "すべての{type}の更新を確認しますか？",
            "message": "ライブラリ内のすべての{type}で更新を確認します。コレクションが大きい場合は時間がかかることがあります。",
@@ -1156,6 +1195,7 @@
                "editModelName": "モデル名を編集",
                "editFileName": "ファイル名を編集",
                "editBaseModel": "ベースモデルを編集",
                "editVersionName": "バージョン名を編集",
                "viewOnCivitai": "Civitaiで表示",
                "viewOnCivitaiText": "Civitaiで表示",
                "viewCreatorProfile": "作成者プロフィールを表示",
@@ -1187,7 +1227,9 @@
            },
            "notes": {
                "saved": "メモが正常に保存されました",
-                "saveFailed": "メモの保存に失敗しました"
+                "saveFailed": "メモの保存に失敗しました",
                "showMore": "もっと見る",
                "showLess": "折りたたむ"
            },
            "usageTips": {
                "addPresetParameter": "プリセットパラメータを追加...",
@@ -1292,12 +1334,15 @@
                    "earlyAccess": "早期アクセス",
                    "earlyAccessTooltip": "このバージョンは現在 Civitai の早期アクセスが必要です",
                    "ignored": "無視中",
-                    "ignoredTooltip": "このバージョンの更新通知は無効です"
+                    "ignoredTooltip": "このバージョンの更新通知は無効です",
                    "onSiteOnly": "サイト内のみ",
                    "onSiteOnlyTooltip": "このバージョンはCivitaiサイト内でのみ利用可能で、ダウンロードはできません"
                },
                "actions": {
                    "download": "ダウンロード",
                    "downloadTooltip": "このバージョンをダウンロード",
                    "downloadEarlyAccessTooltip": "Civitai からこの早期アクセス版をダウンロード",
                    "downloadNotAllowedTooltip": "このバージョンはCivitaiサイト内でのみ利用可能で、ダウンロードはできません",
                    "delete": "削除",
                    "deleteTooltip": "このローカルバージョンを削除",
                    "ignore": "無視",
@@ -1627,6 +1672,10 @@
            "noRecipeId": "レシピIDが利用できません",
            "sendToWorkflowFailed": "ワークフローへのレシピ送信に失敗しました：{message}",
            "copyFailed": "レシピ構文のコピーエラー：{message}",
            "createError": "レシピ作成中にエラーが発生しました：{message}",
            "createFailed": "レシピの作成に失敗しました：{error}",
            "createMissingData": "レシピ作成に必要なデータが不足しています",
            "created": "レシピを作成しました",
            "noMissingLoras": "ダウンロードする不足LoRAがありません",
            "missingLorasInfoFailed": "不足LoRAの情報取得に失敗しました",
            "preparingForDownloadFailed": "ダウンロード用LoRAの準備中にエラーが発生しました",
@@ -1665,6 +1714,9 @@
            "batchImportBrowseFailed": "Failed to browse directory: {message}",
            "batchImportDirectorySelected": "Directory selected: {path}",
            "noRecipesSelected": "レシピが選択されていません",
            "repairBulkComplete": "修復完了：{repaired} 件修復、{skipped} 件スキップ（合計 {total} 件）",
            "repairBulkSkipped": "選択した {total} 件のレシピは修復不要です",
            "repairBulkFailed": "選択したレシピの修復に失敗しました：{message}",
            "noMissingLorasInSelection": "選択したレシピに不足している LoRA が見つかりませんでした",
            "noLoraRootConfigured": "LoRA ルートディレクトリが設定されていません。設定でデフォルトの LoRA ルートを設定してください。"
        },
@@ -1695,6 +1747,11 @@
            "bulkContentRatingSet": "{count} 件のモデルのコンテンツレーティングを {level} に設定しました",
            "bulkContentRatingPartial": "{success} 件のモデルのコンテンツレーティングを {level} に設定、{failed} 件は失敗しました",
            "bulkContentRatingFailed": "選択したモデルのコンテンツレーティングを更新できませんでした",
            "bulkFavoriteUpdating": "{count} 個のモデルをお気に入りに追加中...",
            "bulkUnfavoriteUpdating": "{count} 個のモデルをお気に入りから削除中...",
            "bulkFavoritePartialAdded": "{success} 個のモデルをお気に入りに追加、{failed} 個失敗",
            "bulkFavoritePartialRemoved": "{success} 個のモデルをお気に入りから削除、{failed} 個失敗",
            "bulkFavoriteFailed": "お気に入り状態の更新に失敗しました",
            "bulkUpdatesChecking": "選択された{type}の更新を確認しています...",
            "bulkUpdatesSuccess": "{count} 件の選択された{type}に利用可能な更新があります",
            "bulkUpdatesNone": "選択された{type}には更新が見つかりませんでした",
@@ -1897,9 +1954,32 @@
            "warning": "要注意",
            "error": "対応が必要"
        },
        "issues": {
            "civitai_api_key": {
                "title": "Civitai API キー"
            },
            "cache_health": {
                "title": "モデルキャッシュの健全性"
            },
            "filename_conflicts": {
                "title": "ファイル名重複競合"
            },
            "ui_version": {
                "title": "UI バージョン"
            }
        },
        "actions": {
            "runAgain": "再実行",
-            "exportBundle": "パッケージをエクスポート"
+            "exportBundle": "パッケージをエクスポート",
            "open-settings": "設定を開く",
            "open-settings-syntax-format": "フルパス構文に切り替え",
            "repair-cache": "キャッシュを再構築",
            "resolve-filename-conflicts": "競合を解決",
            "reload-page": "UI をリロード"
        },
        "labels": {
            "conflicts": "競合",
            "version": "バージョン"
        },
        "toast": {
            "loadFailed": "診断の読み込みに失敗しました: {message}",
@@ -1911,6 +1991,15 @@
            "conflictsResolveFailed": "ファイル名競合の解決に失敗しました: {message}"
        }
    },
    "conflictConfirm": {
        "title": "ファイル名の競合を解決",
        "message": "重複したファイル名に4文字のハッシュを追加してリネームします。",
        "note": "この操作はディスク上のファイルをリネームします。A1111 構文形式を使用している場合、既存のワークフロー内のモデル参照を更新する必要があるかもしれません。",
        "detail": "例：<code>filename_v1.2</code> → <code>filename_v1.2-ab3c</code>",
        "impact": "<strong>{groups}</strong> 組の重複にわたって <strong>{count}</strong> 個のファイルをリネームします",
        "confirm": "ファイルをリネーム",
        "cancel": "キャンセル"
    },
    "banners": {
        "versionMismatch": {
            "title": "アプリケーション更新が検出されました",
--- a/locales/ko.json
+++ b/locales/ko.json
@@ -15,7 +15,10 @@
            "settings": "설정",
            "help": "도움말",
            "add": "추가",
-            "close": "닫기"
+            "close": "닫기",
            "menu": "메뉴",
            "remove": "제거",
            "change": "변경"
        },
        "status": {
            "loading": "로딩 중...",
@@ -231,7 +234,10 @@
            "license": "라이선스",
            "noCreditRequired": "크레딧 표기 없음",
            "allowSellingGeneratedContent": "판매 허용",
            "allowSellingGeneratedContentTooltip": "생성된 이미지 판매 허용",
            "noCreditRequiredTooltip": "크리에이터 저작자 표시 없이 모델 사용 가능",
            "noTags": "태그 없음",
            "autoTags": "자동 태그",
            "noBaseModelMatches": "현재 검색과 일치하는 베이스 모델이 없습니다.",
            "clearAll": "모든 필터 지우기",
            "any": "아무",
@@ -265,10 +271,10 @@
        },
        "downloadBackend": {
            "label": "다운로드 백엔드",
-            "help": "모델 파일을 다운로드하는 방식을 선택합니다. Python은 내장 다운로더를 사용하고, aria2는 실험적인 외부 다운로더 프로세스를 사용합니다.",
+            "help": "모델 파일을 다운로드하는 방식을 선택합니다. Python은 내장 다운로더를 사용하고, aria2는 권장되는 외부 다운로더 프로세스를 사용합니다.",
            "options": {
                "python": "Python(내장)",
-                "aria2": "aria2(실험적)"
+                "aria2": "aria2(권장)"
            }
        },
        "aria2cPath": {
@@ -575,7 +581,13 @@
        },
        "misc": {
            "includeTriggerWords": "LoRA 문법에 트리거 단어 포함",
-            "includeTriggerWordsHelp": "LoRA 문법을 클립보드에 복사할 때 학습된 트리거 단어를 포함합니다"
+            "includeTriggerWordsHelp": "LoRA 문법을 클립보드에 복사할 때 학습된 트리거 단어를 포함합니다",
            "loraSyntaxFormat": "LoRA 구문 형식",
            "loraSyntaxFormatHelp": "LoRA 구문 형식. 전체 경로는 하위 폴더 경로(<lora:style/anime/x:1.0>)를 포함하여 손실 없는 모델 해상도를 제공합니다. 레거시는 파일 이름만(<lora:x:1.0>) 사용 — A1111 규칙이지만, 폴더 간 파일명 중복 시 모호할 수 있습니다.",
            "loraSyntaxFormatOptions": {
                "full": "전체 경로(하위 폴더/이름)",
                "legacy": "레거시 A1111(이름만)"
            }
        },
        "metadataArchive": {
            "enableArchiveDb": "메타데이터 아카이브 데이터베이스 활성화",
@@ -639,8 +651,6 @@
            },
            "refresh": {
                "title": "모델 목록 새로고침",
                "quick": "변경 사항 동기화",
                "quickTooltip": "새로운 모델 파일이나 누락된 파일을 찾아 목록을 최신 상태로 유지합니다.",
                "full": "캐시 재구성",
                "fullTooltip": "메타데이터 파일에서 모든 모델 정보를 다시 불러옵니다. 라이브러리가 오래되어 보이거나 수동 수정 후에 사용하세요."
            },
@@ -681,16 +691,29 @@
            "setContentRating": "모든 모델에 콘텐츠 등급 설정",
            "copyAll": "모든 문법 복사",
            "refreshAll": "모든 메타데이터 새로고침",
            "repairMetadata": "선택한 레시피 메타데이터 복구",
            "checkUpdates": "선택 항목 업데이트 확인",
            "moveAll": "모두 폴더로 이동",
            "autoOrganize": "자동 정리 선택",
            "skipMetadataRefresh": "선택한 모델의 메타데이터 새로고침 건너뛰기",
            "resumeMetadataRefresh": "선택한 모델의 메타데이터 새로고침 재개",
            "setFavorite": "즐겨찾기로 설정",
            "setFavoriteCount": "즐겨찾기로 설정 ({favorited}/{total})",
            "unfavorite": "즐겨찾기 해제",
            "deleteAll": "선택된 항목 삭제",
            "downloadMissingLoras": "누락된 LoRA 다운로드",
            "downloadExamples": "예시 이미지 다운로드",
            "clear": "선택 지우기",
            "skipMetadataRefreshCount": "건너뛰기（{count}개 모델）",
            "resumeMetadataRefreshCount": "재개（{count}개 모델）",
            "sendToWorkflow": "워크플로우로 보내기",
            "sections": {
                "workflow": "워크플로우",
                "metadata": "메타데이터",
                "attributes": "속성",
                "organize": "정리",
                "download": "다운로드"
            },
            "autoOrganizeProgress": {
                "initializing": "자동 정리 초기화 중...",
                "starting": "{type}에 대한 자동 정리 시작...",
@@ -803,8 +826,6 @@
            },
            "refresh": {
                "title": "레시피 목록 새로고침",
                "quick": "변경 사항 동기화",
                "quickTooltip": "변경 사항 동기화 - 캐시를 재구성하지 않고 빠른 새로고침",
                "full": "캐시 재구성",
                "fullTooltip": "캐시 재구성 - 모든 레시피 파일을 완전히 다시 스캔"
            },
@@ -944,6 +965,13 @@
        "empty": {
            "noFolders": "폴더를 찾을 수 없습니다",
            "dragHint": "항목을 여기로 드래그하여 폴더를 만듭니다"
        },
        "folderUpdateCheck": {
            "label": "이 폴더의 업데이트 확인",
            "loading": "이 폴더의 {type} 업데이트를 확인하는 중...",
            "success": "이 폴더에서 {type}s에 대한 {count}개 업데이트를 찾았습니다",
            "none": "이 폴더의 모든 {type}s가 최신 상태입니다",
            "error": "폴더의 {type} 업데이트 확인 실패: {message}"
        }
    },
    "statistics": {
@@ -988,9 +1016,9 @@
        "download": {
            "title": "URL에서 모델 다운로드",
            "titleWithType": "URL에서 {type} 다운로드",
            "url": "Civitai URL",
            "civitaiUrl": "Civitai URL:",
            "placeholder": "https://civitai.com/models/...",
            "urlHint": "한 줄에 하나의 CivitAI 또는 CivArchive URL을 입력하세요. 여러 URL을 일괄 다운로드할 수 있습니다.",
            "locationPreview": "다운로드 위치 미리보기",
            "useDefaultPath": "기본 경로 사용",
            "useDefaultPathTooltip": "활성화하면 구성된 경로 템플릿을 사용하여 파일이 자동으로 정리됩니다",
@@ -1012,6 +1040,11 @@
            "downloadedTooltip": "이전에 다운로드했지만 현재 라이브러리에 없습니다.",
            "alreadyInLibrary": "이미 라이브러리에 있음",
            "autoOrganizedPath": "[경로 템플릿으로 자동 정리됨]",
            "fileSelection": {
                "title": "파일 형식 선택",
                "files": "개 파일",
                "select": "파일 선택"
            },
            "errors": {
                "invalidUrl": "잘못된 Civitai URL 형식",
                "noVersions": "이 모델에 사용 가능한 버전이 없습니다"
@@ -1076,6 +1109,12 @@
            "countMessage": "개의 모델이 영구적으로 삭제됩니다.",
            "action": "모두 삭제"
        },
        "bulkDeleteRecipes": {
            "title": "여러 레시피 삭제",
            "message": "선택된 모든 레시피와 관련 파일을 삭제하시겠습니까?",
            "countMessage": "개의 레시피가 영구적으로 삭제됩니다.",
            "action": "모두 삭제"
        },
        "checkUpdates": {
            "title": "{type} 전체 업데이트를 확인할까요?",
            "message": "라이브러리에 있는 모든 {type}의 업데이트를 확인합니다. 컬렉션이 클수록 시간이 조금 더 걸릴 수 있습니다.",
@@ -1156,6 +1195,7 @@
                "editModelName": "모델명 편집",
                "editFileName": "파일명 편집",
                "editBaseModel": "베이스 모델 편집",
                "editVersionName": "버전명 편집",
                "viewOnCivitai": "Civitai에서 보기",
                "viewOnCivitaiText": "Civitai에서 보기",
                "viewCreatorProfile": "제작자 프로필 보기",
@@ -1187,7 +1227,9 @@
            },
            "notes": {
                "saved": "메모가 성공적으로 저장됨",
-                "saveFailed": "메모 저장 실패"
+                "saveFailed": "메모 저장 실패",
                "showMore": "더 보기",
                "showLess": "접기"
            },
            "usageTips": {
                "addPresetParameter": "프리셋 매개변수 추가...",
@@ -1292,12 +1334,15 @@
                    "earlyAccess": "얼리 액세스",
                    "earlyAccessTooltip": "이 버전은 현재 Civitai 얼리 액세스가 필요합니다",
                    "ignored": "무시됨",
-                    "ignoredTooltip": "이 버전은 업데이트 알림이 비활성화되어 있습니다"
+                    "ignoredTooltip": "이 버전은 업데이트 알림이 비활성화되어 있습니다",
                    "onSiteOnly": "사이트 내 전용",
                    "onSiteOnlyTooltip": "이 버전은 Civitai 사이트 내에서만 사용 가능하며 다운로드할 수 없습니다"
                },
                "actions": {
                    "download": "다운로드",
                    "downloadTooltip": "이 버전 다운로드",
                    "downloadEarlyAccessTooltip": "Civitai에서 이 얼리 액세스 버전 다운로드",
                    "downloadNotAllowedTooltip": "이 버전은 Civitai 사이트 내에서만 사용 가능하며 다운로드할 수 없습니다",
                    "delete": "삭제",
                    "deleteTooltip": "이 로컬 버전 삭제",
                    "ignore": "무시",
@@ -1627,6 +1672,10 @@
            "noRecipeId": "사용 가능한 레시피 ID가 없습니다",
            "sendToWorkflowFailed": "워크플로우에 레시피 보내기 실패: {message}",
            "copyFailed": "레시피 문법 복사 오류: {message}",
            "createError": "레시피 생성 중 오류 발생：{message}",
            "createFailed": "레시피 생성 실패：{error}",
            "createMissingData": "레시피 생성에 필요한 데이터가 없습니다",
            "created": "레시피가 생성되었습니다",
            "noMissingLoras": "다운로드할 누락된 LoRA가 없습니다",
            "missingLorasInfoFailed": "누락된 LoRA 정보를 가져오는데 실패했습니다",
            "preparingForDownloadFailed": "LoRA 다운로드 준비 오류",
@@ -1665,6 +1714,9 @@
            "batchImportBrowseFailed": "Failed to browse directory: {message}",
            "batchImportDirectorySelected": "Directory selected: {path}",
            "noRecipesSelected": "선택한 레시피가 없습니다",
            "repairBulkComplete": "복구 완료: {repaired}개 복구, {skipped}개 건너뜀 (총 {total}개)",
            "repairBulkSkipped": "선택한 {total}개 레시피는 복구가 필요하지 않습니다",
            "repairBulkFailed": "선택한 레시피 복구 실패: {message}",
            "noMissingLorasInSelection": "선택한 레시피에서 누락된 LoRA를 찾을 수 없습니다",
            "noLoraRootConfigured": "LoRA 루트 디렉토리가 구성되지 않았습니다. 설정에서 기본 LoRA 루트를 설정하세요."
        },
@@ -1695,6 +1747,11 @@
            "bulkContentRatingSet": "{count}개 모델의 콘텐츠 등급을 {level}(으)로 설정했습니다",
            "bulkContentRatingPartial": "{success}개 모델의 콘텐츠 등급을 {level}(으)로 설정했고, {failed}개는 실패했습니다",
            "bulkContentRatingFailed": "선택한 모델의 콘텐츠 등급을 업데이트하지 못했습니다",
            "bulkFavoriteUpdating": "{count}개 모델을 즐겨찾기에 추가 중...",
            "bulkUnfavoriteUpdating": "{count}개 모델을 즐겨찾기에서 제거 중...",
            "bulkFavoritePartialAdded": "{success}개 모델을 즐겨찾기에 추가, {failed}개 실패",
            "bulkFavoritePartialRemoved": "{success}개 모델을 즐겨찾기에서 제거, {failed}개 실패",
            "bulkFavoriteFailed": "즐겨찾기 상태 업데이트 실패",
            "bulkUpdatesChecking": "선택한 {type}의 업데이트를 확인하는 중...",
            "bulkUpdatesSuccess": "선택한 {count}개의 {type}에 사용할 수 있는 업데이트가 있습니다",
            "bulkUpdatesNone": "선택한 {type}에 대한 업데이트가 없습니다",
@@ -1897,9 +1954,32 @@
            "warning": "주의 필요",
            "error": "조치 필요"
        },
        "issues": {
            "civitai_api_key": {
                "title": "Civitai API 키"
            },
            "cache_health": {
                "title": "모델 캐시 상태"
            },
            "filename_conflicts": {
                "title": "파일명 중복 충돌"
            },
            "ui_version": {
                "title": "UI 버전"
            }
        },
        "actions": {
            "runAgain": "다시 실행",
-            "exportBundle": "번들 내보내기"
+            "exportBundle": "번들 내보내기",
            "open-settings": "설정 열기",
            "open-settings-syntax-format": "전체 경로 구문으로 전환",
            "repair-cache": "캐시 재구축",
            "resolve-filename-conflicts": "충돌 해결",
            "reload-page": "UI 새로고침"
        },
        "labels": {
            "conflicts": "충돌",
            "version": "버전"
        },
        "toast": {
            "loadFailed": "진단 로드 실패: {message}",
@@ -1911,6 +1991,15 @@
            "conflictsResolveFailed": "파일명 충돌 해결 실패: {message}"
        }
    },
    "conflictConfirm": {
        "title": "파일명 충돌 해결",
        "message": "중복 파일명에 4자리 해시를 추가하여 이름을 변경합니다.",
        "note": "이 작업은 디스크에 있는 파일의 이름을 변경합니다. A1111 구문 형식을 사용하는 경우 기존 워크플로우의 모델 참조를 업데이트해야 할 수 있습니다.",
        "detail": "예시: <code>filename_v1.2</code> → <code>filename_v1.2-ab3c</code>",
        "impact": "<strong>{groups}</strong>개 중복 그룹에서 <strong>{count}</strong>개 파일 이름을 변경합니다",
        "confirm": "파일 이름 변경",
        "cancel": "취소"
    },
    "banners": {
        "versionMismatch": {
            "title": "애플리케이션 업데이트 감지",
--- a/locales/ru.json
+++ b/locales/ru.json
@@ -15,7 +15,10 @@
            "settings": "Настройки",
            "help": "Справка",
            "add": "Добавить",
-            "close": "Закрыть"
+            "close": "Закрыть",
            "menu": "Меню",
            "remove": "Удалить",
            "change": "Изменить"
        },
        "status": {
            "loading": "Загрузка...",
@@ -231,7 +234,10 @@
            "license": "Лицензия",
            "noCreditRequired": "Без указания авторства",
            "allowSellingGeneratedContent": "Продажа разрешена",
            "allowSellingGeneratedContentTooltip": "Разрешить продажу сгенерированных изображений",
            "noCreditRequiredTooltip": "Использование модели без указания автора",
            "noTags": "Без тегов",
            "autoTags": "Авто-теги",
            "noBaseModelMatches": "Нет базовых моделей, соответствующих текущему поиску.",
            "clearAll": "Очистить все фильтры",
            "any": "Любой",
@@ -265,10 +271,10 @@
        },
        "downloadBackend": {
            "label": "Бэкенд загрузки",
-            "help": "Выберите способ загрузки файлов моделей. Python использует встроенный загрузчик. aria2 использует экспериментальный внешний процесс загрузки.",
+            "help": "Выберите способ загрузки файлов моделей. Python использует встроенный загрузчик. aria2 использует рекомендуемый внешний процесс загрузки.",
            "options": {
                "python": "Python (встроенный)",
-                "aria2": "aria2 (экспериментальный)"
+                "aria2": "aria2 (рекомендуемый)"
            }
        },
        "aria2cPath": {
@@ -575,7 +581,13 @@
        },
        "misc": {
            "includeTriggerWords": "Включать триггерные слова в синтаксис LoRA",
-            "includeTriggerWordsHelp": "Включать обученные триггерные слова при копировании синтаксиса LoRA в буфер обмена"
+            "includeTriggerWordsHelp": "Включать обученные триггерные слова при копировании синтаксиса LoRA в буфер обмена",
            "loraSyntaxFormat": "Формат синтаксиса LoRA",
            "loraSyntaxFormatHelp": "Формат синтаксиса LoRA. Полный путь включает подпапку (<lora:style/anime/x:1.0>) для безпотерьного разрешения модели. Устаревший использует только имя файла (<lora:x:1.0>) — соглашение A1111, может быть неоднозначным при дублировании имён файлов в разных папках.",
            "loraSyntaxFormatOptions": {
                "full": "Полный путь (подпапка/имя)",
                "legacy": "Устаревший A1111 (только имя)"
            }
        },
        "metadataArchive": {
            "enableArchiveDb": "Включить архив метаданных",
@@ -639,8 +651,6 @@
            },
            "refresh": {
                "title": "Обновить список моделей",
                "quick": "Синхронизировать изменения",
                "quickTooltip": "Находит новые или отсутствующие файлы моделей, чтобы список оставался актуальным.",
                "full": "Перестроить кэш",
                "fullTooltip": "Перечитывает все данные моделей из файлов метаданных — используйте, если библиотека выглядит устаревшей или после ручных правок."
            },
@@ -681,16 +691,29 @@
            "setContentRating": "Установить рейтинг контента для всех",
            "copyAll": "Копировать весь синтаксис",
            "refreshAll": "Обновить все метаданные",
            "repairMetadata": "Восстановить метаданные для выбранных",
            "checkUpdates": "Проверить обновления для выбранных",
            "moveAll": "Переместить все в папку",
            "autoOrganize": "Автоматически организовать выбранные",
            "skipMetadataRefresh": "Пропустить обновление метаданных для выбранных",
            "resumeMetadataRefresh": "Возобновить обновление метаданных для выбранных",
            "setFavorite": "Добавить в избранное",
            "setFavoriteCount": "Добавить в избранное ({favorited}/{total})",
            "unfavorite": "Удалить из избранного",
            "deleteAll": "Удалить выбранные",
            "downloadMissingLoras": "Скачать отсутствующие LoRAs",
            "downloadExamples": "Загрузить примеры изображений",
            "clear": "Очистить выбор",
            "skipMetadataRefreshCount": "Пропустить（{count} моделей）",
            "resumeMetadataRefreshCount": "Возобновить（{count} моделей）",
            "sendToWorkflow": "Отправить в Workflow",
            "sections": {
                "workflow": "Workflow",
                "metadata": "Метаданные",
                "attributes": "Атрибуты",
                "organize": "Организовать",
                "download": "Скачать"
            },
            "autoOrganizeProgress": {
                "initializing": "Инициализация автоматической организации...",
                "starting": "Запуск автоматической организации для {type}...",
@@ -803,8 +826,6 @@
            },
            "refresh": {
                "title": "Обновить список рецептов",
                "quick": "Синхронизировать изменения",
                "quickTooltip": "Синхронизировать изменения - быстрое обновление без перестроения кэша",
                "full": "Перестроить кэш",
                "fullTooltip": "Перестроить кэш - полное повторное сканирование всех файлов рецептов"
            },
@@ -944,6 +965,13 @@
        "empty": {
            "noFolders": "Папки не найдены",
            "dragHint": "Перетащите элементы сюда, чтобы создать папки"
        },
        "folderUpdateCheck": {
            "label": "Проверить обновления в этой папке",
            "loading": "Проверка обновлений {type} в этой папке...",
            "success": "Найдено {count} обновление(й) для {type}s в этой папке",
            "none": "Все {type}s в этой папке актуальны",
            "error": "Не удалось проверить папку на наличие обновлений {type}: {message}"
        }
    },
    "statistics": {
@@ -988,9 +1016,9 @@
        "download": {
            "title": "Скачать модель по URL",
            "titleWithType": "Скачать {type} по URL",
            "url": "Civitai URL",
            "civitaiUrl": "Civitai URL:",
            "placeholder": "https://civitai.com/models/...",
            "urlHint": "Введите один URL CivitAI или CivArchive в каждой строке. Поддерживается пакетная загрузка нескольких URL.",
            "locationPreview": "Предпросмотр места загрузки",
            "useDefaultPath": "Использовать путь по умолчанию",
            "useDefaultPathTooltip": "При включении файлы автоматически организуются с использованием настроенных шаблонов путей",
@@ -1012,6 +1040,11 @@
            "downloadedTooltip": "Ранее загружено, но сейчас этого нет в вашей библиотеке.",
            "alreadyInLibrary": "Уже в библиотеке",
            "autoOrganizedPath": "[Автоматически организовано по шаблону пути]",
            "fileSelection": {
                "title": "Выбрать формат файла",
                "files": "файлов",
                "select": "Выбрать файл"
            },
            "errors": {
                "invalidUrl": "Неверный формат URL Civitai",
                "noVersions": "Нет доступных версий для этой модели"
@@ -1076,6 +1109,12 @@
            "countMessage": "моделей будут удалены навсегда.",
            "action": "Удалить все"
        },
        "bulkDeleteRecipes": {
            "title": "Удалить несколько рецептов",
            "message": "Вы уверены, что хотите удалить все выбранные рецепты и связанные с ними файлы?",
            "countMessage": "рецептов будут удалены навсегда.",
            "action": "Удалить все"
        },
        "checkUpdates": {
            "title": "Проверить обновления для всех {typePlural}?",
            "message": "Будут проверены обновления для всех {typePlural} в вашей библиотеке. Для больших коллекций это может занять немного больше времени.",
@@ -1156,6 +1195,7 @@
                "editModelName": "Редактировать название модели",
                "editFileName": "Редактировать имя файла",
                "editBaseModel": "Редактировать базовую модель",
                "editVersionName": "Редактировать название версии",
                "viewOnCivitai": "Посмотреть на Civitai",
                "viewOnCivitaiText": "Посмотреть на Civitai",
                "viewCreatorProfile": "Посмотреть профиль создателя",
@@ -1187,7 +1227,9 @@
            },
            "notes": {
                "saved": "Заметки успешно сохранены",
-                "saveFailed": "Не удалось сохранить заметки"
+                "saveFailed": "Не удалось сохранить заметки",
                "showMore": "Показать больше",
                "showLess": "Свернуть"
            },
            "usageTips": {
                "addPresetParameter": "Добавить предустановленный параметр...",
@@ -1292,12 +1334,15 @@
                    "earlyAccess": "Ранний доступ",
                    "earlyAccessTooltip": "Для этой версии сейчас требуется ранний доступ Civitai",
                    "ignored": "Игнорируется",
-                    "ignoredTooltip": "Уведомления об обновлениях для этой версии отключены"
+                    "ignoredTooltip": "Уведомления об обновлениях для этой версии отключены",
                    "onSiteOnly": "Только на Сайте",
                    "onSiteOnlyTooltip": "Эта версия доступна только для генерации на сайте Civitai"
                },
                "actions": {
                    "download": "Скачать",
                    "downloadTooltip": "Скачать эту версию",
                    "downloadEarlyAccessTooltip": "Скачать эту версию раннего доступа с Civitai",
                    "downloadNotAllowedTooltip": "Эта версия доступна только для генерации на сайте Civitai",
                    "delete": "Удалить",
                    "deleteTooltip": "Удалить эту локальную версию",
                    "ignore": "Игнорировать",
@@ -1627,6 +1672,10 @@
            "noRecipeId": "ID рецепта недоступен",
            "sendToWorkflowFailed": "Не удалось отправить рецепт в рабочий процесс: {message}",
            "copyFailed": "Ошибка копирования синтаксиса рецепта: {message}",
            "createError": "Ошибка при создании рецепта：{message}",
            "createFailed": "Не удалось создать рецепт：{error}",
            "createMissingData": "Отсутствуют необходимые данные для создания рецепта",
            "created": "Рецепт успешно создан",
            "noMissingLoras": "Нет отсутствующих LoRAs для загрузки",
            "missingLorasInfoFailed": "Не удалось получить информацию для отсутствующих LoRAs",
            "preparingForDownloadFailed": "Ошибка подготовки LoRAs для загрузки",
@@ -1665,6 +1714,9 @@
            "batchImportBrowseFailed": "Failed to browse directory: {message}",
            "batchImportDirectorySelected": "Directory selected: {path}",
            "noRecipesSelected": "Рецепты не выбраны",
            "repairBulkComplete": "Восстановление завершено: {repaired} восстановлено, {skipped} пропущено (из {total})",
            "repairBulkSkipped": "Ни один из {total} выбранных рецептов не требует восстановления",
            "repairBulkFailed": "Не удалось восстановить выбранные рецепты: {message}",
            "noMissingLorasInSelection": "В выбранных рецептах не найдены отсутствующие LoRAs",
            "noLoraRootConfigured": "Корневой каталог LoRA не настроен. Пожалуйста, установите корневой каталог LoRA по умолчанию в настройках."
        },
@@ -1695,6 +1747,11 @@
            "bulkContentRatingSet": "Рейтинг контента установлен на {level} для {count} модель(ей)",
            "bulkContentRatingPartial": "Рейтинг контента {level} установлен для {success} модель(ей), {failed} не удалось",
            "bulkContentRatingFailed": "Не удалось обновить рейтинг контента для выбранных моделей",
            "bulkFavoriteUpdating": "Добавление {count} моделей в избранное...",
            "bulkUnfavoriteUpdating": "Удаление {count} моделей из избранного...",
            "bulkFavoritePartialAdded": "{success} моделей добавлено в избранное, {failed} не удалось",
            "bulkFavoritePartialRemoved": "{success} моделей удалено из избранного, {failed} не удалось",
            "bulkFavoriteFailed": "Не удалось обновить статус избранного",
            "bulkUpdatesChecking": "Проверка обновлений для выбранных {type}...",
            "bulkUpdatesSuccess": "Доступны обновления для {count} выбранных {type}",
            "bulkUpdatesNone": "Обновления для выбранных {type} не найдены",
@@ -1897,9 +1954,32 @@
            "warning": "Требует внимания",
            "error": "Требуется действие"
        },
        "issues": {
            "civitai_api_key": {
                "title": "Civitai API Key"
            },
            "cache_health": {
                "title": "Model Cache Health"
            },
            "filename_conflicts": {
                "title": "Duplicate Filename Conflicts"
            },
            "ui_version": {
                "title": "UI Version"
            }
        },
        "actions": {
            "runAgain": "Запустить снова",
-            "exportBundle": "Экспортировать пакет"
+            "exportBundle": "Экспортировать пакет",
            "open-settings": "Open Settings",
            "open-settings-syntax-format": "Switch to Full Path Syntax",
            "repair-cache": "Rebuild Cache",
            "resolve-filename-conflicts": "Resolve Conflicts",
            "reload-page": "Reload UI"
        },
        "labels": {
            "conflicts": "Conflicts",
            "version": "Version"
        },
        "toast": {
            "loadFailed": "Не удалось загрузить диагностику: {message}",
@@ -1911,6 +1991,15 @@
            "conflictsResolveFailed": "Не удалось разрешить конфликты имён файлов: {message}"
        }
    },
    "conflictConfirm": {
        "title": "Разрешить конфликты имён файлов",
        "message": "Переименование с добавлением 4-символьного хеша к каждому дублирующемуся имени файла.",
        "note": "Эта операция переименовывает файлы на диске. Если вы используете синтаксис A1111, ссылки на модели в существующих рабочих процессах могут потребовать обновления.",
        "detail": "Пример: <code>filename_v1.2</code> → <code>filename_v1.2-ab3c</code>",
        "impact": "Будет переименовано <strong>{count}</strong> файл(ов) в <strong>{groups}</strong> группе(ах) дубликатов",
        "confirm": "Переименовать файлы",
        "cancel": "Отмена"
    },
    "banners": {
        "versionMismatch": {
            "title": "Обнаружено обновление приложения",
--- a/locales/zh-CN.json
+++ b/locales/zh-CN.json
@@ -15,7 +15,10 @@
            "settings": "设置",
            "help": "帮助",
            "add": "添加",
-            "close": "关闭"
+            "close": "关闭",
            "menu": "菜单",
            "remove": "移除",
            "change": "更换"
        },
        "status": {
            "loading": "加载中...",
@@ -231,7 +234,10 @@
            "license": "许可证",
            "noCreditRequired": "无需署名",
            "allowSellingGeneratedContent": "允许销售",
            "allowSellingGeneratedContentTooltip": "允许出售生成的图片",
            "noCreditRequiredTooltip": "使用模型时无需注明原作者",
            "noTags": "无标签",
            "autoTags": "自动标签",
            "noBaseModelMatches": "没有基础模型符合当前搜索。",
            "clearAll": "清除所有筛选",
            "any": "任一",
@@ -265,10 +271,10 @@
        },
        "downloadBackend": {
            "label": "下载后端",
-            "help": "选择模型文件的下载方式。Python 使用内置下载器。aria2 使用实验性的外部下载进程。",
+            "help": "选择模型文件的下载方式。Python 使用内置下载器。aria2 使用推荐的外部下载进程。",
            "options": {
                "python": "Python（内置）",
-                "aria2": "aria2（实验性）"
+                "aria2": "aria2（推荐）"
            }
        },
        "aria2cPath": {
@@ -575,7 +581,13 @@
        },
        "misc": {
            "includeTriggerWords": "复制 LoRA 语法时包含触发词",
-            "includeTriggerWordsHelp": "复制 LoRA 语法到剪贴板时包含训练触发词"
+            "includeTriggerWordsHelp": "复制 LoRA 语法到剪贴板时包含训练触发词",
            "loraSyntaxFormat": "LoRA 语法格式",
            "loraSyntaxFormatHelp": "LoRA 语法格式。完整路径（Full）包含子文件夹路径 (<lora:style/anime/x:1.0>)，解析精确无歧义。旧版（Legacy）仅使用文件名 (<lora:x:1.0>)——A1111 原始约定，同名文件跨文件夹时可能产生歧义。",
            "loraSyntaxFormatOptions": {
                "full": "完整路径（子文件夹/名称）",
                "legacy": "旧版 A1111（仅名称）"
            }
        },
        "metadataArchive": {
            "enableArchiveDb": "启用元数据归档数据库",
@@ -639,8 +651,6 @@
            },
            "refresh": {
                "title": "刷新模型列表",
                "quick": "同步变更",
                "quickTooltip": "扫描新的或缺失的模型文件，保持列表最新。",
                "full": "重建缓存",
                "fullTooltip": "从元数据文件重新加载所有模型信息；用于列表过时或手动编辑后。"
            },
@@ -681,16 +691,29 @@
            "setContentRating": "为所选中设置内容评级",
            "copyAll": "复制所选中语法",
            "refreshAll": "刷新所选中元数据",
            "repairMetadata": "修复所选中元数据",
            "checkUpdates": "检查所选更新",
            "moveAll": "移动所选中到文件夹",
            "autoOrganize": "自动整理所选模型",
            "skipMetadataRefresh": "跳过所选模型的元数据刷新",
            "resumeMetadataRefresh": "恢复所选模型的元数据刷新",
            "setFavorite": "设为收藏",
            "setFavoriteCount": "设为收藏 ({favorited}/{total})",
            "unfavorite": "取消收藏",
            "deleteAll": "删除已选",
            "downloadMissingLoras": "下载缺失的 LoRAs",
            "downloadExamples": "下载示例图片",
            "clear": "清除选择",
            "skipMetadataRefreshCount": "跳过（{count} 个模型）",
            "resumeMetadataRefreshCount": "恢复（{count} 个模型）",
            "sendToWorkflow": "发送到工作流",
            "sections": {
                "workflow": "工作流",
                "metadata": "元数据",
                "attributes": "属性",
                "organize": "整理",
                "download": "下载"
            },
            "autoOrganizeProgress": {
                "initializing": "正在初始化自动整理...",
                "starting": "正在为 {type} 启动自动整理...",
@@ -803,8 +826,6 @@
            },
            "refresh": {
                "title": "刷新配方列表",
                "quick": "同步变更",
                "quickTooltip": "同步变更 - 快速刷新而不重建缓存",
                "full": "重建缓存",
                "fullTooltip": "重建缓存 - 重新扫描所有配方文件"
            },
@@ -944,6 +965,13 @@
        "empty": {
            "noFolders": "未找到文件夹",
            "dragHint": "拖拽项目到此处以创建文件夹"
        },
        "folderUpdateCheck": {
            "label": "检查此文件夹的更新",
            "loading": "正在检查此文件夹中的{type}更新...",
            "success": "在此文件夹中找到 {count} 个{type}更新",
            "none": "此文件夹中的所有{type}都是最新版本",
            "error": "检查文件夹{type}更新失败: {message}"
        }
    },
    "statistics": {
@@ -988,9 +1016,9 @@
        "download": {
            "title": "从 URL 下载模型",
            "titleWithType": "从 URL 下载 {type}",
            "url": "Civitai URL",
            "civitaiUrl": "Civitai URL:",
            "placeholder": "https://civitai.com/models/...",
            "urlHint": "每行输入一个 CivitAI 或 CivArchive URL。支持批量下载多个 URL。",
            "locationPreview": "下载位置预览",
            "useDefaultPath": "使用默认路径",
            "useDefaultPathTooltip": "启用后，文件将自动按配置的路径模板进行整理",
@@ -1012,6 +1040,11 @@
            "downloadedTooltip": "之前已下载，但当前不在你的库中。",
            "alreadyInLibrary": "已存在于库中",
            "autoOrganizedPath": "【已按路径模板自动整理】",
            "fileSelection": {
                "title": "选择文件格式",
                "files": "个文件",
                "select": "选择文件"
            },
            "errors": {
                "invalidUrl": "无效的 Civitai URL 格式",
                "noVersions": "此模型没有可用版本"
@@ -1076,6 +1109,12 @@
            "countMessage": "模型将被永久删除。",
            "action": "全部删除"
        },
        "bulkDeleteRecipes": {
            "title": "删除多个配方",
            "message": "你确定要删除所有选中的配方及其相关文件吗？",
            "countMessage": "配方将被永久删除。",
            "action": "全部删除"
        },
        "checkUpdates": {
            "title": "检查所有 {type} 的更新？",
            "message": "这会为库中的每个 {type} 检查更新，大型集合可能需要一些时间。",
@@ -1156,6 +1195,7 @@
                "editModelName": "编辑模型名称",
                "editFileName": "编辑文件名",
                "editBaseModel": "编辑基础模型",
                "editVersionName": "编辑版本名称",
                "viewOnCivitai": "在 Civitai 查看",
                "viewOnCivitaiText": "在 Civitai 查看",
                "viewCreatorProfile": "查看创作者主页",
@@ -1187,7 +1227,9 @@
            },
            "notes": {
                "saved": "备注保存成功",
-                "saveFailed": "备注保存失败"
+                "saveFailed": "备注保存失败",
                "showMore": "展开",
                "showLess": "收起"
            },
            "usageTips": {
                "addPresetParameter": "添加预设参数...",
@@ -1292,12 +1334,15 @@
                    "earlyAccess": "抢先体验",
                    "earlyAccessTooltip": "此版本当前需要 Civitai 抢先体验权限",
                    "ignored": "已忽略",
-                    "ignoredTooltip": "此版本已关闭更新通知"
+                    "ignoredTooltip": "此版本已关闭更新通知",
                    "onSiteOnly": "仅站内生成",
                    "onSiteOnlyTooltip": "此版本仅在 Civitai 站内可用，无法下载"
                },
                "actions": {
                    "download": "下载",
                    "downloadTooltip": "下载此版本",
                    "downloadEarlyAccessTooltip": "从 Civitai 下载此抢先体验版本",
                    "downloadNotAllowedTooltip": "此版本仅在 Civitai 站内可用，无法下载",
                    "delete": "删除",
                    "deleteTooltip": "删除此本地版本",
                    "ignore": "忽略",
@@ -1627,6 +1672,10 @@
            "noRecipeId": "无配方 ID",
            "sendToWorkflowFailed": "发送配方到工作流失败：{message}",
            "copyFailed": "复制配方语法出错：{message}",
            "createError": "创建配方时出错：{message}",
            "createFailed": "创建配方失败：{error}",
            "createMissingData": "缺少创建配方所需的数据",
            "created": "配方创建成功",
            "noMissingLoras": "没有缺失的 LoRA 可下载",
            "missingLorasInfoFailed": "获取缺失 LoRA 信息失败",
            "preparingForDownloadFailed": "准备下载 LoRA 时出错",
@@ -1665,6 +1714,9 @@
            "batchImportBrowseFailed": "浏览目录失败：{message}",
            "batchImportDirectorySelected": "已选择目录：{path}",
            "noRecipesSelected": "未选择任何配方",
            "repairBulkComplete": "修复完成：{repaired} 个已修复，{skipped} 个已跳过（共 {total} 个）",
            "repairBulkSkipped": "所选 {total} 个配方无需修复",
            "repairBulkFailed": "修复所选配方失败：{message}",
            "noMissingLorasInSelection": "在选定的配方中未找到缺失的 LoRAs",
            "noLoraRootConfigured": "未配置 LoRA 根目录。请在设置中设置默认的 LoRA 根目录。"
        },
@@ -1695,6 +1747,11 @@
            "bulkContentRatingSet": "已将 {count} 个模型的内容评级设置为 {level}",
            "bulkContentRatingPartial": "已将 {success} 个模型的内容评级设置为 {level}，{failed} 个失败",
            "bulkContentRatingFailed": "未能更新所选模型的内容评级",
            "bulkFavoriteUpdating": "正在将 {count} 个模型添加到收藏...",
            "bulkUnfavoriteUpdating": "正在将 {count} 个模型从收藏移除...",
            "bulkFavoritePartialAdded": "已将 {success} 个模型添加到收藏，{failed} 个失败",
            "bulkFavoritePartialRemoved": "已将 {success} 个模型从收藏移除，{failed} 个失败",
            "bulkFavoriteFailed": "更新收藏状态失败",
            "bulkUpdatesChecking": "正在检查所选 {type} 的更新...",
            "bulkUpdatesSuccess": "{count} 个所选 {type} 有可用更新",
            "bulkUpdatesNone": "所选 {type} 未发现更新",
@@ -1897,9 +1954,32 @@
            "warning": "需要关注",
            "error": "需要处理"
        },
        "issues": {
            "civitai_api_key": {
                "title": "Civitai API 密钥"
            },
            "cache_health": {
                "title": "模型缓存健康状态"
            },
            "filename_conflicts": {
                "title": "文件名重复冲突"
            },
            "ui_version": {
                "title": "UI 版本"
            }
        },
        "actions": {
            "runAgain": "重新检查",
-            "exportBundle": "导出诊断包"
+            "exportBundle": "导出诊断包",
            "open-settings": "打开设置",
            "open-settings-syntax-format": "切换为完整路径语法",
            "repair-cache": "重建缓存",
            "resolve-filename-conflicts": "解决冲突",
            "reload-page": "刷新 UI"
        },
        "labels": {
            "conflicts": "冲突详情",
            "version": "版本信息"
        },
        "toast": {
            "loadFailed": "加载诊断结果失败：{message}",
@@ -1911,6 +1991,15 @@
            "conflictsResolveFailed": "解决文件名冲突失败：{message}"
        }
    },
    "conflictConfirm": {
        "title": "解决文件名冲突",
        "message": "通过在每个重复文件名后附加 4 位哈希值来重命名文件。",
        "note": "此操作会重命名磁盘上的文件。如果使用 A1111 语法格式，现有工作流中的模型引用可能需要更新。",
        "detail": "示例：<code>filename_v1.2</code> → <code>filename_v1.2-ab3c</code>",
        "impact": "将重命名 <strong>{count}</strong> 个文件（共 <strong>{groups}</strong> 组重复）",
        "confirm": "重命名文件",
        "cancel": "取消"
    },
    "banners": {
        "versionMismatch": {
            "title": "检测到应用更新",
--- a/locales/zh-TW.json
+++ b/locales/zh-TW.json
@@ -15,7 +15,10 @@
            "settings": "設定",
            "help": "說明",
            "add": "新增",
-            "close": "關閉"
+            "close": "關閉",
            "menu": "選單",
            "remove": "移除",
            "change": "更換"
        },
        "status": {
            "loading": "載入中...",
@@ -231,7 +234,10 @@
            "license": "授權",
            "noCreditRequired": "無需署名",
            "allowSellingGeneratedContent": "允許銷售",
            "allowSellingGeneratedContentTooltip": "允許出售生成的圖片",
            "noCreditRequiredTooltip": "使用模型時無需註明原作者",
            "noTags": "無標籤",
            "autoTags": "自動標籤",
            "noBaseModelMatches": "沒有基礎模型符合目前的搜尋。",
            "clearAll": "清除所有篩選",
            "any": "任一",
@@ -265,10 +271,10 @@
        },
        "downloadBackend": {
            "label": "下載後端",
-            "help": "選擇模型檔案的下載方式。Python 使用內建下載器。aria2 使用實驗性的外部下載程序。",
+            "help": "選擇模型檔案的下載方式。Python 使用內建下載器。aria2 使用推薦的外部下載程序。",
            "options": {
                "python": "Python（內建）",
-                "aria2": "aria2（實驗性）"
+                "aria2": "aria2（推薦）"
            }
        },
        "aria2cPath": {
@@ -575,7 +581,13 @@
        },
        "misc": {
            "includeTriggerWords": "在 LoRA 語法中包含觸發詞",
-            "includeTriggerWordsHelp": "複製 LoRA 語法到剪貼簿時包含訓練觸發詞"
+            "includeTriggerWordsHelp": "複製 LoRA 語法到剪貼簿時包含訓練觸發詞",
            "loraSyntaxFormat": "LoRA 語法格式",
            "loraSyntaxFormatHelp": "LoRA 語法格式。完整路徑（Full）包含子資料夾路徑 (<lora:style/anime/x:1.0>)，解析精確無歧義。舊版（Legacy）僅使用檔名 (<lora:x:1.0>)——A1111 原始約定，同名檔案跨資料夾時可能產生歧義。",
            "loraSyntaxFormatOptions": {
                "full": "完整路徑（子資料夾/名稱）",
                "legacy": "舊版 A1111（僅名稱）"
            }
        },
        "metadataArchive": {
            "enableArchiveDb": "啟用中繼資料封存資料庫",
@@ -639,8 +651,6 @@
            },
            "refresh": {
                "title": "重新整理模型列表",
                "quick": "同步變更",
                "quickTooltip": "掃描新的或缺少的模型檔案，讓清單保持最新。",
                "full": "重建快取",
                "fullTooltip": "從中繼資料檔重新載入所有模型資訊；適用於清單過時或手動編輯後。"
            },
@@ -681,16 +691,29 @@
            "setContentRating": "為全部設定內容分級",
            "copyAll": "複製全部語法",
            "refreshAll": "刷新全部 metadata",
            "repairMetadata": "修復所選中元數據",
            "checkUpdates": "檢查所選更新",
            "moveAll": "全部移動到資料夾",
            "autoOrganize": "自動整理所選模型",
            "skipMetadataRefresh": "跳過所選模型的元數據更新",
            "resumeMetadataRefresh": "恢復所選模型的元數據更新",
            "setFavorite": "設為收藏",
            "setFavoriteCount": "設為收藏 ({favorited}/{total})",
            "unfavorite": "取消收藏",
            "deleteAll": "刪除所選",
            "downloadMissingLoras": "下載缺失的 LoRAs",
            "downloadExamples": "下載範例圖片",
            "clear": "清除選取",
            "skipMetadataRefreshCount": "跳過（{count} 個模型）",
            "resumeMetadataRefreshCount": "恢復（{count} 個模型）",
            "sendToWorkflow": "發送到工作流",
            "sections": {
                "workflow": "工作流",
                "metadata": "元數據",
                "attributes": "屬性",
                "organize": "整理",
                "download": "下載"
            },
            "autoOrganizeProgress": {
                "initializing": "正在初始化自動整理...",
                "starting": "正在開始自動整理 {type}...",
@@ -803,8 +826,6 @@
            },
            "refresh": {
                "title": "重新整理配方列表",
                "quick": "同步變更",
                "quickTooltip": "同步變更 - 快速重新整理而不重建快取",
                "full": "重建快取",
                "fullTooltip": "重建快取 - 重新掃描所有配方檔案"
            },
@@ -944,6 +965,13 @@
        "empty": {
            "noFolders": "未找到資料夾",
            "dragHint": "將項目拖到此處以建立資料夾"
        },
        "folderUpdateCheck": {
            "label": "檢查此資料夾的更新",
            "loading": "正在檢查此資料夾中的{type}更新...",
            "success": "在此資料夾中找到 {count} 個{type}更新",
            "none": "此資料夾中的所有{type}都是最新版本",
            "error": "檢查資料夾{type}更新失敗: {message}"
        }
    },
    "statistics": {
@@ -988,9 +1016,9 @@
        "download": {
            "title": "從網址下載模型",
            "titleWithType": "從網址下載 {type}",
            "url": "Civitai 網址",
            "civitaiUrl": "Civitai 網址：",
            "placeholder": "https://civitai.com/models/...",
            "urlHint": "每行輸入一個 CivitAI 或 CivArchive URL。支援批量下載多個 URL。",
            "locationPreview": "下載位置預覽",
            "useDefaultPath": "使用預設路徑",
            "useDefaultPathTooltip": "啟用後，檔案將依照設定的路徑範本自動整理",
@@ -1012,6 +1040,11 @@
            "downloadedTooltip": "先前已下載，但目前不在你的庫中。",
            "alreadyInLibrary": "已在庫存",
            "autoOrganizedPath": "[依路徑範本自動整理]",
            "fileSelection": {
                "title": "選擇檔案格式",
                "files": "個檔案",
                "select": "選擇檔案"
            },
            "errors": {
                "invalidUrl": "Civitai 網址格式無效",
                "noVersions": "此模型無可用版本"
@@ -1076,6 +1109,12 @@
            "countMessage": "模型將被永久刪除。",
            "action": "全部刪除"
        },
        "bulkDeleteRecipes": {
            "title": "刪除多個配方",
            "message": "您確定要刪除所有選取的配方及其相關檔案嗎？",
            "countMessage": "配方將被永久刪除。",
            "action": "全部刪除"
        },
        "checkUpdates": {
            "title": "要檢查所有 {type} 的更新嗎？",
            "message": "這會為資料庫中的每個 {type} 檢查更新，大型收藏可能會花上一些時間。",
@@ -1156,6 +1195,7 @@
                "editModelName": "編輯模型名稱",
                "editFileName": "編輯檔案名稱",
                "editBaseModel": "編輯基礎模型",
                "editVersionName": "編輯版本名稱",
                "viewOnCivitai": "在 Civitai 查看",
                "viewOnCivitaiText": "在 Civitai 查看",
                "viewCreatorProfile": "查看創作者個人檔案",
@@ -1187,7 +1227,9 @@
            },
            "notes": {
                "saved": "備註已儲存",
-                "saveFailed": "儲存備註失敗"
+                "saveFailed": "儲存備註失敗",
                "showMore": "展開",
                "showLess": "收起"
            },
            "usageTips": {
                "addPresetParameter": "新增預設參數...",
@@ -1292,12 +1334,15 @@
                    "earlyAccess": "搶先體驗",
                    "earlyAccessTooltip": "此版本目前需要 Civitai 搶先體驗權限",
                    "ignored": "已忽略",
-                    "ignoredTooltip": "此版本已關閉更新通知"
+                    "ignoredTooltip": "此版本已關閉更新通知",
                    "onSiteOnly": "僅站內生成",
                    "onSiteOnlyTooltip": "此版本僅在 Civitai 站內可用，無法下載"
                },
                "actions": {
                    "download": "下載",
                    "downloadTooltip": "下載此版本",
                    "downloadEarlyAccessTooltip": "從 Civitai 下載此搶先體驗版本",
                    "downloadNotAllowedTooltip": "此版本僅在 Civitai 站內可用，無法下載",
                    "delete": "刪除",
                    "deleteTooltip": "刪除此本地版本",
                    "ignore": "忽略",
@@ -1627,6 +1672,10 @@
            "noRecipeId": "無配方 ID",
            "sendToWorkflowFailed": "傳送配方到工作流失敗：{message}",
            "copyFailed": "複製配方語法錯誤：{message}",
            "createError": "建立配方時發生錯誤：{message}",
            "createFailed": "建立配方失敗：{error}",
            "createMissingData": "缺少建立配方所需的資料",
            "created": "配方建立成功",
            "noMissingLoras": "無缺少的 LoRA 可下載",
            "missingLorasInfoFailed": "取得缺少 LoRA 資訊失敗",
            "preparingForDownloadFailed": "準備下載 LoRA 時發生錯誤",
@@ -1665,6 +1714,9 @@
            "batchImportBrowseFailed": "瀏覽目錄失敗：{message}",
            "batchImportDirectorySelected": "已選擇目錄：{path}",
            "noRecipesSelected": "未選取任何食譜",
            "repairBulkComplete": "修復完成：{repaired} 個已修復，{skipped} 個已跳過（共 {total} 個）",
            "repairBulkSkipped": "所選 {total} 個配方無需修復",
            "repairBulkFailed": "修復所選配方失敗：{message}",
            "noMissingLorasInSelection": "在選取的食譜中未找到缺失的 LoRAs",
            "noLoraRootConfigured": "未配置 LoRA 根目錄。請在設定中設定預設的 LoRA 根目錄。"
        },
@@ -1695,6 +1747,11 @@
            "bulkContentRatingSet": "已將 {count} 個模型的內容分級設定為 {level}",
            "bulkContentRatingPartial": "已將 {success} 個模型的內容分級設定為 {level}，{failed} 個失敗",
            "bulkContentRatingFailed": "無法更新所選模型的內容分級",
            "bulkFavoriteUpdating": "正在將 {count} 個模型加入收藏...",
            "bulkUnfavoriteUpdating": "正在將 {count} 個模型從收藏移除...",
            "bulkFavoritePartialAdded": "已將 {success} 個模型加入收藏，{failed} 個失敗",
            "bulkFavoritePartialRemoved": "已將 {success} 個模型從收藏移除，{failed} 個失敗",
            "bulkFavoriteFailed": "更新收藏狀態失敗",
            "bulkUpdatesChecking": "正在檢查所選 {type} 的更新...",
            "bulkUpdatesSuccess": "{count} 個所選 {type} 有可用更新",
            "bulkUpdatesNone": "所選 {type} 未找到更新",
@@ -1897,9 +1954,32 @@
            "warning": "需要注意",
            "error": "需要處理"
        },
        "issues": {
            "civitai_api_key": {
                "title": "Civitai API 金鑰"
            },
            "cache_health": {
                "title": "模型快取健康狀態"
            },
            "filename_conflicts": {
                "title": "檔案名稱重複衝突"
            },
            "ui_version": {
                "title": "UI 版本"
            }
        },
        "actions": {
            "runAgain": "重新執行",
-            "exportBundle": "匯出套件"
+            "exportBundle": "匯出套件",
            "open-settings": "開啟設定",
            "open-settings-syntax-format": "切換為完整路徑語法",
            "repair-cache": "重建快取",
            "resolve-filename-conflicts": "解決衝突",
            "reload-page": "重新載入 UI"
        },
        "labels": {
            "conflicts": "衝突詳情",
            "version": "版本"
        },
        "toast": {
            "loadFailed": "載入診斷失敗：{message}",
@@ -1911,6 +1991,15 @@
            "conflictsResolveFailed": "解決檔案名稱衝突失敗：{message}"
        }
    },
    "conflictConfirm": {
        "title": "解決檔案名稱衝突",
        "message": "通過在每個重複檔案名稱後附加 4 位元哈希值來重新命名檔案。",
        "note": "此操作會重新命名磁碟上的檔案。如果使用 A1111 語法格式，現有工作流程中的模型參考可能需要更新。",
        "detail": "示例：<code>filename_v1.2</code> → <code>filename_v1.2-ab3c</code>",
        "impact": "將重新命名 <strong>{count}</strong> 個檔案（共 <strong>{groups}</strong> 組重複）",
        "confirm": "重新命名檔案",
        "cancel": "取消"
    },
    "banners": {
        "versionMismatch": {
            "title": "偵測到應用程式更新",
--- a/py/config.py
+++ b/py/config.py
@@ -172,6 +172,12 @@ class Config:
        self.extra_unet_roots: List[str] = []
        self.extra_embeddings_roots: List[str] = []
        self.recipes_path: str = ""
        # Load extra folder paths from active library settings before symlink scan
        # so both primary and extra paths are discovered in a single pass.
        if not standalone_mode:
            self._load_extra_paths_from_settings()
        # Scan symbolic links during initialization
        self._initialize_symlink_mappings()
@@ -179,6 +185,96 @@ class Config:
            # Save the paths to settings.json when running in ComfyUI mode
            self.save_folder_paths_to_settings()
    def _load_extra_paths_from_settings(self) -> None:
        """Read extra folder paths from the active library and apply them.
        Called during ``Config.__init__`` before the symlink scan so both primary and
        extra paths are discovered in a single pass.  Mirrors the extra-path
        portion of ``_apply_library_paths`` without replacing the primary roots
        that were already resolved from ComfyUI's ``folder_paths``.
        """
        try:
            from .services.settings_manager import get_settings_manager
            settings_manager = get_settings_manager()
            library_name = settings_manager.get_active_library_name()
            libraries = settings_manager.get_libraries()
            if not library_name or library_name not in libraries:
                return
            library_config = libraries[library_name]
            if not isinstance(library_config, dict):
                return
            extra_folder_paths = library_config.get("extra_folder_paths")
            if not isinstance(extra_folder_paths, dict):
                return
            extra_lora = extra_folder_paths.get("loras", []) or []
            extra_checkpoint = extra_folder_paths.get("checkpoints", []) or []
            extra_unet = extra_folder_paths.get("unet", []) or []
            extra_embedding = extra_folder_paths.get("embeddings", []) or []
            if not any([extra_lora, extra_checkpoint, extra_unet, extra_embedding]):
                return
            filtered_extra_lora = self._filter_overlapping_extra_lora_paths(
                self.loras_roots, extra_lora
            )
            self.extra_loras_roots = self._prepare_lora_paths(filtered_extra_lora)
            (
                _,
                self.extra_checkpoints_roots,
                self.extra_unet_roots,
            ) = self._prepare_checkpoint_paths(extra_checkpoint, extra_unet)
            self.extra_embeddings_roots = self._prepare_embedding_paths(
                extra_embedding
            )
            recipes_path = library_config.get("recipes_path", "")
            if isinstance(recipes_path, str) and recipes_path:
                self.recipes_path = recipes_path
            if self.extra_loras_roots:
                logger.info(
                    "Found extra LoRA roots:"
                    + "\n - "
                    + "\n - ".join(self.extra_loras_roots)
                )
            if self.extra_checkpoints_roots:
                logger.info(
                    "Found extra checkpoint roots:"
                    + "\n - "
                    + "\n - ".join(self.extra_checkpoints_roots)
                )
            if self.extra_unet_roots:
                logger.info(
                    "Found extra diffusion model roots:"
                    + "\n - "
                    + "\n - ".join(self.extra_unet_roots)
                )
            if self.extra_embeddings_roots:
                logger.info(
                    "Found extra embedding roots:"
                    + "\n - "
                    + "\n - ".join(self.extra_embeddings_roots)
                )
            logger.info(
                "Applied library settings for '%s' with extra paths: loras=%s, "
                "checkpoints=%s, embeddings=%s",
                library_name,
                extra_lora,
                extra_checkpoint,
                extra_embedding,
            )
        except Exception as exc:
            logger.debug(
                "Could not load extra paths from library settings: %s", exc
            )
    def save_folder_paths_to_settings(self):
        """Persist ComfyUI-derived folder paths to the multi-library settings."""
        try:
--- a/py/lora_manager.py
+++ b/py/lora_manager.py
@@ -184,39 +184,6 @@ class LoraManager:
    async def _initialize_services(cls):
        """Initialize all services using the ServiceRegistry"""
        try:
            # Apply library settings to load extra folder paths before scanning
            # Only apply if extra paths haven't been loaded yet (preserves test mocks)
            try:
                from .services.settings_manager import get_settings_manager
                settings_manager = get_settings_manager()
                library_name = settings_manager.get_active_library_name()
                libraries = settings_manager.get_libraries()
                if library_name and library_name in libraries:
                    library_config = libraries[library_name]
                    # Only apply settings if extra paths are not already configured
                    # This preserves values set by tests via monkeypatch
                    extra_paths = library_config.get("extra_folder_paths", {})
                    has_extra_paths = (
                        config.extra_loras_roots
                        or config.extra_checkpoints_roots
                        or config.extra_unet_roots
                        or config.extra_embeddings_roots
                    )
                    if not has_extra_paths and any(extra_paths.values()):
                        config.apply_library_settings(library_config)
                        logger.info(
                            "Applied library settings for '%s' with extra paths: loras=%s, checkpoints=%s, embeddings=%s",
                            library_name,
                            extra_paths.get("loras", []),
                            extra_paths.get("checkpoints", []),
                            extra_paths.get("embeddings", []),
                        )
            except Exception as exc:
                logger.warning(
                    "Failed to apply library settings during initialization: %s", exc
                )
            # Initialize CivitaiClient first to ensure it's ready for other services
            await ServiceRegistry.get_civitai_client()
--- a/py/metadata_collector/metadata_processor.py
+++ b/py/metadata_collector/metadata_processor.py
@@ -560,8 +560,14 @@ class MetadataProcessor:
        params["loras"] = " ".join(lora_parts)
-        # Set default clip_skip value
+        # Extract clip_skip from any SAMPLING node that provides it
-        params["clip_skip"] = "1"  # Common default
+        for sampler_info in metadata.get(SAMPLING, {}).values():
            clip_skip = sampler_info.get("parameters", {}).get("clip_skip")
            if clip_skip is not None:
                params["clip_skip"] = clip_skip
                break
        if params["clip_skip"] is None:
            params["clip_skip"] = "1"
        return params
--- a/py/metadata_collector/node_extractors.py
+++ b/py/metadata_collector/node_extractors.py
@@ -144,6 +144,118 @@ class TSCCheckpointLoaderExtractor(NodeMetadataExtractor):
                metadata[PROMPTS][node_id]["positive_encoded"] = positive_conditioning
                metadata[PROMPTS][node_id]["negative_encoded"] = negative_conditioning
 class EasyComfyLoaderExtractor(NodeMetadataExtractor):
    @staticmethod
    def extract(node_id, inputs, outputs, metadata):
        if not inputs:
            return
        if "ckpt_name" in inputs:
            _store_checkpoint_metadata(metadata, node_id, inputs["ckpt_name"])
        # Only extract from optional_lora_stack — skip the single lora_name to
        # avoid double-counting LoRAs that come through the LORA_STACK path.
        active_loras = []
        optional_lora_stack = inputs.get("optional_lora_stack")
        if optional_lora_stack is not None and isinstance(optional_lora_stack, (list, tuple)):
            for item in optional_lora_stack:
                if isinstance(item, (list, tuple)) and len(item) >= 2:
                    lora_path = item[0]
                    model_strength = item[1]
                    lora_name = os.path.splitext(os.path.basename(lora_path))[0]
                    active_loras.append({
                        "name": lora_name,
                        "strength": model_strength
                    })
        if active_loras:
            metadata[LORAS][node_id] = {
                "lora_list": active_loras,
                "node_id": node_id
            }
        positive_text = inputs.get("positive", "")
        negative_text = inputs.get("negative", "")
        if positive_text or negative_text:
            if node_id not in metadata[PROMPTS]:
                metadata[PROMPTS][node_id] = {"node_id": node_id}
            metadata[PROMPTS][node_id]["positive_text"] = positive_text
            metadata[PROMPTS][node_id]["negative_text"] = negative_text
        if "clip_skip" in inputs:
            clip_skip = inputs["clip_skip"]
            if node_id not in metadata[SAMPLING]:
                metadata[SAMPLING][node_id] = {"parameters": {}, "node_id": node_id}
            metadata[SAMPLING][node_id]["parameters"]["clip_skip"] = clip_skip
        width = inputs.get("empty_latent_width")
        height = inputs.get("empty_latent_height")
        if width is not None and height is not None:
            if SIZE not in metadata:
                metadata[SIZE] = {}
            metadata[SIZE][node_id] = {
                "width": int(width),
                "height": int(height),
                "node_id": node_id
            }
    @staticmethod
    def update(node_id, outputs, metadata):
        # outputs: [(pipe_dict, model, vae), ...]
        if not outputs or not isinstance(outputs, list) or len(outputs) == 0:
            return
        first_output = outputs[0]
        if not isinstance(first_output, tuple) or len(first_output) < 1:
            return
        pipe = first_output[0]
        if not isinstance(pipe, dict):
            return
        positive_conditioning = pipe.get("positive")
        negative_conditioning = pipe.get("negative")
        if positive_conditioning is not None or negative_conditioning is not None:
            if node_id not in metadata[PROMPTS]:
                metadata[PROMPTS][node_id] = {"node_id": node_id}
            if positive_conditioning is not None:
                metadata[PROMPTS][node_id]["positive_encoded"] = positive_conditioning
            if negative_conditioning is not None:
                metadata[PROMPTS][node_id]["negative_encoded"] = negative_conditioning
 class EasyPreSamplingExtractor(NodeMetadataExtractor):
    @staticmethod
    def extract(node_id, inputs, outputs, metadata):
        if not inputs:
            return
        sampling_params = {}
        for key in ("steps", "cfg", "sampler_name", "scheduler", "denoise", "seed"):
            if key in inputs:
                sampling_params[key] = inputs[key]
        metadata[SAMPLING][node_id] = {
            "parameters": sampling_params,
            "node_id": node_id,
            IS_SAMPLER: True
        }
 class EasySeedExtractor(NodeMetadataExtractor):
    @staticmethod
    def extract(node_id, inputs, outputs, metadata):
        if not inputs or "seed" not in inputs:
            return
        metadata[SAMPLING][node_id] = {
            "parameters": {"seed": inputs["seed"]},
            "node_id": node_id,
            IS_SAMPLER: False
        }
 class CLIPTextEncodeExtractor(NodeMetadataExtractor):
    @staticmethod
    def extract(node_id, inputs, outputs, metadata):
@@ -1013,9 +1125,12 @@ NODE_EXTRACTORS = {
    "KSamplerSelect": KSamplerSelectExtractor,  # Add KSamplerSelect
    "BasicScheduler": BasicSchedulerExtractor,  # Add BasicScheduler
    "AlignYourStepsScheduler": BasicSchedulerExtractor,  # Add AlignYourStepsScheduler
    # ComfyUI-Easy-Use pre-sampling / seed
    "samplerSettings": EasyPreSamplingExtractor,  # easy preSampling
    "easySeed": EasySeedExtractor,  # easy seed
    # Loaders
    "CheckpointLoaderSimple": CheckpointLoaderExtractor,
-    "comfyLoader": CheckpointLoaderExtractor,  # easy comfyLoader
+    "comfyLoader": EasyComfyLoaderExtractor,  # ComfyUI-Easy-Use easy comfyLoader
    "CheckpointLoaderSimpleWithImages": CheckpointLoaderExtractor,  # CheckpointLoader|pysssss
    "TSC_EfficientLoader": TSCCheckpointLoaderExtractor,  # Efficient Nodes
    "NunchakuFluxDiTLoader": NunchakuFluxDiTLoaderExtractor,  # ComfyUI-Nunchaku
--- a/py/nodes/lora_loader.py
+++ b/py/nodes/lora_loader.py
@@ -9,6 +9,7 @@ from ..utils.utils import get_lora_info_absolute
 from .utils import (
    FlexibleOptionalInputType,
    any_type,
    apply_lora_syntax_format,
    detect_nunchaku_model_kind,
    extract_lora_name,
    get_loras_list,
@@ -52,7 +53,7 @@ def _collect_widget_entries(kwargs):
    for lora in get_loras_list(kwargs):
        if not lora.get("active", False):
            continue
-        lora_name = lora["name"]
+        lora_name = apply_lora_syntax_format(lora["name"])
        model_strength = float(lora["strength"])
        clip_strength = float(lora.get("clipStrength", model_strength))
        lora_path, trigger_words = get_lora_info_absolute(lora_name)
--- a/py/nodes/lora_stacker.py
+++ b/py/nodes/lora_stacker.py
@@ -1,6 +1,6 @@
 import os
 from ..utils.utils import get_lora_info
-from .utils import FlexibleOptionalInputType, any_type, extract_lora_name, get_loras_list
+from .utils import FlexibleOptionalInputType, any_type, apply_lora_syntax_format, extract_lora_name, get_loras_list
 import logging
@@ -48,7 +48,7 @@ class LoraStackerLM:
            if not lora.get('active', False):
                continue
-            lora_name = lora['name']
+            lora_name = apply_lora_syntax_format(lora['name'])
            model_strength = float(lora['strength'])
            # Get clip strength - use model strength as default if not specified
            clip_strength = float(lora.get('clipStrength', model_strength))
--- a/py/nodes/utils.py
+++ b/py/nodes/utils.py
@@ -44,11 +44,29 @@ import folder_paths  # type: ignore
 logger = logging.getLogger(__name__)
 def get_lora_syntax_format():
    try:
        from ..services.settings_manager import get_settings_manager
        return get_settings_manager().get("lora_syntax_format", "legacy")
    except Exception:
        return "legacy"
 def apply_lora_syntax_format(name):
    fmt = get_lora_syntax_format()
    if fmt == "legacy":
        return name.replace("\\", "/").rstrip("/").split("/")[-1]
    return name
 def extract_lora_name(lora_path):
-    """Extract the lora name from a lora path (e.g., 'IL\\aorunIllstrious.safetensors' -> 'aorunIllstrious')"""
+    normalized = lora_path.replace("\\", "/")
-    # Get the basename without extension
+    basename = os.path.basename(normalized)
-    basename = os.path.basename(lora_path)
+    name_no_ext = os.path.splitext(basename)[0]
-    return os.path.splitext(basename)[0]
+    dirname = os.path.dirname(normalized)
    if dirname and dirname not in (".", "/") and not normalized.startswith("/"):
        return apply_lora_syntax_format(f"{dirname}/{name_no_ext}")
    return apply_lora_syntax_format(name_no_ext)
 def get_loras_list(kwargs):
--- a/py/nodes/wanvideo_lora_select.py
+++ b/py/nodes/wanvideo_lora_select.py
@@ -1,10 +1,22 @@
-import folder_paths # type: ignore
+import os
-from ..utils.utils import get_lora_info
+from ..utils.utils import get_lora_info_absolute
 from ..config import config
 from .utils import FlexibleOptionalInputType, any_type, get_loras_list
 import logging
 logger = logging.getLogger(__name__)
 def _relpath_within_loras(abs_path):
    """Return abs_path relative to the first matching lora root, or basename as fallback."""
    all_roots = list(config.loras_roots or []) + list(config.extra_loras_roots or [])
    for root in all_roots:
        try:
            return os.path.relpath(abs_path, root)
        except ValueError:
            continue
    return os.path.basename(abs_path)
 class WanVideoLoraSelectLM:
    NAME = "WanVideo Lora Select (LoraManager)"
    CATEGORY = "Lora Manager/stackers"
@@ -56,13 +68,13 @@ class WanVideoLoraSelectLM:
            clip_strength = float(lora.get('clipStrength', model_strength))
            # Get lora path and trigger words
-            lora_path, trigger_words = get_lora_info(lora_name)
+            lora_path, trigger_words = get_lora_info_absolute(lora_name)
            # Create lora item for WanVideo format
            lora_item = {
-                "path": folder_paths.get_full_path("loras", lora_path),
+                "path": lora_path,
                "strength": model_strength,
-                "name": lora_path.split(".")[0],
+                "name": os.path.splitext(_relpath_within_loras(lora_path))[0],
                "blocks": selected_blocks,
                "layer_filter": layer_filter,
                "low_mem_load": low_mem_load,
--- a/py/nodes/wanvideo_lora_select_from_text.py
+++ b/py/nodes/wanvideo_lora_select_from_text.py
@@ -1,11 +1,23 @@
-import folder_paths # type: ignore
+import os
-from ..utils.utils import get_lora_info
+from ..utils.utils import get_lora_info_absolute
 from ..config import config
 from .utils import any_type 
 import logging
 # 初始化日志记录器
 logger = logging.getLogger(__name__)
 def _relpath_within_loras(abs_path):
    """Return abs_path relative to the first matching lora root, or basename as fallback."""
    all_roots = list(config.loras_roots or []) + list(config.extra_loras_roots or [])
    for root in all_roots:
        try:
            return os.path.relpath(abs_path, root)
        except ValueError:
            continue
    return os.path.basename(abs_path)
 # 定义新节点的类
 class WanVideoLoraTextSelectLM:
    # 节点在UI中显示的名称
@@ -87,12 +99,12 @@ class WanVideoLoraTextSelectLM:
            else:
                continue
-            lora_path, trigger_words = get_lora_info(lora_name_raw)
+            lora_path, trigger_words = get_lora_info_absolute(lora_name_raw)
            lora_item = {
-                "path": folder_paths.get_full_path("loras", lora_path),
+                "path": lora_path,
                "strength": model_strength,
-                "name": lora_path.split(".")[0],
+                "name": os.path.splitext(_relpath_within_loras(lora_path))[0],
                "blocks": selected_blocks,
                "layer_filter": layer_filter,
                "low_mem_load": low_mem_load,
--- a/py/recipes/base.py
+++ b/py/recipes/base.py
@@ -7,7 +7,7 @@ import re
 from typing import Dict, List, Any, Optional, Tuple
 from abc import ABC, abstractmethod
 from ..config import config
-from ..utils.constants import VALID_LORA_TYPES
+from ..utils.constants import VALID_LORA_TYPES, VALID_CHECKPOINT_SUB_TYPES
 from ..utils.civitai_utils import rewrite_preview_url
 logger = logging.getLogger(__name__)
@@ -58,9 +58,52 @@ class RecipeMetadataParser(ABC):
            civitai_info, error_msg = civitai_info_tuple if isinstance(civitai_info_tuple, tuple) else (civitai_info_tuple, None)
            if not civitai_info or error_msg == "Model not found":
-                # Model not found or deleted
+                # CivitAI may fail to resolve a hash that is still being
-                lora_entry['isDeleted'] = True
+                # computed (known CivitAI issue). Before marking as deleted,
-                lora_entry['thumbnailUrl'] = '/loras_static/images/no-preview.png'
+                # try to reconcile with a local model that has the same
                # filename and matching AutoV3 hash.
                reconciled = False
                file_name = lora_entry.get("file_name")
                if file_name and recipe_scanner and hash_value:
                    lora_scanner = getattr(recipe_scanner, "_lora_scanner", None)
                    if lora_scanner:
                        try:
                            # Local import to avoid circular dependency:
                            # base.py → file_utils → settings_manager → ...
                            #  → recipe_scanner → enrichment → base.py
                            from ..utils.file_utils import calculate_autov3  # fmt: skip
                            cache = await lora_scanner.get_cached_data()
                            for item in getattr(cache, "raw_data", []):
                                if item.get("file_name") == file_name:
                                    local_path = item.get("file_path")
                                    if local_path and os.path.exists(local_path):
                                        local_autov3 = calculate_autov3(local_path)
                                        if local_autov3 and local_autov3 == hash_value:
                                            lora_entry["existsLocally"] = True
                                            lora_entry["localPath"] = local_path
                                            lora_entry["hash"] = item.get("sha256", hash_value)
                                            if "preview_url" in item:
                                                lora_entry["thumbnailUrl"] = config.get_preview_static_url(item["preview_url"])
                                            civ = item.get("civitai") or {}
                                            if isinstance(civ, dict):
                                                if civ.get("id") is not None:
                                                    lora_entry["id"] = civ["id"]
                                                if civ.get("modelId") is not None:
                                                    lora_entry["modelId"] = civ["modelId"]
                                                if civ.get("name"):
                                                    lora_entry["version"] = civ["name"]
                                                # model_name is the CivitAI model display
                                                # name stored directly in the cache column.
                                                cached_model_name = item.get("model_name")
                                                if cached_model_name:
                                                    lora_entry["name"] = cached_model_name
                                            reconciled = True
                                            break
                        except Exception:
                            pass
                if not reconciled:
                    lora_entry['isDeleted'] = True
                    lora_entry['thumbnailUrl'] = '/loras_static/images/no-preview.png'
                return lora_entry
            # Get model type and validate
@@ -173,6 +216,20 @@ class RecipeMetadataParser(ABC):
                checkpoint['isDeleted'] = True
                return checkpoint
            # Validate that the model type is actually a checkpoint.
            # Unlike populate_lora_from_civitai which has this check,
            # this function was missing type validation — allowing LoRA
            # version data to be saved as the recipe's checkpoint when the
            # wrong version ID was passed downstream (fixed in v2.7+).
            model_type = civitai_data.get('model', {}).get('type', '').lower()
            if model_type not in VALID_CHECKPOINT_SUB_TYPES:
                logger.warning(
                    f"Cannot populate checkpoint: model version {civitai_data.get('id')} "
                    f"has type '{model_type}', expected one of {VALID_CHECKPOINT_SUB_TYPES}. "
                    f"Skipping checkpoint enrichment."
                )
                return checkpoint
            if 'model' in civitai_data and 'name' in civitai_data['model']:
                checkpoint['name'] = civitai_data['model']['name']
--- a/py/recipes/enrichment.py
+++ b/py/recipes/enrichment.py
@@ -16,55 +16,65 @@ class RecipeEnricher:
    async def enrich_recipe(
        recipe: Dict[str, Any],
        civitai_client: Any,
-        request_params: Optional[Dict[str, Any]] = None
+        request_params: Optional[Dict[str, Any]] = None,
        prefetched_civitai_meta_raw: Optional[Dict[str, Any]] = None,
        prefetched_model_version_id: Optional[int] = None,
    ) -> bool:
        """
        Enrich a recipe dictionary in-place with metadata from Civitai and embedded params.
-        
+
        Args:
            recipe: The recipe dictionary to enrich. Must have 'gen_params' initialized.
            civitai_client: Authenticated Civitai client instance.
            request_params: (Optional) Parameters from a user request (e.g. import).
-            
+            prefetched_civitai_meta_raw: (Optional) Pre-fetched raw meta from Civitai
                get_image_info, avoiding a duplicate API call.
            prefetched_model_version_id: (Optional) Pre-fetched model version ID.
        Returns:
            bool: True if the recipe was modified, False otherwise.
        """
        updated = False
        gen_params = recipe.get("gen_params", {})
-        
+
-        # 1. Fetch Civitai Info if available
+        # 1. Obtain Civitai metadata
        civitai_meta = None
-        model_version_id = None
+        model_version_id = prefetched_model_version_id
-        
+
-        source_url = recipe.get("source_url") or recipe.get("source_path", "")
+        source_path = recipe.get("source_path", "")
-        
+
-        # Check if it's a Civitai image URL
+        if prefetched_civitai_meta_raw is not None:
-        image_id = extract_civitai_image_id(str(source_url))
+            raw_meta = prefetched_civitai_meta_raw
-        if image_id:
+            if isinstance(raw_meta, dict):
-            try:
+                if "meta" in raw_meta and isinstance(raw_meta["meta"], dict):
-                image_info = await civitai_client.get_image_info(
+                    civitai_meta = raw_meta["meta"]
-                    image_id, source_url=str(source_url)
+                else:
-                )
+                    civitai_meta = raw_meta
-                if image_info:
+        else:
-                    # Handle nested meta often found in Civitai API responses
+            image_id = extract_civitai_image_id(str(source_path))
-                    raw_meta = image_info.get("meta")
+            if image_id:
-                    if isinstance(raw_meta, dict):
+                try:
-                        if "meta" in raw_meta and isinstance(raw_meta["meta"], dict):
+                    image_info = await civitai_client.get_image_info(
-                            civitai_meta = raw_meta["meta"]
+                        image_id, source_url=str(source_path)
-                        else:
+                    )
-                            civitai_meta = raw_meta
+                    if image_info:
-                    
+                        raw_meta = image_info.get("meta")
-                    model_version_id = image_info.get("modelVersionId")
+                        if isinstance(raw_meta, dict):
-                    
+                            if "meta" in raw_meta and isinstance(raw_meta["meta"], dict):
-                    # If not at top level, check resources in meta
+                                civitai_meta = raw_meta["meta"]
-                    if not model_version_id and civitai_meta:
+                            else:
-                        resources = civitai_meta.get("civitaiResources", [])
+                                civitai_meta = raw_meta
-                        for res in resources:
+
-                            if res.get("type") == "checkpoint":
+                        model_version_id = image_info.get("modelVersionId")
-                                model_version_id = res.get("modelVersionId")
+                except Exception as e:
-                                break
+                    logger.warning(f"Failed to fetch Civitai image info: {e}")
-            except Exception as e:
+
-                logger.warning(f"Failed to fetch Civitai image info: {e}")
+        if not model_version_id and civitai_meta:
            resources = civitai_meta.get("civitaiResources", [])
            for res in resources:
                if res.get("type") == "checkpoint":
                    model_version_id = res.get("modelVersionId")
                    break
        # 2. Merge Parameters
        # Priority: request_params > civitai_meta > embedded (existing gen_params)
@@ -180,27 +190,42 @@ class RecipeEnricher:
            existing_cp = recipe.get("checkpoint")
            if existing_cp is None:
                existing_cp = {}
            # Extract baseModel from raw civitai_info before populate_checkpoint_from_civitai
            # (populate may reject non-checkpoint types and lose this data)
            base_model_from_civitai: str = ""
            if isinstance(civitai_info, dict):
                base_model_from_civitai = civitai_info.get("baseModel", "") or ""
            elif isinstance(civitai_info, tuple) and len(civitai_info) > 0 and isinstance(civitai_info[0], dict):
                base_model_from_civitai = civitai_info[0].get("baseModel", "") or ""
            checkpoint_data = await RecipeMetadataParser.populate_checkpoint_from_civitai(existing_cp, civitai_info)
-            # 1. First, resolve base_model using full data before we format it away
+
            # 1. Resolve base_model from checkpoint_data first, then fall back to raw civitai_info
            current_base_model = recipe.get("base_model")
-            resolved_base_model = checkpoint_data.get("baseModel")
+            resolved_base_model = checkpoint_data.get("baseModel") or base_model_from_civitai
            if resolved_base_model:
                # Update if empty OR if it matches our generic prefix but is less specific
                is_generic = not current_base_model or current_base_model.lower() in ["flux", "sdxl", "sd15"]
                if is_generic and resolved_base_model != current_base_model:
                    recipe["base_model"] = resolved_base_model
-            
+
-            # 2. Format according to requirements: type, modelId, modelVersionId, modelName, modelVersionName
+            # 2. Only format and save checkpoint if it has real data (not just type after type rejection)
-            formatted_checkpoint = {
+            has_checkpoint_data = any([
-                "type": "checkpoint",
+                checkpoint_data.get("modelId"),
-                "modelId": checkpoint_data.get("modelId"),
+                checkpoint_data.get("id") or checkpoint_data.get("modelVersionId"),
-                "modelVersionId": checkpoint_data.get("id") or checkpoint_data.get("modelVersionId"),
+                checkpoint_data.get("name"),
-                "modelName": checkpoint_data.get("name"), # In base.py, 'name' is populated from civitai_data['model']['name']
+                checkpoint_data.get("version"),
-                "modelVersionName": checkpoint_data.get("version") # In base.py, 'version' is populated from civitai_data['name']
+            ])
-            }
+            if has_checkpoint_data:
-            # Remove None values
+                formatted_checkpoint = {
-            recipe["checkpoint"] = {k: v for k, v in formatted_checkpoint.items() if v is not None}
+                    "type": "checkpoint",
-            
+                    "modelId": checkpoint_data.get("modelId"),
                    "modelVersionId": checkpoint_data.get("id") or checkpoint_data.get("modelVersionId"),
                    "modelName": checkpoint_data.get("name"),
                    "modelVersionName": checkpoint_data.get("version"),
                }
                recipe["checkpoint"] = {k: v for k, v in formatted_checkpoint.items() if v is not None}
            return True
        else:
            # Fallback to name extraction if we don't already have one
--- a/py/recipes/parsers/civitai_image.py
+++ b/py/recipes/parsers/civitai_image.py
@@ -6,6 +6,7 @@ from typing import Dict, Any, Union
 from ..base import RecipeMetadataParser
 from ..constants import GEN_PARAM_KEYS
 from ...services.metadata_service import get_default_metadata_provider
 from ...config import config
 logger = logging.getLogger(__name__)
@@ -73,7 +74,8 @@ class CivitaiApiMetadataParser(RecipeMetadataParser):
        return False
    async def parse_metadata(  # type: ignore[override]
-        self, user_comment, recipe_scanner=None, civitai_client=None
+        self, user_comment, recipe_scanner=None, civitai_client=None,
        local_cache: dict[str, Any] | None = None,
    ) -> Dict[str, Any]:
        """Parse metadata from Civitai image format
@@ -81,6 +83,8 @@ class CivitaiApiMetadataParser(RecipeMetadataParser):
            user_comment: The metadata from the image (dict)
            recipe_scanner: Optional recipe scanner service
            civitai_client: Optional Civitai API client (deprecated, use metadata_provider instead)
            local_cache: Optional dict mapping sha256/autov3 hash → scanner cache item.
                         When provided, matching models skip CivitAI API calls.
        Returns:
            Dict containing parsed recipe data
@@ -185,8 +189,77 @@ class CivitaiApiMetadataParser(RecipeMetadataParser):
            # Process standard resources array
            if "resources" in metadata and isinstance(metadata["resources"], list):
                for resource in metadata["resources"]:
                    resource_type = resource.get("type", "lora")
                    # Track resources with type "model" — these are checkpoint models.
                    # The resources array is the most reliable source for checkpoint
                    # identification because it has an explicit type field and hash,
                    # unlike modelVersionIds which is a flat list with no type info.
                    if resource_type == "model":
                        checkpoint_entry = {
                            "id": 0,
                            "modelId": 0,
                            "name": resource.get("name", "Unknown Model"),
                            "version": "",
                            "type": resource.get("type", "model"),
                            "existsLocally": False,
                            "localPath": None,
                            "file_name": resource.get("name", ""),
                            "hash": resource.get("hash", "") or "",
                            "thumbnailUrl": "/loras_static/images/no-preview.png",
                            "baseModel": "",
                            "size": 0,
                            "downloadUrl": "",
                            "isDeleted": False,
                        }
                        # Try to look up base model from the checkpoint hash
                        cp_hash = checkpoint_entry.get("hash")
                        if cp_hash and metadata_provider:
                            local_cached = local_cache.get(cp_hash) if local_cache else None
                            if local_cached:
                                self._populate_entry_from_cache(
                                    checkpoint_entry, local_cached
                                )
                                bm = checkpoint_entry.get("baseModel", "")
                                if bm and not result["base_model"]:
                                    result["base_model"] = bm
                            else:
                                try:
                                    civitai_info = (
                                        await metadata_provider.get_model_by_hash(
                                            cp_hash
                                        )
                                    )
                                    civitai_data, error_msg = (
                                        (civitai_info, None)
                                        if not isinstance(civitai_info, tuple)
                                        else civitai_info
                                    )
                                    if civitai_data and error_msg != "Model not found":
                                        if 'model' in civitai_data and 'name' in civitai_data['model']:
                                            checkpoint_entry['name'] = civitai_data['model']['name']
                                        checkpoint_entry['id'] = civitai_data.get('id', 0)
                                        checkpoint_entry['modelId'] = civitai_data.get('modelId', 0)
                                        if 'name' in civitai_data:
                                            checkpoint_entry['version'] = civitai_data['name']
                                        base_model = civitai_data.get('baseModel', '')
                                        if base_model:
                                            checkpoint_entry['baseModel'] = base_model
                                            if not result['base_model']:
                                                result['base_model'] = base_model
                                except Exception as e:
                                    logger.error(
                                        f"Error fetching checkpoint info for hash "
                                        f"{cp_hash}: {e}"
                                    )
                        if result["model"] is None:
                            result["model"] = checkpoint_entry
                        continue
                    # Modified to process resources without a type field as potential LoRAs
-                    if resource.get("type", "lora") == "lora":
+                    if resource_type == "lora":
                        lora_hash = resource.get("hash", "")
                        # Try to get hash from the hashes field if not present in resource
@@ -220,34 +293,45 @@ class CivitaiApiMetadataParser(RecipeMetadataParser):
                        }
                        # Try to get info from Civitai if hash is available
-                        if lora_entry["hash"] and metadata_provider:
+                        if lora_hash and metadata_provider:
-                            try:
+                            local_cached = local_cache.get(lora_hash) if local_cache else None
-                                civitai_info = (
+                            if local_cached:
-                                    await metadata_provider.get_model_by_hash(lora_hash)
+                                self._populate_entry_from_cache(
                                    lora_entry, local_cached
                                )
-
+                                # Track by version ID for deduplication
-                                populated_entry = await self.populate_lora_from_civitai(
+                                if lora_entry.get("id"):
                                    lora_entry,
                                    civitai_info,
                                    recipe_scanner,
                                    base_model_counts,
                                    lora_hash,
                                )
                                if populated_entry is None:
                                    continue  # Skip invalid LoRA types
                                lora_entry = populated_entry
                                # If we have a version ID from Civitai, track it for deduplication
                                if "id" in lora_entry and lora_entry["id"]:
                                    added_loras[str(lora_entry["id"])] = len(
                                        result["loras"]
                                    )
-                            except Exception as e:
+                            else:
-                                logger.error(
+                                try:
-                                    f"Error fetching Civitai info for LoRA hash {lora_entry['hash']}: {e}"
+                                    civitai_info = (
-                                )
+                                        await metadata_provider.get_model_by_hash(lora_hash)
                                    )
                                    populated_entry = await self.populate_lora_from_civitai(
                                        lora_entry,
                                        civitai_info,
                                        recipe_scanner,
                                        base_model_counts,
                                        lora_hash,
                                    )
                                    if populated_entry is None:
                                        continue  # Skip invalid LoRA types
                                    lora_entry = populated_entry
                                    # If we have a version ID from Civitai, track it for deduplication
                                    if "id" in lora_entry and lora_entry["id"]:
                                        added_loras[str(lora_entry["id"])] = len(
                                            result["loras"]
                                        )
                                except Exception as e:
                                    logger.error(
                                        f"Error fetching Civitai info for LoRA hash {lora_entry['hash']}: {e}"
                                    )
                        # Track by hash if we have it
                        if lora_hash:
@@ -625,3 +709,41 @@ class CivitaiApiMetadataParser(RecipeMetadataParser):
        except Exception as e:
            logger.error(f"Error parsing Civitai image metadata: {e}", exc_info=True)
            return {"error": str(e), "loras": []}
    @staticmethod
    def _populate_entry_from_cache(
        entry: dict[str, Any],
        cache_item: dict[str, Any],
    ) -> None:
        """Fill a lora/checkpoint entry from a scanner cache item.
        Avoids CivitAI API calls for models that exist locally.
        Mirrors the population logic in
        ``RecipeMetadataParser.populate_lora_from_civitai()`` but operates
        entirely on cached data.
        """
        civ = cache_item.get("civitai") or {}
        if isinstance(civ, dict):
            if civ.get("id") is not None:
                entry["id"] = civ["id"]
            if civ.get("modelId") is not None:
                entry["modelId"] = civ["modelId"]
            if civ.get("name"):
                entry["version"] = civ["name"]
            cached_name = cache_item.get("model_name")
            if cached_name:
                entry["name"] = cached_name
        entry["existsLocally"] = True
        local_path = cache_item.get("file_path")
        if local_path:
            entry["localPath"] = local_path
        sha256 = cache_item.get("sha256")
        if sha256:
            entry["hash"] = sha256
        if "preview_url" in cache_item:
            entry["thumbnailUrl"] = config.get_preview_static_url(
                cache_item["preview_url"]
            )
        base_model = cache_item.get("base_model", "")
        if base_model:
            entry["baseModel"] = base_model
--- a/py/routes/handlers/misc_handlers.py
+++ b/py/routes/handlers/misc_handlers.py
@@ -33,6 +33,7 @@ from ...services.metadata_service import (
    update_metadata_providers,
 )
 from ...services.service_registry import ServiceRegistry
 from ...services.model_lifecycle_service import delete_model_artifacts
 from ...services.settings_manager import get_settings_manager
 from ...services.websocket_manager import ws_manager
 from ...services.downloader import get_downloader
@@ -685,6 +686,9 @@ class DoctorHandler:
        )
    async def resolve_filename_conflicts(self, request: web.Request) -> web.Response:
        if self._settings.get("lora_syntax_format", "legacy") == "full":
            return web.json_response({"success": True, "renamed": [], "count": 0})
        renamed: list[dict[str, Any]] = []
        try:
@@ -989,11 +993,29 @@ class DoctorHandler:
        }
    async def _check_filename_conflicts(self) -> dict[str, Any]:
        # When full path syntax is active, duplicate filenames across subfolders
        # are not ambiguous (<lora:subfolder/name:strength>), so skip the check.
        if self._settings.get("lora_syntax_format", "legacy") == "full":
            return {
                "id": "filename_conflicts",
                "title": "Duplicate Filename Conflicts",
                "status": "ok",
                "summary": "Full path syntax is active — duplicate filenames across folders are not ambiguous.",
                "details": [],
                "actions": [],
            }
        all_conflicts: list[dict[str, Any]] = []
        total_conflict_groups = 0
        total_conflict_files = 0
        for model_type, label, factory in self._scanner_factories:
            # Duplicate filename detection targets LoRAs which use basename-only
            # syntax (<lora:name:strength>). Checkpoints/embeddings reference
            # models via relative paths with extensions, so conflicts there would
            # be false positives.
            if model_type != "lora":
                continue
            try:
                scanner = await factory()
                hash_index = getattr(scanner, "_hash_index", None)
@@ -1041,12 +1063,22 @@ class DoctorHandler:
                "total_conflict_files": total_conflict_files,
            }
        ]
-        for conflict in all_conflicts:
+
        # Show at most 5 conflict groups inline; note any remainder.
        MAX_VISIBLE_CONFLICTS = 5
        visible_conflicts = all_conflicts[:MAX_VISIBLE_CONFLICTS]
        for conflict in visible_conflicts:
            details.append(
-                f"[{conflict['label']}] '{conflict['filename']}' "
+                f"'{conflict['filename']}' "
                f"found in {len(conflict['paths'])} locations"
            )
        hidden_count = len(all_conflicts) - MAX_VISIBLE_CONFLICTS
        if hidden_count > 0:
            details.append(
                f"...and {hidden_count} more duplicate filename group(s)"
            )
        return {
            "id": "filename_conflicts",
            "title": "Duplicate Filename Conflicts",
@@ -1057,7 +1089,11 @@ class DoctorHandler:
                {
                    "id": "resolve-filename-conflicts",
                    "label": "Resolve Conflicts",
-                }
+                },
                {
                    "id": "open-settings-syntax-format",
                    "label": "Switch to Full Path Syntax",
                },
            ],
        }
@@ -1791,29 +1827,33 @@ class ModelLibraryHandler:
                    exists = True
                    model_type = "embedding"
                if exists:
                    return web.json_response(
                        {
                            "success": True,
                            "exists": True,
                            "modelType": model_type,
                            "hasBeenDownloaded": False,
                        }
                    )
                history_service = await self._get_download_history_service()
                has_been_downloaded = False
-                history_type = model_type
+                history_type = None
-                if history_type:
+                for candidate_type in ("lora", "checkpoint", "embedding"):
-                    has_been_downloaded = await history_service.has_been_downloaded(
+                    if await history_service.has_been_downloaded(
-                        history_type,
+                        candidate_type,
                        model_version_id,
-                    )
+                    ):
-                else:
+                        has_been_downloaded = True
-                    for candidate_type in ("lora", "checkpoint", "embedding"):
+                        history_type = candidate_type
-                        if await history_service.has_been_downloaded(
+                        break
                            candidate_type,
                            model_version_id,
                        ):
                            has_been_downloaded = True
                            history_type = candidate_type
                            break
                return web.json_response(
                    {
                        "success": True,
-                        "exists": exists,
+                        "exists": False,
-                        "modelType": model_type if exists else history_type,
+                        "modelType": history_type,
                        "hasBeenDownloaded": has_been_downloaded,
                    }
                )
@@ -1833,40 +1873,46 @@ class ModelLibraryHandler:
            model_type = None
            versions = []
            downloaded_version_ids = []
            history_service = await self._get_download_history_service()
            if lora_versions:
-                model_type = "lora"
+                return web.json_response(
-                versions = self._with_downloaded_flag(lora_versions)
+                    {
-                downloaded_version_ids = await history_service.get_downloaded_version_ids(
+                        "success": True,
-                    model_type,
+                        "modelType": "lora",
-                    model_id,
+                        "versions": self._with_downloaded_flag(lora_versions),
                        "downloadedVersionIds": [],
                    }
                )
-            elif checkpoint_versions:
+            if checkpoint_versions:
-                model_type = "checkpoint"
+                return web.json_response(
-                versions = self._with_downloaded_flag(checkpoint_versions)
+                    {
-                downloaded_version_ids = await history_service.get_downloaded_version_ids(
+                        "success": True,
-                    model_type,
+                        "modelType": "checkpoint",
-                    model_id,
+                        "versions": self._with_downloaded_flag(checkpoint_versions),
                        "downloadedVersionIds": [],
                    }
                )
-            elif embedding_versions:
+            if embedding_versions:
-                model_type = "embedding"
+                return web.json_response(
-                versions = self._with_downloaded_flag(embedding_versions)
+                    {
-                downloaded_version_ids = await history_service.get_downloaded_version_ids(
+                        "success": True,
-                    model_type,
+                        "modelType": "embedding",
-                    model_id,
+                        "versions": self._with_downloaded_flag(embedding_versions),
                        "downloadedVersionIds": [],
                    }
                )
-            else:
+
-                for candidate_type in ("lora", "checkpoint", "embedding"):
+            history_service = await self._get_download_history_service()
-                    candidate_downloaded_version_ids = (
+            for candidate_type in ("lora", "checkpoint", "embedding"):
-                        await history_service.get_downloaded_version_ids(
+                candidate_downloaded_version_ids = (
-                            candidate_type,
+                    await history_service.get_downloaded_version_ids(
-                            model_id,
+                        candidate_type,
-                        )
+                        model_id,
                    )
-                    if candidate_downloaded_version_ids:
+                )
-                        model_type = candidate_type
+                if candidate_downloaded_version_ids:
-                        downloaded_version_ids = candidate_downloaded_version_ids
+                    model_type = candidate_type
-                        break
+                    downloaded_version_ids = candidate_downloaded_version_ids
                    break
            return web.json_response(
                {
@@ -1880,6 +1926,86 @@ class ModelLibraryHandler:
            logger.error("Failed to check model existence: %s", exc, exc_info=True)
            return web.json_response({"success": False, "error": str(exc)}, status=500)
    async def check_models_exist(self, request: web.Request) -> web.Response:
        try:
            model_ids_raw = request.query.get("modelIds", "")
            if not model_ids_raw:
                return web.json_response(
                    {"success": True, "results": []}
                )
            raw_ids = model_ids_raw.split(",")
            seen: set[int] = set()
            model_ids: list[int] = []
            for raw in raw_ids:
                stripped = raw.strip()
                if not stripped:
                    continue
                try:
                    mid = int(stripped)
                except ValueError:
                    continue
                if mid not in seen:
                    seen.add(mid)
                    model_ids.append(mid)
            if not model_ids:
                return web.json_response(
                    {"success": True, "results": []}
                )
            lora_scanner = await self._service_registry.get_lora_scanner()
            checkpoint_scanner = await self._service_registry.get_checkpoint_scanner()
            embedding_scanner = await self._service_registry.get_embedding_scanner()
            results: list[dict] = []
            for model_id in model_ids:
                lora_versions = await lora_scanner.get_model_versions_by_id(model_id)
                if lora_versions:
                    results.append({
                        "modelId": model_id,
                        "modelType": "lora",
                        "versions": self._with_downloaded_flag(lora_versions),
                        "downloadedVersionIds": [],
                    })
                    continue
                if checkpoint_scanner:
                    checkpoint_versions = await checkpoint_scanner.get_model_versions_by_id(model_id)
                    if checkpoint_versions:
                        results.append({
                            "modelId": model_id,
                            "modelType": "checkpoint",
                            "versions": self._with_downloaded_flag(checkpoint_versions),
                            "downloadedVersionIds": [],
                        })
                        continue
                if embedding_scanner:
                    embedding_versions = await embedding_scanner.get_model_versions_by_id(model_id)
                    if embedding_versions:
                        results.append({
                            "modelId": model_id,
                            "modelType": "embedding",
                            "versions": self._with_downloaded_flag(embedding_versions),
                            "downloadedVersionIds": [],
                        })
                        continue
                results.append({
                    "modelId": model_id,
                    "modelType": None,
                    "versions": [],
                    "downloadedVersionIds": [],
                })
            return web.json_response(
                {"success": True, "results": results}
            )
        except Exception as exc:
            logger.error("Failed to check models existence: %s", exc, exc_info=True)
            return web.json_response({"success": False, "error": str(exc)}, status=500)
    async def get_model_version_download_status(
        self, request: web.Request
    ) -> web.Response:
@@ -1974,7 +2100,7 @@ class ModelLibraryHandler:
                    file_path=file_path if isinstance(file_path, str) else None,
                )
            else:
-                await history_service.mark_not_downloaded(model_type, model_version_id)
+                await history_service.mark_as_deleted(model_type, model_version_id)
            return web.json_response(
                {
@@ -1992,6 +2118,89 @@ class ModelLibraryHandler:
            )
            return web.json_response({"success": False, "error": str(exc)}, status=500)
    async def delete_model_version(self, request: web.Request) -> web.Response:
        try:
            model_version_id_str = request.query.get("modelVersionId")
            if not model_version_id_str:
                return web.json_response(
                    {"success": False, "error": "Missing required parameter: modelVersionId"},
                    status=400,
                )
            try:
                model_version_id = int(model_version_id_str)
            except ValueError:
                return web.json_response(
                    {"success": False, "error": "Parameter modelVersionId must be an integer"},
                    status=400,
                )
            lora_scanner = await self._service_registry.get_lora_scanner()
            checkpoint_scanner = await self._service_registry.get_checkpoint_scanner()
            embedding_scanner = await self._service_registry.get_embedding_scanner()
            found_type = None
            file_path = None
            found_cache = None
            for model_type, scanner in (
                ("lora", lora_scanner),
                ("checkpoint", checkpoint_scanner),
                ("embedding", embedding_scanner),
            ):
                cache = await scanner.get_cached_data()
                if cache and model_version_id in cache.version_index:
                    found_type = model_type
                    found_cache = cache
                    entry = cache.version_index[model_version_id]
                    file_path = entry.get("file_path")
                    break
            if not file_path:
                return web.json_response(
                    {"success": False, "error": "Model version not found in any scanner cache"},
                    status=404,
                )
            target_dir = os.path.dirname(file_path)
            base_name = os.path.basename(file_path)
            file_name, extension = os.path.splitext(base_name)
            await delete_model_artifacts(target_dir, file_name, main_extension=extension)
            if found_cache:
                found_cache.raw_data = [
                    item
                    for item in found_cache.raw_data
                    if item.get("file_path") != file_path
                ]
                await found_cache.resort()
            scanner_map = {
                "lora": lora_scanner,
                "checkpoint": checkpoint_scanner,
                "embedding": embedding_scanner,
            }
            scanner = scanner_map.get(found_type)
            if scanner:
                persist = getattr(scanner, "_persist_current_cache", None)
                if callable(persist):
                    await persist()
            history_service = await self._get_download_history_service()
            await history_service.mark_as_deleted(found_type, model_version_id)
            return web.json_response(
                {
                    "success": True,
                    "modelType": found_type,
                    "modelVersionId": model_version_id,
                }
            )
        except Exception as exc:
            logger.error(
                "Failed to delete model version: %s", exc, exc_info=True
            )
            return web.json_response({"success": False, "error": str(exc)}, status=500)
    async def get_model_versions_status(self, request: web.Request) -> web.Response:
        try:
            model_id_str = request.query.get("modelId")
@@ -3025,8 +3234,10 @@ class MiscHandlerSet:
            "update_node_widget": self.node_registry.update_node_widget,
            "get_registry": self.node_registry.get_registry,
            "check_model_exists": self.model_library.check_model_exists,
            "check_models_exist": self.model_library.check_models_exist,
            "get_model_version_download_status": self.model_library.get_model_version_download_status,
            "set_model_version_download_status": self.model_library.set_model_version_download_status,
            "delete_model_version": self.model_library.delete_model_version,
            "get_civitai_user_models": self.model_library.get_civitai_user_models,
            "download_metadata_archive": self.metadata_archive.download_metadata_archive,
            "remove_metadata_archive": self.metadata_archive.remove_metadata_archive,
--- a/py/routes/handlers/model_handlers.py
+++ b/py/routes/handlers/model_handlers.py
@@ -301,6 +301,15 @@ class ModelListingHandler:
        for tag in exclude_tags:
            if tag:
                tag_filters[tag] = "exclude"
        auto_tag_filters: Dict[str, str] = {}
        for tag in request.query.getall("auto_tag_include", []):
            if tag:
                auto_tag_filters[tag] = "include"
        for tag in request.query.getall("auto_tag_exclude", []):
            if tag:
                auto_tag_filters[tag] = "exclude"
        favorites_only = request.query.get("favorites_only", "false").lower() == "true"
        search_options = {
@@ -367,6 +376,7 @@ class ModelListingHandler:
            "fuzzy_search": fuzzy_search,
            "base_models": base_models,
            "tags": tag_filters,
            "auto_tags": auto_tag_filters,
            "tag_logic": tag_logic,
            "search_options": search_options,
            "hash_filters": hash_filters,
@@ -778,7 +788,7 @@ class ModelManagementHandler:
            metadata_updates = {k: v for k, v in data.items() if k != "file_path"}
-            await self._metadata_sync.save_metadata_updates(
+            updated_metadata = await self._metadata_sync.save_metadata_updates(
                file_path=file_path,
                updates=metadata_updates,
                metadata_loader=self._metadata_sync.load_local_metadata,
@@ -789,7 +799,12 @@ class ModelManagementHandler:
                cache = await self._service.scanner.get_cached_data()
                await cache.resort()
-            return web.json_response({"success": True})
+            from ...services.auto_tag_service import extract_auto_tags
            auto_tags = extract_auto_tags(updated_metadata)
            return web.json_response(
                {"success": True, "auto_tags": auto_tags}
            )
        except Exception as exc:
            self._logger.error("Error saving metadata: %s", exc, exc_info=True)
            return web.Response(text=str(exc), status=500)
@@ -806,14 +821,16 @@ class ModelManagementHandler:
            if not isinstance(new_tags, list):
                return web.Response(text="Tags must be a list", status=400)
-            tags = await self._tag_update_service.add_tags(
+            tags, auto_tags = await self._tag_update_service.add_tags(
                file_path=file_path,
                new_tags=new_tags,
                metadata_loader=self._metadata_sync.load_local_metadata,
                update_cache=self._service.scanner.update_single_model_cache,
            )
-            return web.json_response({"success": True, "tags": tags})
+            return web.json_response(
                {"success": True, "tags": tags, "auto_tags": auto_tags}
            )
        except Exception as exc:
            self._logger.error("Error adding tags: %s", exc, exc_info=True)
            return web.Response(text=str(exc), status=500)
@@ -1160,6 +1177,12 @@ class ModelQueryHandler:
    async def find_filename_conflicts(self, request: web.Request) -> web.Response:
        try:
            settings = get_settings_manager()
            if settings.get("lora_syntax_format", "legacy") == "full":
                return web.json_response(
                    {"success": True, "conflicts": [], "count": 0}
                )
            duplicates = self._service.find_duplicate_filenames()
            result = []
            cache = await self._service.scanner.get_cached_data()
@@ -1449,6 +1472,21 @@ class ModelDownloadHandler:
            )
            return web.Response(status=500, text=str(exc))
    async def skip_download_get(self, request: web.Request) -> web.Response:
        try:
            download_id = request.query.get("download_id")
            if not download_id:
                return web.json_response(
                    {"success": False, "error": "Download ID is required"}, status=400
                )
            result = await self._download_coordinator.skip_download(download_id)
            return web.json_response(result)
        except Exception as exc:
            self._logger.error(
                "Error skipping download via GET: %s", exc, exc_info=True
            )
            return web.json_response({"success": False, "error": str(exc)}, status=500)
    async def cancel_download_get(self, request: web.Request) -> web.Response:
        try:
            download_id = request.query.get("download_id")
@@ -1937,6 +1975,10 @@ class ModelUpdateHandler:
        if target_model_ids:
            target_model_ids = sorted(set(target_model_ids))
        folder_path: Optional[str] = payload.get("folder_path")
        if folder_path is not None and not isinstance(folder_path, str):
            folder_path = None
        provider = await self._get_civitai_provider()
        if provider is None:
            return web.json_response(
@@ -1951,6 +1993,7 @@ class ModelUpdateHandler:
                provider,
                force_refresh=force_refresh,
                target_model_ids=target_model_ids or None,
                folder_path=folder_path,
            )
            if self._service.scanner.is_cancelled():
                return web.json_response(
@@ -1973,10 +2016,21 @@ class ModelUpdateHandler:
            self._logger.error("Failed to refresh model updates: %s", exc, exc_info=True)
            return web.json_response({"success": False, "error": str(exc)}, status=500)
        hide_early_access = False
        if self._settings is not None:
            try:
                hide_early_access = bool(
                    self._settings.get("hide_early_access_updates", False)
                )
            except Exception:
                pass
        serialized_records = []
        for record in records.values():
            has_update_fn = getattr(record, "has_update", None)
-            if callable(has_update_fn) and has_update_fn():
+            if callable(has_update_fn) and has_update_fn(
                hide_early_access=hide_early_access
            ):
                serialized_records.append(self._serialize_record(record))
        return web.json_response(
@@ -2423,6 +2477,7 @@ class ModelUpdateHandler:
            "shouldIgnore": version.should_ignore,
            "earlyAccessEndsAt": version.early_access_ends_at,
            "isEarlyAccess": is_early_access,
            "usageControl": version.usage_control,
            "filePath": context.get("file_path"),
            "fileName": context.get("file_name"),
        }
@@ -2537,6 +2592,7 @@ class ModelHandlerSet:
            "download_model": self.download.download_model,
            "download_model_get": self.download.download_model_get,
            "cancel_download_get": self.download.cancel_download_get,
            "skip_download_get": self.download.skip_download_get,
            "pause_download_get": self.download.pause_download_get,
            "resume_download_get": self.download.resume_download_get,
            "get_download_progress": self.download.get_download_progress,
--- a/py/routes/handlers/preview_handlers.py
+++ b/py/routes/handlers/preview_handlers.py
@@ -3,6 +3,7 @@
 from __future__ import annotations
 import logging
 import mimetypes
 import urllib.parse
 from pathlib import Path
@@ -12,6 +13,12 @@ from ...config import config as global_config
 logger = logging.getLogger(__name__)
 _CHUNK_SIZE = 256 * 1024  # 256 KB
 # Video file extensions that bypass native sendfile on Windows
 # to avoid IOCP/ProactorEventLoop crashes during client disconnect.
 _VIDEO_EXTENSIONS = frozenset({".mp4", ".webm", ".mov", ".avi", ".mkv"})
 class PreviewHandler:
    """Serve preview assets for the active library at request time."""
@@ -48,8 +55,51 @@ class PreviewHandler:
            logger.debug("Preview file not found at %s", str(resolved))
            raise web.HTTPNotFound(text="Preview file not found")
        # Video files: stream manually to avoid Windows native sendfile crash.
        # aiohttp's FileResponse uses _sendfile_native on Windows (IOCP-based),
        # which breaks when the client disconnects mid-transfer — this happens
        # constantly when users scroll through a gallery of animated previews.
        suffix = resolved.suffix.lower()
        if suffix in _VIDEO_EXTENSIONS:
            return await self._stream_file(request, resolved)
        # aiohttp's FileResponse handles range requests and content headers for us.
-        return web.FileResponse(path=resolved, chunk_size=256 * 1024)
+        return web.FileResponse(path=resolved, chunk_size=_CHUNK_SIZE)
    async def _stream_file(
        self, request: web.Request, path: Path
    ) -> web.StreamResponse:
        """Stream a file chunk-by-chunk, bypassing native sendfile.
        This avoids the Windows IOCP ``_sendfile_native`` crash that occurs
        when the client disconnects during a large file transfer.
        """
        content_type, _ = mimetypes.guess_type(str(path))
        if content_type is None:
            content_type = "application/octet-stream"
        file_size = path.stat().st_size
        resp = web.StreamResponse()
        resp.content_type = content_type
        resp.content_length = file_size
        await resp.prepare(request)
        try:
            with open(path, "rb") as f:
                while True:
                    chunk = f.read(_CHUNK_SIZE)
                    if not chunk:
                        break
                    await resp.write(chunk)
        except (ConnectionResetError, ConnectionAbortedError):
            # Client disconnected during streaming — expected when scrolling
            # rapidly through a library with animated previews.
            pass
        except OSError as exc:
            logger.debug("I/O error streaming preview %s: %s", path, exc)
        return resp
 __all__ = ["PreviewHandler"]
--- a/py/routes/handlers/recipe_handlers.py
+++ b/py/routes/handlers/recipe_handlers.py
--- a/py/routes/misc_route_registrar.py
+++ b/py/routes/misc_route_registrar.py
@@ -43,6 +43,7 @@ MISC_ROUTE_DEFINITIONS: tuple[RouteDefinition, ...] = (
    RouteDefinition("POST", "/api/lm/update-node-widget", "update_node_widget"),
    RouteDefinition("GET", "/api/lm/get-registry", "get_registry"),
    RouteDefinition("GET", "/api/lm/check-model-exists", "check_model_exists"),
    RouteDefinition("GET", "/api/lm/check-models-exist", "check_models_exist"),
    RouteDefinition(
        "GET",
        "/api/lm/model-version-download-status",
@@ -90,6 +91,9 @@ MISC_ROUTE_DEFINITIONS: tuple[RouteDefinition, ...] = (
    RouteDefinition(
        "GET", "/api/lm/base-models/cache-status", "get_base_model_cache_status"
    ),
    RouteDefinition(
        "GET", "/api/lm/delete-model-version", "delete_model_version"
    ),
 )
--- a/py/routes/model_route_registrar.py
+++ b/py/routes/model_route_registrar.py
@@ -101,6 +101,7 @@ COMMON_ROUTE_DEFINITIONS: tuple[RouteDefinition, ...] = (
    RouteDefinition("POST", "/api/lm/download-model", "download_model"),
    RouteDefinition("GET", "/api/lm/download-model-get", "download_model_get"),
    RouteDefinition("GET", "/api/lm/cancel-download-get", "cancel_download_get"),
    RouteDefinition("GET", "/api/lm/skip-download", "skip_download_get"),
    RouteDefinition("GET", "/api/lm/pause-download", "pause_download_get"),
    RouteDefinition("GET", "/api/lm/resume-download", "resume_download_get"),
    RouteDefinition(
--- a/py/routes/recipe_route_registrar.py
+++ b/py/routes/recipe_route_registrar.py
@@ -58,6 +58,7 @@ ROUTE_DEFINITIONS: tuple[RouteDefinition, ...] = (
    RouteDefinition("POST", "/api/lm/recipes/repair", "repair_recipes"),
    RouteDefinition("POST", "/api/lm/recipes/cancel-repair", "cancel_repair"),
    RouteDefinition("POST", "/api/lm/recipe/{recipe_id}/repair", "repair_recipe"),
    RouteDefinition("POST", "/api/lm/recipes/repair-bulk", "repair_recipes_bulk"),
    RouteDefinition("GET", "/api/lm/recipes/repair-progress", "get_repair_progress"),
    RouteDefinition("POST", "/api/lm/recipes/batch-import/start", "start_batch_import"),
    RouteDefinition(
@@ -70,6 +71,13 @@ ROUTE_DEFINITIONS: tuple[RouteDefinition, ...] = (
        "POST", "/api/lm/recipes/batch-import/directory", "start_directory_import"
    ),
    RouteDefinition("POST", "/api/lm/recipes/browse-directory", "browse_directory"),
    RouteDefinition(
        "GET", "/api/lm/recipes/check-image-exists", "check_image_exists"
    ),
    RouteDefinition("GET", "/api/lm/recipes/import-from-url", "import_from_url"),
    RouteDefinition(
        "POST", "/api/lm/recipes/create-from-example", "create_from_example"
    ),
 )
--- a/py/routes/update_routes.py
+++ b/py/routes/update_routes.py
@@ -11,6 +11,7 @@ from typing import Dict, List
 from ..utils.settings_paths import ensure_settings_file
 from ..services.downloader import get_downloader
 from ..services.service_registry import ServiceRegistry
 logger = logging.getLogger(__name__)
@@ -212,8 +213,19 @@ class UpdateRoutes:
            zip_path = tmp_zip_path
-            # Skip both settings.json, civitai and model cache folder
+            # Close the downloaded-versions SQLite connection before cleaning,
-            UpdateRoutes._clean_plugin_folder(plugin_root, skip_files=['settings.json', 'civitai', 'model_cache'])
+            # so that shutil.rmtree() does not fail on Windows (the process
            # cannot delete a file with an outstanding open handle).
            try:
                history_svc = ServiceRegistry._services.get("downloaded_version_history_service")
                if history_svc is not None:
                    history_svc.close()
                    logger.info("Closed downloaded-version history database connection")
            except Exception:
                logger.debug("Could not close downloaded-version history database", exc_info=True)
            # Skip settings.json, civitai, model cache and runtime cache folders
            UpdateRoutes._clean_plugin_folder(plugin_root, skip_files=['settings.json', 'civitai', 'model_cache', 'cache', 'wildcards', 'backups'])
            # Extract ZIP to temp dir
            with tempfile.TemporaryDirectory() as tmp_dir:
@@ -222,16 +234,17 @@ class UpdateRoutes:
                    # Find extracted folder (GitHub ZIP contains a root folder)
                    extracted_root = next(os.scandir(tmp_dir)).path
-                    # Copy files, skipping settings.json and civitai folder
+                    # Copy files, skipping user data that should be preserved
                    skip_items = {'settings.json', 'civitai', 'wildcards', 'backups'}
                    for item in os.listdir(extracted_root):
-                        if item == 'settings.json' or item == 'civitai':
+                        if item in skip_items:
                            continue
                        src = os.path.join(extracted_root, item)
                        dst = os.path.join(plugin_root, item)
                        if os.path.isdir(src):
                            if os.path.exists(dst):
                                shutil.rmtree(dst)
-                            shutil.copytree(src, dst, ignore=shutil.ignore_patterns('settings.json', 'civitai'))
+                            shutil.copytree(src, dst, ignore=shutil.ignore_patterns(*skip_items))
                        else:
                            shutil.copy2(src, dst)
@@ -239,15 +252,17 @@ class UpdateRoutes:
                    # for ComfyUI Manager to work properly
                    tracking_info_file = os.path.join(plugin_root, '.tracking')
                    tracking_files = []
                    skip_tracked = {'civitai', 'wildcards', 'backups'}
                    for root, dirs, files in os.walk(extracted_root):
-                        # Skip civitai folder and its contents
+                        # Skip user data directories and their contents
                        rel_root = os.path.relpath(root, extracted_root)
-                        if rel_root == 'civitai' or rel_root.startswith('civitai' + os.sep):
+                        top_dir = rel_root.split(os.sep)[0] if rel_root != '.' else ''
                        if top_dir in skip_tracked:
                            continue
                        for file in files:
                            rel_path = os.path.relpath(os.path.join(root, file), extracted_root)
-                            # Skip settings.json and any file under civitai
+                            # Skip settings.json and any file under user data dirs
-                            if rel_path == 'settings.json' or rel_path.startswith('civitai' + os.sep):
+                            if rel_path == 'settings.json' or rel_path.split(os.sep)[0] in skip_tracked:
                                continue
                            tracking_files.append(rel_path.replace("\\", "/"))
                    with open(tracking_info_file, "w", encoding='utf-8') as file:
--- a/py/services/aria2_downloader.py
+++ b/py/services/aria2_downloader.py
@@ -14,12 +14,30 @@ from typing import Any, Dict, Optional, Tuple
 import aiohttp
-from .downloader import DownloadProgress, get_downloader
+from .downloader import DownloadProgress, get_downloader, is_ssl_cert_verify_error
 from .aria2_transfer_state import Aria2TransferStateStore
 from .settings_manager import get_settings_manager
 logger = logging.getLogger(__name__)
 def _try_certifi_ca_path() -> str | None:
    """Return the certifi CA bundle path if available, else None."""
    try:
        import certifi  # type: ignore[import-untyped]
        path = certifi.where()
        if os.path.isfile(path):
            logger.debug(
                "aria2 --ca-certificate: using certifi CA bundle at %s", path
            )
            return path
    except ImportError:
        pass
    logger.debug("aria2 --ca-certificate: certifi not available")
    return None
 CIVITAI_DOWNLOAD_URL_PREFIXES = (
    "https://civitai.com/api/download/",
    "https://civitai.red/api/download/",
@@ -39,7 +57,7 @@ class Aria2Transfer:
 class Aria2Downloader:
-    """Manage an aria2 RPC daemon for experimental model downloads."""
+    """Manage an aria2 RPC daemon for recommended model downloads."""
    _instance = None
    _lock = asyncio.Lock()
@@ -391,6 +409,15 @@ class Aria2Downloader:
                    f"Failed to resolve authenticated Civitai redirect: status={response.status} body={body[:300]}"
                )
        except aiohttp.ClientError as exc:
            if is_ssl_cert_verify_error(exc):
                logger.error(
                    "SSL certificate verification failed during Civitai redirect "
                    "resolution for %s. This is usually caused by an outdated CA "
                    "certificate bundle. Recommended fixes:\n"
                    "  1. pip install --upgrade certifi\n"
                    "  2. pip install pip-system-certs",
                    url,
                )
            raise Aria2Error(
                f"Failed to resolve authenticated Civitai redirect: {exc}"
            ) from exc
@@ -414,6 +441,11 @@ class Aria2Downloader:
                f"--rpc-listen-port={self._rpc_port}",
                f"--rpc-secret={self._rpc_secret}",
                "--check-certificate=true",
                # Point aria2 at certifi's CA bundle when available so it uses
                # the same certificate store as Python downloads.
                *((
                    f"--ca-certificate={ca_cert}",
                ) if (ca_cert := _try_certifi_ca_path()) else ()),
                "--allow-overwrite=true",
                "--auto-file-renaming=false",
                "--file-allocation=none",
--- a/py/services/auto_tag_service.py
+++ b/py/services/auto_tag_service.py
@@ -0,0 +1,139 @@
 """
 Auto-tag extraction service for model cards.
 Extracts implicit model attributes (HIGH/LOW, I2V/T2V/TI2V, Lightning, Turbo)
 from filename, base_model, and CivitAI version name — no manual tagging required.
 """
 from __future__ import annotations
 import re
 from typing import Dict, List, Set
 # ── Tag category definitions ──────────────────────────────────────────
 # Each category maps a display label to a regex pattern.
 # Patterns are case-insensitive and matched against filename, base_model,
 # and civitai version name.
 # Use (?<![a-zA-Z0-9]) and (?![a-zA-Z0-9]) instead of \b because
 # Python's \b treats underscore as a word character, so \bHIGH\b
 # won't match '_HIGH_' in filenames.
 _B = r"(?<![a-zA-Z0-9])"   # left boundary
 _E = r"(?![a-zA-Z0-9])"    # right boundary
 AUTO_TAG_CATEGORIES: Dict[str, str] = {
    "HIGH":      _B + r"HIGH" + _E,
    "LOW":       _B + r"(?<!F)LOW" + _E,
    "I2V":       _B + r"I2V" + _E,
    "T2V":       _B + r"T2V" + _E,
    "TI2V":      _B + r"TI2V" + _E,
    "Lightning": _B + r"Lightning" + _E,
    "Turbo":     _B + r"Turbo" + _E,
 }
 # Tags that belong to the "mode" group (HIGH/LOW)
 MODE_TAGS = {"HIGH", "LOW"}
 # Tags that belong to the "video mode" group (I2V/T2V/TI2V)
 VIDEO_MODE_TAGS = {"I2V", "T2V", "TI2V"}
 # Tags that belong to the "speed/optimization" group
 SPEED_TAGS = {"Lightning", "Turbo"}
 # ── Display category groups (for settings UI) ─────────────────────────
 AUTO_TAG_GROUPS = {
    "mode":   {"HIGH", "LOW"},
    "video":  {"I2V", "T2V", "TI2V"},
    "speed":  {"Lightning", "Turbo"},
 }
 # Default enabled categories
 DEFAULT_ENABLED_GROUPS = {"mode", "video"}
 def _collect_sources(model_data: Dict) -> List[str]:
    """Collect all text sources from model data for tag matching."""
    sources: List[str] = []
    file_name = model_data.get("file_name", "")
    if file_name:
        sources.append(file_name)
    base_model = model_data.get("base_model", "")
    if base_model:
        sources.append(base_model)
    civitai = model_data.get("civitai", {})
    if isinstance(civitai, dict):
        version_name = civitai.get("name", "")
        if version_name:
            sources.append(version_name)
    return sources
 def extract_auto_tags(model_data: Dict) -> List[str]:
    """Extract auto-detected tags from model metadata.
    Uses a two-layer approach:
      Layer 1 — Regex-based detection against filename, base_model, and
                CivitAI version name.
      Layer 2 — Merge in any user-defined tags that overlap with known
                auto-tag categories. This provides a manual fallback when
                auto-detection fails (e.g. "I2V HN" or unlabeled models).
    HIGH/LOW tags are only returned when the base_model indicates a Wan
    family model — no other model architecture uses this distinction.
    Args:
        model_data: Model metadata dict with keys:
            file_name, base_model, civitai (with optional 'name' field),
            tags (user-defined tag list, used as fallback).
    Returns:
        Sorted list of unique auto-tag strings (e.g. ["I2V"]).
    """
    sources = _collect_sources(model_data)
    base_model = model_data.get("base_model", "")
    is_wan = "wan" in base_model.lower()
    found: Set[str] = set()
    # ── Layer 1: regex-based detection ────────────────────────────
    if sources:
        for label, pattern in AUTO_TAG_CATEGORIES.items():
            # HIGH/LOW are Wan-specific — skip for non-Wan to avoid noise
            if label in ("HIGH", "LOW"):
                if not is_wan:
                    continue
                # Use case-insensitive character class + case-sensitive boundary,
                # so "HighNoise" (camelCase) matches but "highlight" doesn't.
                # Boundary: not followed by lowercase letter (= word has ended).
                ci = "".join(f"[{c.lower()}{c.upper()}]" for c in label)
                if label == "LOW":
                    regex = re.compile(r"(?<![Ff])" + ci + r"(?![a-z])")
                else:
                    regex = re.compile(ci + r"(?![a-z])")
            else:
                regex = re.compile(pattern, re.IGNORECASE)
            for source in sources:
                if regex.search(source):
                    found.add(label)
                    break
    # ── Layer 2: user-defined tags as manual fallback ─────────────
    # When auto-detection fails (abbreviated names like "Hi"/"Lo",
    # "I2V HN", or unlabeled models), users can add canonical tags
    # (HIGH, LOW, I2V, etc.) to the model's regular tags for correct
    # badge display and filtering. Matching is case-insensitive so
    # "high"/"High"/"HIGH" all resolve to the canonical label.
    user_tags = model_data.get("tags")
    if user_tags:
        label_map = {label.lower(): label for label in AUTO_TAG_CATEGORIES}
        for t in user_tags:
            canonical = label_map.get(t.lower())
            if canonical:
                found.add(canonical)
    return sorted(found)
--- a/py/services/base_model_service.py
+++ b/py/services/base_model_service.py
@@ -77,6 +77,7 @@ class BaseModelService(ABC):
        base_models: list = None,
        model_types: list = None,
        tags: Optional[Dict[str, str]] = None,
        auto_tags: Optional[Dict[str, str]] = None,
        search_options: dict = None,
        hash_filters: dict = None,
        favorites_only: bool = False,
@@ -95,6 +96,11 @@ class BaseModelService(ABC):
            sorted_data = await self._fetch_with_usage_sort(sort_params)
        else:
            sorted_data = await self.cache_repository.fetch_sorted(sort_params)
        # Pre-compute auto_tags for every item — needed for both filtering
        # and display. Computation is cheap (string regex on 2-3 fields).
        from .auto_tag_service import extract_auto_tags
        for item in sorted_data:
            item["auto_tags"] = extract_auto_tags(item)
        fetch_duration = time.perf_counter() - t0
        initial_count = len(sorted_data)
@@ -110,6 +116,7 @@ class BaseModelService(ABC):
                base_models=base_models,
                model_types=model_types,
                tags=tags,
                auto_tags=auto_tags,
                favorites_only=favorites_only,
                search_options=search_options,
                tag_logic=tag_logic,
@@ -354,6 +361,7 @@ class BaseModelService(ABC):
        base_models: list = None,
        model_types: list = None,
        tags: Optional[Dict[str, str]] = None,
        auto_tags: Optional[Dict[str, str]] = None,
        favorites_only: bool = False,
        search_options: dict = None,
        tag_logic: str = "any",
@@ -367,6 +375,7 @@ class BaseModelService(ABC):
            base_models=base_models,
            model_types=model_types,
            tags=tags,
            auto_tags=auto_tags,
            favorites_only=favorites_only,
            search_options=normalized_options,
            tag_logic=tag_logic,
@@ -861,22 +870,75 @@ class BaseModelService(ABC):
        """Get the static preview URL for a model file"""
        cache = await self.scanner.get_cached_data()
        name_normalized = model_name.replace("\\", "/")
        name_no_ext = name_normalized
        for ext in (".safetensors", ".ckpt", ".pt", ".bin"):
            if name_no_ext.lower().endswith(ext):
                name_no_ext = name_no_ext[: -len(ext)]
                break
        has_path = "/" in name_no_ext
        basename = os.path.basename(name_no_ext) if has_path else name_no_ext
        best_fallback = None
        for model in cache.raw_data:
-            if model["file_name"] == model_name:
+            file_name = model.get("file_name", "")
            folder = model.get("folder", "")
            file_name_no_ext = file_name
            for ext in (".safetensors", ".ckpt", ".pt", ".bin"):
                if file_name_no_ext.lower().endswith(ext):
                    file_name_no_ext = file_name_no_ext[: -len(ext)]
                    break
            path_name = f"{folder}/{file_name_no_ext}".replace("\\", "/") if folder else file_name_no_ext
            if name_no_ext == file_name_no_ext or name_no_ext == path_name:
                preview_url = model.get("preview_url")
                if preview_url:
                    from ..config import config
                    return config.get_preview_static_url(preview_url)
            if has_path and file_name_no_ext == basename:
                if folder and name_no_ext.startswith(folder.replace("\\", "/") + "/"):
                    best_fallback = model
                elif best_fallback is None:
                    best_fallback = model
        if best_fallback:
            preview_url = best_fallback.get("preview_url")
            if preview_url:
                from ..config import config
                return config.get_preview_static_url(preview_url)
        return "/loras_static/images/no-preview.png"
    async def get_model_civitai_url(self, model_name: str) -> Dict[str, Optional[str]]:
        """Get the Civitai URL for a model file"""
        cache = await self.scanner.get_cached_data()
        name_normalized = model_name.replace("\\", "/")
        name_no_ext = name_normalized
        for ext in (".safetensors", ".ckpt", ".pt", ".bin"):
            if name_no_ext.lower().endswith(ext):
                name_no_ext = name_no_ext[: -len(ext)]
                break
        has_path = "/" in name_no_ext
        basename = os.path.basename(name_no_ext) if has_path else name_no_ext
        best_fallback = None
        for model in cache.raw_data:
-            if model["file_name"] == model_name:
+            file_name = model.get("file_name", "")
            folder = model.get("folder", "")
            file_name_no_ext = file_name
            for ext in (".safetensors", ".ckpt", ".pt", ".bin"):
                if file_name_no_ext.lower().endswith(ext):
                    file_name_no_ext = file_name_no_ext[: -len(ext)]
                    break
            path_name = f"{folder}/{file_name_no_ext}".replace("\\", "/") if folder else file_name_no_ext
            if name_no_ext == file_name_no_ext or name_no_ext == path_name:
                civitai_data = model.get("civitai", {})
                model_id = civitai_data.get("modelId")
                version_id = civitai_data.get("id")
@@ -895,6 +957,27 @@ class BaseModelService(ABC):
                        "version_id": str(version_id) if version_id else None,
                    }
            if has_path and file_name_no_ext == basename:
                if folder and name_no_ext.startswith(folder.replace("\\", "/") + "/"):
                    best_fallback = model
                elif best_fallback is None:
                    best_fallback = model
        if best_fallback:
            civitai_data = best_fallback.get("civitai", {})
            model_id = civitai_data.get("modelId")
            if model_id:
                version_id = civitai_data.get("id")
                civitai_host = self.settings.get("civitai_host", "civitai.com")
                civitai_url = build_civitai_model_page_url(
                    model_id, version_id, host=civitai_host
                )
                return {
                    "civitai_url": civitai_url,
                    "model_id": str(model_id),
                    "version_id": str(version_id) if version_id else None,
                }
        return {"civitai_url": None, "model_id": None, "version_id": None}
    async def get_model_metadata(self, file_path: str) -> Optional[Dict]:
@@ -908,6 +991,17 @@ class BaseModelService(ABC):
        )
        if should_skip or metadata is None:
            return None
        # Prune stale example-image metadata entries whose files no longer
        # exist on disk (e.g. a user deleted the files manually).
        from ..utils.example_images_metadata import MetadataUpdater
        was_modified = await MetadataUpdater.prune_stale_example_images(metadata)
        if was_modified:
            asyncio.create_task(
                MetadataManager.save_metadata(file_path, metadata)
            )
        return self.filter_civitai_data(metadata.to_dict().get("civitai", {}))
    async def get_model_description(self, file_path: str) -> Optional[str]:
--- a/py/services/batch_import_service.py
+++ b/py/services/batch_import_service.py
@@ -224,7 +224,7 @@ class BatchImportService:
                return False
            for recipe in getattr(cache, "raw_data", []):
-                source_path = recipe.get("source_path") or recipe.get("source_url")
+                source_path = recipe.get("source_path")
                if source_path and source_path == source:
                    return True
            return False
--- a/py/services/checkpoint_service.py
+++ b/py/services/checkpoint_service.py
@@ -3,6 +3,7 @@ import logging
 from typing import Dict
 from .base_model_service import BaseModelService
 from .auto_tag_service import extract_auto_tags
 from ..utils.models import CheckpointMetadata
 from ..config import config
@@ -45,7 +46,8 @@ class CheckpointService(BaseModelService):
            "exclude": bool(checkpoint_data.get("exclude", False)),
            "update_available": bool(checkpoint_data.get("update_available", False)),
            "skip_metadata_refresh": bool(checkpoint_data.get("skip_metadata_refresh", False)),
-            "civitai": self.filter_civitai_data(checkpoint_data.get("civitai", {}), minimal=True)
+            "civitai": self.filter_civitai_data(checkpoint_data.get("civitai", {}), minimal=True),
            "auto_tags": checkpoint_data.get("auto_tags") or extract_auto_tags(checkpoint_data),
        }
    def find_duplicate_hashes(self) -> Dict:
--- a/py/services/civarchive_client.py
+++ b/py/services/civarchive_client.py
@@ -186,6 +186,22 @@ class CivArchiveClient:
        if "metadata" in file_data:
            transformed["metadata"] = file_data["metadata"]
        # Infer metadata.format from filename extension
        name = transformed.get("name")
        if name and isinstance(name, str):
            lower_name = name.lower()
            if lower_name.endswith(".safetensors"):
                inferred_format = "SafeTensor"
            elif lower_name.endswith(".ckpt"):
                inferred_format = "PickleTensor"
            else:
                inferred_format = None
            if inferred_format:
                if "metadata" not in transformed:
                    transformed["metadata"] = {}
                if isinstance(transformed["metadata"], dict):
                    transformed["metadata"].setdefault("format", inferred_format)
        if file_data.get("modelVersionId") is not None:
            transformed["modelVersionId"] = file_data.get("modelVersionId")
        elif file_data.get("model_version_id") is not None:
@@ -213,6 +229,20 @@ class CivArchiveClient:
        for file_data in candidates:
            if isinstance(file_data, dict):
                transformed_files.append(self._transform_file_entry(file_data))
        # Sort: .safetensors first, .ckpt second, others last
        # so the backend fallback (no file_params) prefers safetensors
        def _sort_key(f: Dict) -> int:
            fname = f.get("name") or ""
            if isinstance(fname, str):
                lower = fname.lower()
                if lower.endswith(".safetensors"):
                    return 0
                elif lower.endswith(".ckpt"):
                    return 1
            return 2
        transformed_files.sort(key=_sort_key)
        return transformed_files
    def _transform_version(
--- a/py/services/civitai_base_model_service.py
+++ b/py/services/civitai_base_model_service.py
@@ -193,6 +193,9 @@ class CivitaiBaseModelService:
            "zimageturbo": "ZIT",
            "zimagebase": "ZIB",
            "anima": "ANI",
            "ernie": "ERNI",
            "ernie turbo": "ETRB",
            "nucleus": "NUCL",
            "svd": "SVD",
            "ltxv": "LTXV",
            "ltxv2": "LTV2",
@@ -418,6 +421,9 @@ class CivitaiBaseModelService:
                "Kolors",
                "NoobAI",
                "Anima",
                "Ernie",
                "Ernie Turbo",
                "Nucleus",
            ],
        }
--- a/py/services/civitai_client.py
+++ b/py/services/civitai_client.py
@@ -2,6 +2,7 @@ import asyncio
 import copy
 import logging
 import os
 from collections import OrderedDict
 from typing import Any, Optional, Dict, Tuple, List, Sequence
 from .connectivity_guard import (
    OFFLINE_FRIENDLY_MESSAGE,
@@ -45,6 +46,14 @@ class CivitaiClient:
        self._initialized = True
        self.base_url = "https://civitai.red/api/v1"
        # In-memory cache to avoid redundant get_model_version_info calls
        # within the same import/scan flow. Only successful results are cached.
        # Uses OrderedDict with LRU eviction at MAX_CACHE_ENTRIES to prevent
        # unbounded growth in long-running server processes.
        self._version_info_cache: OrderedDict[
            str, Tuple[Optional[Dict], Optional[str]]
        ] = OrderedDict()
        self._MAX_CACHE_ENTRIES = 500
    def _build_image_info_url(self, image_id: str) -> str:
        return f"{self.base_url}/images?imageId={image_id}&nsfw=X"
@@ -57,22 +66,57 @@ class CivitaiClient:
        use_auth: bool = False,
        **kwargs,
    ) -> Tuple[bool, Dict | str]:
-        """Wrapper around downloader.make_request that surfaces rate limits."""
+        """Wrapper around downloader.make_request that surfaces rate limits,
        with retry for transient server errors (5xx, Cloudflare 524, network flakiness)."""
-        downloader = await get_downloader()
+        max_retries = 3
-        success, result = await downloader.make_request(
+        for attempt in range(max_retries):
-            method,
+            downloader = await get_downloader()
-            url,
+            success, result = await downloader.make_request(
-            use_auth=use_auth,
+                method,
-            **kwargs,
+                url,
-        )
+                use_auth=use_auth,
-        if not success and isinstance(result, RateLimitError):
+                **kwargs,
-            if result.provider is None:
+            )
-                result.provider = "civitai_api"
+            if success:
-            raise result
+                return True, result
-        if not success and is_offline_cooldown_error(result):
+
-            return False, OFFLINE_FRIENDLY_MESSAGE
+            if isinstance(result, RateLimitError):
-        return success, result
+                if result.provider is None:
                    result.provider = "civitai_api"
                raise result
            if is_offline_cooldown_error(result):
                return False, OFFLINE_FRIENDLY_MESSAGE
            # Transient server error — retry with exponential backoff
            if self._is_transient_server_error(str(result)):
                if attempt < max_retries - 1:
                    wait = 2**attempt  # 1s, 2s, 4s
                    logger.info(
                        "Transient error on %s %s, retrying in %ds "
                        "(attempt %d/%d): %s",
                        method,
                        url,
                        wait,
                        attempt + 1,
                        max_retries,
                        result,
                    )
                    await asyncio.sleep(wait)
                    continue
                logger.warning(
                    "All %d retries exhausted for %s %s: %s",
                    max_retries,
                    method,
                    url,
                    result,
                )
                return False, result
            return False, result
        return False, "Unexpected error in _make_request"
    @staticmethod
    def _remove_comfy_metadata(model_version: Optional[Dict]) -> None:
@@ -201,6 +245,29 @@ class CivitaiClient:
        return _from_value(payload)
    @staticmethod
    def _is_transient_server_error(message: str) -> bool:
        """Return True when the message indicates a transient upstream failure.
        Recognises Cloudflare 524, generic 5xx, and connectivity-level flakiness
        that should not be treated as a permanent failure.
        """
        normalized = message.lower()
        if "status 5" in normalized or "status 524" in normalized:
            return True
        if any(
            keyword in normalized
            for keyword in (
                "connection refused",
                "connection reset",
                "temporary failure",
                "name resolution",
                "connection closed",
            )
        ):
            return True
        return False
    async def get_model_versions(self, model_id: str) -> Optional[Dict]:
        """Get all versions of a model with local availability info"""
        try:
@@ -223,6 +290,13 @@ class CivitaiClient:
                logger.info("Civitai request skipped: %s", OFFLINE_FRIENDLY_MESSAGE)
                return None
            if message:
                if self._is_transient_server_error(message):
                    logger.info(
                        "Transient server error for model %s: %s",
                        model_id,
                        message,
                    )
                    return None
                raise RuntimeError(message)
            return None
        except RateLimitError:
@@ -257,7 +331,7 @@ class CivitaiClient:
                "GET",
                f"{self.base_url}/models",
                use_auth=True,
-                params={"ids": query},
+                params={"ids": query, "nsfw": "true"},
            )
            if not success:
                return None
@@ -336,6 +410,25 @@ class CivitaiClient:
            return None
        target_version = self._select_target_version(model_data, model_id, version_id)
        # If modelVersions is empty (e.g. CivitAI cache lag for newly published
        # models) but a specific version_id is known, fall back to fetching the
        # version directly via the individual model-versions endpoint, then
        # enrich it with the model-level data we already have.
        if target_version is None and version_id is not None:
            logger.info(
                "modelVersions empty for model %s; falling back to direct "
                "version lookup for %s",
                model_id,
                version_id,
            )
            version = await self._fetch_version_by_id(version_id)
            if version:
                self._enrich_version_with_model_data(version, model_data)
                self._remove_comfy_metadata(version)
                return version
            return None
        if target_version is None:
            return None
@@ -482,6 +575,14 @@ class CivitaiClient:
                - The model version data or None if not found
                - An error message if there was an error, or None on success
        """
        # In-memory cache avoids redundant API calls within the same
        # import/scan flow (e.g. _resolve_base_model_from_checkpoint
        # followed by _resolve_and_populate_checkpoint with the same id).
        if version_id in self._version_info_cache:
            logger.debug("Cache hit for model version info: %s", version_id)
            self._version_info_cache.move_to_end(version_id)  # LRU bump
            return self._version_info_cache[version_id]
        try:
            url = f"{self.base_url}/model-versions/{version_id}"
@@ -491,6 +592,11 @@ class CivitaiClient:
            if success:
                logger.debug("Successfully fetched model version info for: %s", version_id)
                self._remove_comfy_metadata(result)
                self._version_info_cache[version_id] = (result, None)
                self._version_info_cache.move_to_end(version_id)
                # Evict oldest entry when over capacity
                if len(self._version_info_cache) > self._MAX_CACHE_ENTRIES:
                    self._version_info_cache.popitem(last=False)
                return result, None
            # Handle specific error cases
@@ -532,6 +638,13 @@ class CivitaiClient:
            if not success:
                if is_expected_offline_error(result):
                    return None
                if self._is_transient_server_error(str(result)):
                    logger.info(
                        "Transient server error fetching image info for ID %s: %s",
                        image_id,
                        result,
                    )
                    return None
                logger.error(
                    "Failed to fetch image info for ID %s from civitai.red: %s",
                    image_id,
@@ -577,6 +690,59 @@ class CivitaiClient:
            logger.error(error_msg)
            return None
    async def get_model_versions_by_hashes(
        self, hashes: List[str]
    ) -> Optional[List[Dict]]:
        """Fetch full version details for up to 100 SHA256 hashes via the batch endpoint.
        Uses POST /api/v1/model-versions/by-hash which returns full version
        details including ``usageControl`` and ``earlyAccessEndsAt`` that are
        not available from the model-level API.
        Args:
            hashes: List of SHA256 hashes (max 100 per batch; auto-split).
        Returns:
            List of version dicts or None on failure.
        """
        if not hashes:
            return []
        BATCH_SIZE = 100
        all_versions: List[Dict] = []
        for start in range(0, len(hashes), BATCH_SIZE):
            batch = hashes[start : start + BATCH_SIZE]
            try:
                success, result = await self._make_request(
                    "POST",
                    f"{self.base_url}/model-versions/by-hash",
                    use_auth=True,
                    json=batch,
                )
                if not success:
                    logger.warning(
                        "Batch by-hash request failed for %d hashes: %s",
                        len(batch),
                        result,
                    )
                    continue
                if isinstance(result, list):
                    all_versions.extend(result)
                else:
                    logger.debug(
                        "Unexpected by-hash response type: %s", type(result)
                    )
            except RateLimitError:
                raise
            except Exception as exc:  # pragma: no cover - defensive logging
                logger.error(
                    "Error fetching model versions by hashes: %s", exc
                )
        return all_versions if all_versions else None
    async def get_user_models(self, username: str) -> Optional[List[Dict]]:
        """Fetch all models for a specific Civitai user."""
        if not username:
@@ -587,7 +753,7 @@ class CivitaiClient:
                "GET",
                f"{self.base_url}/models",
                use_auth=True,
-                params={"username": username},
+                params={"username": username, "nsfw": "true"},
            )
            if not success:
--- a/py/services/download_coordinator.py
+++ b/py/services/download_coordinator.py
@@ -110,6 +110,23 @@ class DownloadCoordinator:
        return result
    async def skip_download(self, download_id: str) -> Dict[str, Any]:
        """Skip a download while preserving all partial files on disk."""
        download_manager = await self._download_manager_factory()
        result = await download_manager.skip_download(download_id)
        await self._ws_manager.broadcast_download_progress(
            download_id,
            {
                "status": "skipped",
                "progress": 0,
                "download_id": download_id,
                "message": "Download skipped by user (partial files preserved)",
            },
        )
        return result
    async def pause_download(self, download_id: str) -> Dict[str, Any]:
        """Pause an active download and notify listeners."""
--- a/py/services/download_manager.py
+++ b/py/services/download_manager.py
@@ -18,6 +18,7 @@ from ..utils.constants import (
    VALID_LORA_TYPES,
 )
 from ..utils.civitai_utils import normalize_civitai_download_url, rewrite_preview_url
 from ..utils.file_utils import calculate_sha256
 from ..utils.preview_selection import resolve_mature_threshold, select_preview_media
 from ..utils.utils import sanitize_folder_name
 from ..utils.exif_utils import ExifUtils
@@ -2239,8 +2240,11 @@ class DownloadManager:
            entry.file_name = os.path.splitext(os.path.basename(file_path))[0]
            # Update size to actual downloaded file size
            entry.size = os.path.getsize(file_path)
-            # Use SHA256 from API metadata (already set in from_civitai_info)
+            # Compute SHA256 locally when the API response didn't include it
-            # Do not recalculate to avoid blocking during ComfyUI execution
+            if not entry.sha256:
                sha256 = await calculate_sha256(file_path)
                if sha256:
                    entry.sha256 = sha256.lower()
            entries.append(entry)
        return entries
@@ -2400,6 +2404,89 @@ class DownloadManager:
                self._download_tasks.pop(download_id, None)
                await self._aria2_state_store.remove(download_id)
    async def skip_download(self, download_id: str) -> Dict:
        """Skip a download while preserving all partial files on disk.
        Removes all in-memory tracking (asyncio task, semaphore, active/pause
        state) but keeps partial files (.part / .aria2) on disk so that a
        subsequent download-model-get request for the same save path can
        auto-resume from the preserved partial download.
        Args:
            download_id: The unique identifier of the download task
        Returns:
            Dict: Status of the skip operation
        """
        await self._restore_persisted_downloads()
        if download_id not in self._download_tasks and download_id not in self._active_downloads:
            return {"success": False, "error": "Download task not found"}
        download_info = self._active_downloads.get(download_id)
        task = self._download_tasks.get(download_id)
        active_statuses = {"queued", "waiting", "downloading", "paused", "cancelling"}
        if task is None and (
            not isinstance(download_info, dict)
            or download_info.get("status") not in active_statuses
        ):
            return {"success": False, "error": "Download task not found"}
        backend = (
            self._active_downloads.get(download_id, {}).get("transfer_backend")
            or "python"
        )
        try:
            # For aria2: pause the transfer rather than force-removing it, so
            # the .aria2 control file stays on disk for future resume
            if backend == "aria2":
                try:
                    aria2_downloader = await get_aria2_downloader()
                    pause_result = await aria2_downloader.pause_download(download_id)
                    if not pause_result.get("success"):
                        logger.warning(
                            "Failed to pause aria2 transfer for %s during skip: %s",
                            download_id,
                            pause_result.get("error"),
                        )
                except Exception as exc:
                    logger.warning(
                        "Failed to pause aria2 transfer for %s during skip: %s",
                        download_id,
                        exc,
                    )
            # Cancel the asyncio task so the semaphore slot is released
            if task is not None:
                task.cancel()
            # Resume pause event so the task can exit cleanly
            pause_control = self._pause_events.get(download_id)
            if pause_control is not None:
                pause_control.resume()
            # Wait briefly for task to acknowledge cancellation
            if task is not None:
                try:
                    await asyncio.wait_for(asyncio.shield(task), timeout=2.0)
                except (asyncio.CancelledError, asyncio.TimeoutError):
                    pass
            logger.info(f"Download skipped for task {download_id} (partial files preserved)")
            return {"success": True, "message": "Download skipped successfully"}
        except Exception as e:
            logger.error(f"Error skipping download: {e}", exc_info=True)
            return {"success": False, "error": str(e)}
        finally:
            # Clean up local in-memory tracking only - NO file deletion
            self._pause_events.pop(download_id, None)
            self._download_tasks.pop(download_id, None)
            if download_id in self._active_downloads:
                del self._active_downloads[download_id]
            # Preserve aria2 state store entry so the partial download
            # info survives restarts and can be resumed later
    async def pause_download(self, download_id: str) -> Dict:
        """Pause an active download without losing progress."""
--- a/py/services/downloaded_version_history_service.py
+++ b/py/services/downloaded_version_history_service.py
@@ -64,6 +64,7 @@ class DownloadedVersionHistoryService:
        self._db_path = db_path or _resolve_database_path()
        self._settings = settings_manager or get_settings_manager()
        self._lock = asyncio.Lock()
        self._conn: sqlite3.Connection | None = None
        self._schema_initialized = False
        self._ensure_directory()
        self._initialize_schema()
@@ -78,6 +79,12 @@ class DownloadedVersionHistoryService:
        conn.row_factory = sqlite3.Row
        return conn
    def _get_conn(self) -> sqlite3.Connection:
        if self._conn is None:
            self._conn = sqlite3.connect(self._db_path, check_same_thread=False)
            self._conn.row_factory = sqlite3.Row
        return self._conn
    def _initialize_schema(self) -> None:
        if self._schema_initialized:
            return
@@ -89,6 +96,21 @@ class DownloadedVersionHistoryService:
    def get_database_path(self) -> str:
        return self._db_path
    def close(self) -> None:
        """Close the persistent SQLite connection, if open.
        This is called before plugin update operations to release the
        database file lock on Windows, allowing ``shutil.rmtree()`` to
        succeed when the cache resides inside the plugin directory.
        """
        if self._conn is not None:
            try:
                self._conn.close()
            except Exception:
                pass
            finally:
                self._conn = None
    def _get_active_library_name(self) -> str | None:
        try:
            value = self._settings.get_active_library_name()
@@ -116,33 +138,33 @@ class DownloadedVersionHistoryService:
        timestamp = time.time()
        async with self._lock:
-            with self._connect() as conn:
+            conn = self._get_conn()
-                conn.execute(
+            conn.execute(
-                    """
+                """
-                    INSERT INTO downloaded_model_versions (
+                INSERT INTO downloaded_model_versions (
-                        model_type, version_id, model_id, first_seen_at, last_seen_at,
+                    model_type, version_id, model_id, first_seen_at, last_seen_at,
-                        source, last_file_path, last_library_name, is_deleted_override
+                    source, last_file_path, last_library_name, is_deleted_override
-                    ) VALUES (?, ?, ?, ?, ?, ?, ?, ?, 0)
+                ) VALUES (?, ?, ?, ?, ?, ?, ?, ?, 0)
-                    ON CONFLICT(model_type, version_id) DO UPDATE SET
+                ON CONFLICT(model_type, version_id) DO UPDATE SET
-                        model_id = COALESCE(excluded.model_id, downloaded_model_versions.model_id),
+                    model_id = COALESCE(excluded.model_id, downloaded_model_versions.model_id),
-                        last_seen_at = excluded.last_seen_at,
+                    last_seen_at = excluded.last_seen_at,
-                        source = excluded.source,
+                    source = excluded.source,
-                        last_file_path = COALESCE(excluded.last_file_path, downloaded_model_versions.last_file_path),
+                    last_file_path = COALESCE(excluded.last_file_path, downloaded_model_versions.last_file_path),
-                        last_library_name = COALESCE(excluded.last_library_name, downloaded_model_versions.last_library_name),
+                    last_library_name = COALESCE(excluded.last_library_name, downloaded_model_versions.last_library_name),
-                        is_deleted_override = 0
+                    is_deleted_override = 0
-                    """,
+                """,
-                    (
+                (
-                        normalized_type,
+                    normalized_type,
-                        normalized_version_id,
+                    normalized_version_id,
-                        normalized_model_id,
+                    normalized_model_id,
-                        timestamp,
+                    timestamp,
-                        timestamp,
+                    timestamp,
-                        source,
+                    source,
-                        file_path,
+                    file_path,
-                        active_library_name,
+                    active_library_name,
-                    ),
+                ),
-                )
+            )
-                conn.commit()
+            conn.commit()
    async def mark_downloaded_bulk(
        self,
@@ -180,26 +202,26 @@ class DownloadedVersionHistoryService:
            return
        async with self._lock:
-            with self._connect() as conn:
+            conn = self._get_conn()
-                conn.executemany(
+            conn.executemany(
-                    """
+                """
-                    INSERT INTO downloaded_model_versions (
+                INSERT INTO downloaded_model_versions (
-                        model_type, version_id, model_id, first_seen_at, last_seen_at,
+                    model_type, version_id, model_id, first_seen_at, last_seen_at,
-                        source, last_file_path, last_library_name, is_deleted_override
+                    source, last_file_path, last_library_name, is_deleted_override
-                    ) VALUES (?, ?, ?, ?, ?, ?, ?, ?, 0)
+                ) VALUES (?, ?, ?, ?, ?, ?, ?, ?, 0)
-                    ON CONFLICT(model_type, version_id) DO UPDATE SET
+                ON CONFLICT(model_type, version_id) DO UPDATE SET
-                        model_id = COALESCE(excluded.model_id, downloaded_model_versions.model_id),
+                    model_id = COALESCE(excluded.model_id, downloaded_model_versions.model_id),
-                        last_seen_at = excluded.last_seen_at,
+                    last_seen_at = excluded.last_seen_at,
-                        source = excluded.source,
+                    source = excluded.source,
-                        last_file_path = COALESCE(excluded.last_file_path, downloaded_model_versions.last_file_path),
+                    last_file_path = COALESCE(excluded.last_file_path, downloaded_model_versions.last_file_path),
-                        last_library_name = COALESCE(excluded.last_library_name, downloaded_model_versions.last_library_name),
+                    last_library_name = COALESCE(excluded.last_library_name, downloaded_model_versions.last_library_name),
-                        is_deleted_override = 0
+                    is_deleted_override = 0
-                    """,
+                """,
-                    payload,
+                payload,
-                )
+            )
-                conn.commit()
+            conn.commit()
-    async def mark_not_downloaded(self, model_type: str, version_id: int) -> None:
+    async def mark_as_deleted(self, model_type: str, version_id: int) -> None:
        normalized_type = _normalize_model_type(model_type)
        normalized_version_id = _normalize_int(version_id)
        if normalized_type is None or normalized_version_id is None:
@@ -208,28 +230,28 @@ class DownloadedVersionHistoryService:
        timestamp = time.time()
        async with self._lock:
-            with self._connect() as conn:
+            conn = self._get_conn()
-                conn.execute(
+            conn.execute(
-                    """
+                """
-                    INSERT INTO downloaded_model_versions (
+                INSERT INTO downloaded_model_versions (
-                        model_type, version_id, model_id, first_seen_at, last_seen_at,
+                    model_type, version_id, model_id, first_seen_at, last_seen_at,
-                        source, last_file_path, last_library_name, is_deleted_override
+                    source, last_file_path, last_library_name, is_deleted_override
-                    ) VALUES (?, ?, NULL, ?, ?, 'manual', NULL, ?, 1)
+                ) VALUES (?, ?, NULL, ?, ?, 'manual', NULL, ?, 1)
-                    ON CONFLICT(model_type, version_id) DO UPDATE SET
+                ON CONFLICT(model_type, version_id) DO UPDATE SET
-                        last_seen_at = excluded.last_seen_at,
+                    last_seen_at = excluded.last_seen_at,
-                        source = excluded.source,
+                    source = excluded.source,
-                        last_library_name = COALESCE(excluded.last_library_name, downloaded_model_versions.last_library_name),
+                    last_library_name = COALESCE(excluded.last_library_name, downloaded_model_versions.last_library_name),
-                        is_deleted_override = 1
+                    is_deleted_override = 1
-                    """,
+                """,
-                    (
+                (
-                        normalized_type,
+                    normalized_type,
-                        normalized_version_id,
+                    normalized_version_id,
-                        timestamp,
+                    timestamp,
-                        timestamp,
+                    timestamp,
-                        self._get_active_library_name(),
+                    self._get_active_library_name(),
-                    ),
+                ),
-                )
+            )
-                conn.commit()
+            conn.commit()
    async def has_been_downloaded(self, model_type: str, version_id: int) -> bool:
        normalized_type = _normalize_model_type(model_type)
@@ -238,15 +260,15 @@ class DownloadedVersionHistoryService:
            return False
        async with self._lock:
-            with self._connect() as conn:
+            conn = self._get_conn()
-                row = conn.execute(
+            row = conn.execute(
-                    """
+                """
-                    SELECT is_deleted_override
+                SELECT is_deleted_override
-                    FROM downloaded_model_versions
+                FROM downloaded_model_versions
-                    WHERE model_type = ? AND version_id = ?
+                WHERE model_type = ? AND version_id = ?
-                    """,
+                """,
-                    (normalized_type, normalized_version_id),
+                (normalized_type, normalized_version_id),
-                ).fetchone()
+            ).fetchone()
        return bool(row) and not bool(row["is_deleted_override"])
    async def get_downloaded_version_ids(
@@ -258,16 +280,16 @@ class DownloadedVersionHistoryService:
            return []
        async with self._lock:
-            with self._connect() as conn:
+            conn = self._get_conn()
-                rows = conn.execute(
+            rows = conn.execute(
-                    """
+                """
-                    SELECT version_id
+                SELECT version_id
-                    FROM downloaded_model_versions
+                FROM downloaded_model_versions
-                    WHERE model_type = ? AND model_id = ? AND is_deleted_override = 0
+                WHERE model_type = ? AND model_id = ? AND is_deleted_override = 0
-                    ORDER BY version_id ASC
+                ORDER BY version_id ASC
-                    """,
+                """,
-                    (normalized_type, normalized_model_id),
+                (normalized_type, normalized_model_id),
-                ).fetchall()
+            ).fetchall()
        return [int(row["version_id"]) for row in rows]
    async def get_downloaded_version_ids_bulk(
@@ -291,17 +313,17 @@ class DownloadedVersionHistoryService:
        params: list[object] = [normalized_type, *normalized_model_ids]
        async with self._lock:
-            with self._connect() as conn:
+            conn = self._get_conn()
-                rows = conn.execute(
+            rows = conn.execute(
-                    f"""
+                f"""
-                    SELECT model_id, version_id
+                SELECT model_id, version_id
-                    FROM downloaded_model_versions
+                FROM downloaded_model_versions
-                    WHERE model_type = ?
+                WHERE model_type = ?
-                      AND model_id IN ({placeholders})
+                  AND model_id IN ({placeholders})
-                      AND is_deleted_override = 0
+                  AND is_deleted_override = 0
-                    """,
+                """,
-                    params,
+                params,
-                ).fetchall()
+            ).fetchall()
        result: dict[int, set[int]] = {}
        for row in rows:
--- a/py/services/downloader.py
+++ b/py/services/downloader.py
@@ -13,6 +13,7 @@ This module provides a centralized download service with:
 import os
 import logging
 import asyncio
 import ssl
 import aiohttp
 from collections import deque
 from dataclasses import dataclass
@@ -31,6 +32,20 @@ from .errors import RateLimitError
 logger = logging.getLogger(__name__)
 def is_ssl_cert_verify_error(exc: BaseException) -> bool:
    """Check if an exception represents an SSL certificate verification failure.
    Matches ``ssl.SSLCertVerificationError``, ``aiohttp.ClientConnectorCertificateError``
    (which wraps the former), and falls back to the standard OpenSSL error text.
    """
    if isinstance(exc, ssl.SSLCertVerificationError):
        return True
    cert_error = getattr(exc, "certificate_error", None)
    if isinstance(cert_error, ssl.SSLCertVerificationError):
        return True
    return "CERTIFICATE_VERIFY_FAILED" in str(exc)
@dataclass(frozen=True)
 class DownloadProgress:
    """Snapshot of a download transfer at a moment in time."""
@@ -265,9 +280,22 @@ class Downloader:
            logger.debug(
                "Proxy mode: system-level proxy (trust_env) will be used if configured in environment."
            )
        # Build SSL context: prefer certifi's CA bundle for broader
        # CA coverage across different Python environments (especially
        # embedded/compatibility Python builds).
        try:
            import certifi  # type: ignore[import-untyped]
            ca_path = certifi.where()
            ssl_context = ssl.create_default_context(cafile=ca_path)
            logger.debug("SSL: using certifi CA bundle at %s", ca_path)
        except (ImportError, FileNotFoundError, ValueError, OSError):
            ssl_context = ssl.create_default_context()
            logger.debug("SSL: certifi unavailable; using system default CA bundle")
        # Optimize TCP connection parameters
        connector = aiohttp.TCPConnector(
-            ssl=True,
+            ssl=ssl_context,
            limit=8,  # Concurrent connections
            ttl_dns_cache=300,  # DNS cache timeout
            force_close=False,  # Keep connections for reuse
@@ -736,6 +764,17 @@ class Downloader:
                DownloadRestartRequested,
            ) as e:
                retry_count += 1
                if is_ssl_cert_verify_error(e):
                    logger.error(
                        "SSL certificate verification failed when connecting to %s. "
                        "This is usually caused by an outdated CA certificate bundle "
                        "in the Python environment. Recommended fixes:\n"
                        "  1. pip install --upgrade certifi\n"
                        "  2. pip install pip-system-certs",
                        url,
                    )
                logger.warning(
                    f"Network error during download (attempt {retry_count}/{self.max_retries + 1}): {e}"
                )
--- a/py/services/embedding_service.py
+++ b/py/services/embedding_service.py
@@ -3,6 +3,7 @@ import logging
 from typing import Dict
 from .base_model_service import BaseModelService
 from .auto_tag_service import extract_auto_tags
 from ..utils.models import EmbeddingMetadata
 from ..config import config
@@ -45,7 +46,8 @@ class EmbeddingService(BaseModelService):
            "exclude": bool(embedding_data.get("exclude", False)),
            "update_available": bool(embedding_data.get("update_available", False)),
            "skip_metadata_refresh": bool(embedding_data.get("skip_metadata_refresh", False)),
-            "civitai": self.filter_civitai_data(embedding_data.get("civitai", {}), minimal=True)
+            "civitai": self.filter_civitai_data(embedding_data.get("civitai", {}), minimal=True),
            "auto_tags": embedding_data.get("auto_tags") or extract_auto_tags(embedding_data),
        }
    def find_duplicate_hashes(self) -> Dict:
--- a/py/services/lora_service.py
+++ b/py/services/lora_service.py
@@ -5,6 +5,7 @@ from typing import Dict, List, Optional
 from .base_model_service import BaseModelService
 from .model_query import resolve_sub_type
 from .auto_tag_service import extract_auto_tags
 from ..utils.models import LoraMetadata
 from ..config import config
@@ -57,6 +58,7 @@ class LoraService(BaseModelService):
            "civitai": self.filter_civitai_data(
                lora_data.get("civitai", {}), minimal=True
            ),
            "auto_tags": lora_data.get("auto_tags") or extract_auto_tags(lora_data),
        }
    async def _apply_specific_filters(self, data: List[Dict], **kwargs) -> List[Dict]:
@@ -310,8 +312,23 @@ class LoraService(BaseModelService):
        """Return cached raw metadata for a LoRA matching the given filename."""
        cache = await self.scanner.get_cached_data(force_refresh=False)
        fn_normalized = filename.replace("\\", "/")
        fn_no_ext = fn_normalized
        for ext in (".safetensors", ".ckpt", ".pt", ".bin"):
            if fn_no_ext.lower().endswith(ext):
                fn_no_ext = fn_no_ext[: -len(ext)]
                break
        for lora in cache.raw_data if cache else []:
-            if lora.get("file_name") == filename:
+            file_name = lora.get("file_name", "")
            folder = lora.get("folder", "")
            file_name_no_ext = file_name
            for ext in (".safetensors", ".ckpt", ".pt", ".bin"):
                if file_name_no_ext.lower().endswith(ext):
                    file_name_no_ext = file_name_no_ext[: -len(ext)]
                    break
            path_name = f"{folder}/{file_name_no_ext}".replace("\\", "/") if folder else file_name_no_ext
            if fn_no_ext in (file_name_no_ext, path_name):
                return lora
        return None
@@ -399,7 +416,10 @@ class LoraService(BaseModelService):
            locked_loras = locked_loras[:target_count]
        # Filter out locked LoRAs from available pool
-        locked_names = {lora["name"] for lora in locked_loras}
+        locked_names = {
            os.path.basename(lora["name"]) if "/" in str(lora.get("name", "")) else lora["name"]
            for lora in locked_loras
        }
        available_pool = [
            l for l in available_loras if l["file_name"] not in locked_names
        ]
@@ -454,7 +474,7 @@ class LoraService(BaseModelService):
            result_loras.append(
                {
-                    "name": lora["file_name"],
+                    "name": f"{lora['folder']}/{lora['file_name']}" if lora.get("folder") else lora["file_name"],
                    "strength": model_str,
                    "clipStrength": clip_str,
                    "active": True,
@@ -670,8 +690,9 @@ class LoraService(BaseModelService):
        # Return minimal data needed for cycling
        return [
            {
-                "file_name": lora["file_name"],
+                "file_name": f"{lora['folder']}/{lora['file_name']}" if lora.get("folder") else lora["file_name"],
                "model_name": lora.get("model_name", lora["file_name"]),
                "folder": lora.get("folder", ""),
            }
            for lora in available_loras
        ]
--- a/py/services/model_hash_index.py
+++ b/py/services/model_hash_index.py
@@ -7,6 +7,7 @@ class ModelHashIndex:
    def __init__(self):
        self._hash_to_path: Dict[str, str] = {}
        self._filename_to_hash: Dict[str, str] = {}
        self._autov2_to_path: Dict[str, str] = {}
        # New data structures for tracking duplicates
        self._duplicate_hashes: Dict[str, List[str]] = {}  # sha256 -> list of paths
        self._duplicate_filenames: Dict[str, List[str]] = {}  # filename -> list of paths
@@ -63,6 +64,9 @@ class ModelHashIndex:
        # Add new mappings
        self._hash_to_path[sha256] = file_path
        self._filename_to_hash[filename] = sha256
        # AutoV2 = first 10 chars of SHA256
        if len(sha256) >= 10:
            self._autov2_to_path[sha256[:10]] = file_path
    def _get_filename_from_path(self, file_path: str) -> str:
        """Extract filename without extension from path"""
@@ -157,7 +161,12 @@ class ModelHashIndex:
                del self._duplicate_filenames[filename]
                if filename in self._filename_to_hash:
                    del self._filename_to_hash[filename]
-    
+
        # Remove from AutoV2 index
        autov2_keys_to_remove = [k for k, v in self._autov2_to_path.items() if v == file_path]
        for k in autov2_keys_to_remove:
            del self._autov2_to_path[k]
    def remove_by_hash(self, sha256: str) -> None:
        """Remove entry by hash"""
        sha256 = sha256.lower()
@@ -177,6 +186,10 @@ class ModelHashIndex:
        # Remove hash-to-path mapping
        del self._hash_to_path[sha256]
        autov2_key = sha256[:10]
        if autov2_key in self._autov2_to_path:
            del self._autov2_to_path[autov2_key]
        # Update filename-to-hash and duplicate filenames for all paths
        for path_to_remove in paths_to_remove:
            fname = self._get_filename_from_path(path_to_remove)
@@ -195,13 +208,24 @@ class ModelHashIndex:
                    # If only one entry remains, it's no longer a duplicate
                    del self._duplicate_filenames[fname]
-    def has_hash(self, sha256: str) -> bool:
+    def has_hash(self, hash_value: str) -> bool:
-        """Check if hash exists in index"""
+        """Check if hash exists in index (SHA256 or AutoV2)"""
-        return sha256.lower() in self._hash_to_path
+        normalized = hash_value.lower()
-    
+        if normalized in self._hash_to_path:
-    def get_path(self, sha256: str) -> Optional[str]:
+            return True
-        """Get file path for a hash"""
+        if len(normalized) == 10:
-        return self._hash_to_path.get(sha256.lower())
+            return normalized in self._autov2_to_path
        return False
    def get_path(self, hash_value: str) -> Optional[str]:
        """Get file path for a hash (SHA256 or AutoV2)"""
        normalized = hash_value.lower()
        path = self._hash_to_path.get(normalized)
        if path is not None:
            return path
        if len(normalized) == 10:
            return self._autov2_to_path.get(normalized)
        return None
    def get_hash(self, file_path: str) -> Optional[str]:
        """Get hash for a file path"""
@@ -209,13 +233,16 @@ class ModelHashIndex:
        return self._filename_to_hash.get(filename)
    def get_hash_by_filename(self, filename: str) -> Optional[str]:
-        """Get hash for a filename without extension"""
+        """Get hash for a filename (bare basename or path-prefixed name)"""
        if "/" in filename or "\\" in filename:
            filename = os.path.splitext(os.path.basename(filename.replace("\\", "/")))[0]
        return self._filename_to_hash.get(filename)
    def clear(self) -> None:
        """Clear all entries"""
        self._hash_to_path.clear()
        self._filename_to_hash.clear()
        self._autov2_to_path.clear()
        self._duplicate_hashes.clear()
        self._duplicate_filenames.clear()
--- a/py/services/model_lifecycle_service.py
+++ b/py/services/model_lifecycle_service.py
@@ -111,6 +111,11 @@ class ModelLifecycleService:
            self._scanner._hash_index.remove_by_path(file_path)
        await self._sync_update_for_model(model_id)
        persist_current_cache = getattr(self._scanner, "_persist_current_cache", None)
        if callable(persist_current_cache):
            await persist_current_cache()
        return {"success": True, "deleted_files": deleted_files}
    @staticmethod
--- a/py/services/model_metadata_provider.py
+++ b/py/services/model_metadata_provider.py
@@ -5,7 +5,7 @@ import logging
 import random
 from typing import Optional, Dict, Tuple, Any, List, Sequence
 from .downloader import get_downloader
-from .errors import RateLimitError
+from .errors import RateLimitError, ResourceNotFoundError
 try:
    from bs4 import BeautifulSoup
@@ -108,6 +108,18 @@ class ModelMetadataProvider(ABC):
    ) -> Optional[Dict[int, Dict]]:
        """Fetch model versions for multiple model ids when supported."""
        raise NotImplementedError
    async def get_model_versions_by_hashes(
        self, hashes: List[str]
    ) -> Optional[List[Dict]]:
        """Fetch full version details for multiple SHA256 hashes.
        Used specifically to retrieve ``usageControl`` which is only
        available from the per-version / by-hash API, not from model-level
        responses.  Providers that cannot resolve hashes should let the
        default ``NotImplementedError`` propagate.
        """
        raise NotImplementedError
    @abstractmethod
    async def get_model_version(self, model_id: int = None, version_id: int = None) -> Optional[Dict]:
@@ -140,6 +152,11 @@ class CivitaiModelMetadataProvider(ModelMetadataProvider):
        self, model_ids: Sequence[int]
    ) -> Optional[Dict[int, Dict]]:
        return await self.client.get_model_versions_bulk(model_ids)
    async def get_model_versions_by_hashes(
        self, hashes: List[str]
    ) -> Optional[List[Dict]]:
        return await self.client.get_model_versions_by_hashes(hashes)
    async def get_model_version(self, model_id: int = None, version_id: int = None) -> Optional[Dict]:
        return await self.client.get_model_version(model_id, version_id)
@@ -465,6 +482,7 @@ class FallbackMetadataProvider(ModelMetadataProvider):
        return None, "Model not found"
    async def get_model_versions(self, model_id: str) -> Optional[Dict]:
        not_found_confirmed = False
        for provider, label in self._iter_providers():
            try:
                result = await self._call_with_rate_limit(
@@ -475,8 +493,24 @@ class FallbackMetadataProvider(ModelMetadataProvider):
                if result:
                    return result
            except RateLimitError as exc:
                if not_found_confirmed:
                    logger.debug(
                        "Suppressing rate limit from %s for model %s: "
                        "already confirmed as not found by another provider",
                        label,
                        model_id,
                    )
                    return None
                exc.provider = exc.provider or label
                raise exc
            except ResourceNotFoundError:
                not_found_confirmed = True
                logger.debug(
                    "Provider %s reports model %s as not found",
                    label,
                    model_id,
                )
                continue
            except Exception as e:
                logger.debug("Provider %s failed for get_model_versions: %s", label, e)
                continue
@@ -519,6 +553,32 @@ class FallbackMetadataProvider(ModelMetadataProvider):
                continue
        return None, "No provider could retrieve the data"
    async def get_model_versions_by_hashes(
        self, hashes: List[str]
    ) -> Optional[List[Dict]]:
        for provider, label in self._iter_providers():
            try:
                result = await self._call_with_rate_limit(
                    label,
                    provider.get_model_versions_by_hashes,
                    hashes,
                )
                if result is not None:
                    return result
            except NotImplementedError:
                continue
            except RateLimitError as exc:
                exc.provider = exc.provider or label
                raise exc
            except Exception as e:
                logger.debug(
                    "Provider %s failed for get_model_versions_by_hashes: %s",
                    label,
                    e,
                )
                continue
        return None
    async def get_user_models(self, username: str) -> Optional[List[Dict]]:
        for provider, label in self._iter_providers():
            try:
@@ -593,6 +653,15 @@ class RateLimitRetryingProvider(ModelMetadataProvider):
            model_ids,
        )
    async def get_model_versions_by_hashes(
        self, hashes: List[str]
    ) -> Optional[List[Dict]]:
        return await self._rate_limit_helper.run(
            self._label,
            self._provider.get_model_versions_by_hashes,
            hashes,
        )
    async def get_model_version(self, model_id: int = None, version_id: int = None) -> Optional[Dict]:
        return await self._rate_limit_helper.run(
            self._label,
@@ -669,6 +738,17 @@ class ModelMetadataProviderManager:
        provider = self._get_provider(provider_name)
        return await provider.get_model_version_info(version_id)
    async def get_model_versions_by_hashes(
        self,
        hashes: List[str],
        provider_name: str = None,
    ) -> Optional[List[Dict]]:
        provider = self._get_provider(provider_name)
        try:
            return await provider.get_model_versions_by_hashes(hashes)
        except NotImplementedError:
            return None
    async def get_user_models(self, username: str, provider_name: str = None) -> Optional[List[Dict]]:
        """Fetch models owned by the specified user"""
        provider = self._get_provider(provider_name)
--- a/py/services/model_query.py
+++ b/py/services/model_query.py
@@ -96,6 +96,7 @@ class FilterCriteria:
    folder_exclude: Optional[Sequence[str]] = None
    base_models: Optional[Sequence[str]] = None
    tags: Optional[Dict[str, str]] = None
    auto_tags: Optional[Dict[str, str]] = None
    favorites_only: bool = False
    search_options: Optional[Dict[str, Any]] = None
    model_types: Optional[Sequence[str]] = None
@@ -359,10 +360,37 @@ class ModelFilterSet:
                ]
            model_types_duration = time.perf_counter() - t0
        auto_tags_duration = 0
        auto_tag_filters = criteria.auto_tags or {}
        if auto_tag_filters:
            t0 = time.perf_counter()
            include_at = set()
            exclude_at = set()
            for tag, state in auto_tag_filters.items():
                if not tag:
                    continue
                if state == "exclude":
                    exclude_at.add(tag)
                else:
                    include_at.add(tag)
            if include_at:
                items = [
                    item for item in items
                    if any(tag in include_at for tag in (item.get("auto_tags") or []))
                ]
            if exclude_at:
                items = [
                    item for item in items
                    if not any(tag in exclude_at for tag in (item.get("auto_tags") or []))
                ]
            auto_tags_duration = time.perf_counter() - t0
        duration = time.perf_counter() - overall_start
        if duration > 0.1:  # Only log if it's potentially slow
            logger.debug(
-                "ModelFilterSet.apply took %.3fs (sfw: %.3fs, fav: %.3fs, folder: %.3fs, base: %.3fs, tags: %.3fs, types: %.3fs). "
+                "ModelFilterSet.apply took %.3fs (sfw: %.3fs, fav: %.3fs, folder: %.3fs, base: %.3fs, tags: %.3fs, types: %.3fs, auto_tags: %.3fs). "
                "Count: %d -> %d",
                duration,
                sfw_duration,
@@ -371,6 +399,7 @@ class ModelFilterSet:
                base_models_duration,
                tags_duration,
                model_types_duration,
                auto_tags_duration,
                initial_count,
                len(items),
            )
--- a/py/services/model_scanner.py
+++ b/py/services/model_scanner.py
@@ -9,7 +9,7 @@ from typing import Any, Awaitable, Callable, Dict, List, Mapping, Optional, Set,
 from ..utils.models import BaseModelMetadata
 from ..config import config
-from ..utils.file_utils import find_preview_file, get_preview_extension
+from ..utils.file_utils import find_preview_file, get_preview_extension, calculate_sha256
 from ..utils.metadata_manager import MetadataManager
 from ..utils.civitai_utils import resolve_license_info
 from .model_cache import ModelCache
@@ -1067,6 +1067,19 @@ class ModelScanner:
        model_data = self._build_cache_entry(metadata, folder=normalized_folder)
        # Compute SHA256 hash when metadata provided none (e.g., CivitAI API response has empty hashes)
        if not model_data.get('sha256') and file_path:
            try:
                logger.info(f"Computing SHA256 hash for {file_path} (was empty from metadata)")
                sha256 = await calculate_sha256(file_path)
                if sha256:
                    model_data['sha256'] = sha256.lower()
                    if isinstance(metadata, BaseModelMetadata):
                        metadata.sha256 = sha256.lower()
                    await MetadataManager.save_metadata(file_path, metadata)
            except Exception as e:
                logger.error(f"Failed to compute SHA256 for {file_path}: {e}")
        # Skip excluded models
        if model_data.get('exclude', False):
            excluded_models.append(model_data['file_path'])
@@ -1101,7 +1114,15 @@ class ModelScanner:
    def _log_duplicate_filename_summary(self) -> None:
        """Log a batched summary of duplicate filename conflicts once per scan."""
-        if self._hash_index is None:
+        # Duplicate filename detection is only relevant for LoRAs, which use
        # basename-only syntax (<lora:name:strength>). Checkpoints and embeddings
        # use full relative paths for resolution, so conflicts are not ambiguous.
        if self._hash_index is None or self.model_type != "lora":
            return
        # When full path syntax is active, duplicate filenames across subfolders
        # are fully qualified, so there is no ambiguity — skip the warning.
        if get_settings_manager().get("lora_syntax_format", "legacy") == "full":
            return
        duplicates = self._hash_index.get_duplicate_filenames()
@@ -1473,6 +1494,15 @@ class ModelScanner:
                file_path_override=normalized_new_path,
            )
            # Ensure sha256 is populated even when metadata doesn't have it
            if not cache_entry.get('sha256') and normalized_new_path and os.path.exists(normalized_new_path):
                try:
                    sha256 = await calculate_sha256(normalized_new_path)
                    if sha256:
                        cache_entry['sha256'] = sha256.lower()
                except Exception as e:
                    logger.error(f"Failed to compute SHA256 for {normalized_new_path}: {e}")
            if recalculate_type:
                cache_entry = self.adjust_cached_entry(cache_entry)
@@ -1572,12 +1602,39 @@ class ModelScanner:
        """Get model information by name"""
        try:
            cache = await self.get_cached_data()
-            
+
            name_normalized = name.replace("\\", "/")
            name_no_ext = name_normalized
            for ext in (".safetensors", ".ckpt", ".pt", ".bin"):
                if name_no_ext.lower().endswith(ext):
                    name_no_ext = name_no_ext[: -len(ext)]
                    break
            has_path = "/" in name_no_ext
            basename = os.path.basename(name_no_ext) if has_path else name_no_ext
            best_fallback = None
            for model in cache.raw_data:
-                if model.get("file_name") == name:
+                file_name = model.get("file_name", "")
                folder = model.get("folder", "")
                file_name_no_ext = file_name
                for ext in (".safetensors", ".ckpt", ".pt", ".bin"):
                    if file_name_no_ext.lower().endswith(ext):
                        file_name_no_ext = file_name_no_ext[: -len(ext)]
                        break
                path_name = f"{folder}/{file_name_no_ext}".replace("\\", "/") if folder else file_name_no_ext
                if name_no_ext == file_name_no_ext or name_no_ext == path_name:
                    return model
-                    
+
-            return None
+                if has_path and file_name_no_ext == basename:
                    if folder and name_no_ext.startswith(folder.replace("\\", "/") + "/"):
                        best_fallback = model
                    elif best_fallback is None:
                        best_fallback = model
            return best_fallback
        except Exception as e:
            logger.error(f"Error getting model info by name: {e}", exc_info=True)
            return None
--- a/py/services/model_update_service.py
+++ b/py/services/model_update_service.py
@@ -69,6 +69,7 @@ class ModelVersionRecord:
    early_access_ends_at: Optional[str] = None
    sort_index: int = 0
    is_early_access: bool = False
    usage_control: Optional[str] = None  # "Download", "Generation", "InternalGeneration"
@dataclass
@@ -101,11 +102,14 @@ class ModelUpdateRecord:
        return [version.version_id for version in self.versions if version.is_in_library]
-    def has_update(self, hide_early_access: bool = False) -> bool:
+    def has_update(
        self, hide_early_access: bool = False, hide_non_downloadable: bool = True
    ) -> bool:
        """Return True when a non-ignored remote version newer than the newest local copy is available.
        Args:
            hide_early_access: If True, exclude early access versions from update check.
            hide_non_downloadable: If True, exclude versions that don't allow downloads.
        """
        if self.should_ignore_model:
@@ -121,6 +125,7 @@ class ModelUpdateRecord:
                not version.is_in_library
                and not version.should_ignore
                and not (hide_early_access and ModelUpdateRecord._is_early_access_active(version))
                and not (hide_non_downloadable and not ModelUpdateRecord._is_downloadable(version))
                for version in self.versions
            )
@@ -129,6 +134,8 @@ class ModelUpdateRecord:
                continue
            if hide_early_access and ModelUpdateRecord._is_early_access_active(version):
                continue
            if hide_non_downloadable and not ModelUpdateRecord._is_downloadable(version):
                continue
            if version.version_id > max_in_library:
                return True
        return False
@@ -155,11 +162,18 @@ class ModelUpdateRecord:
        # Phase 1: Basic EA flag from bulk API
        return version.is_early_access
    @staticmethod
    def _is_downloadable(version: ModelVersionRecord) -> bool:
        if version.usage_control is None:
            return True
        return version.usage_control == "Download"
    def has_update_for_base(
        self,
        local_version_id: Optional[int],
        local_base_model: Optional[str],
        hide_early_access: bool = False,
        hide_non_downloadable: bool = True,
    ) -> bool:
        """Return True when a newer remote version with the same base model exists.
@@ -167,6 +181,7 @@ class ModelUpdateRecord:
            local_version_id: The current local version id.
            local_base_model: The base model to filter by.
            hide_early_access: If True, exclude early access versions from update check.
            hide_non_downloadable: If True, exclude versions that don't allow downloads.
        """
        if self.should_ignore_model:
@@ -197,6 +212,8 @@ class ModelUpdateRecord:
                continue
            if hide_early_access and ModelUpdateRecord._is_early_access_active(version):
                continue
            if hide_non_downloadable and not ModelUpdateRecord._is_downloadable(version):
                continue
            version_base = _normalize_base_model(version.base_model)
            if version_base != normalized_base:
                continue
@@ -230,6 +247,7 @@ class ModelUpdateService:
            preview_url TEXT,
            is_in_library INTEGER NOT NULL DEFAULT 0,
            should_ignore INTEGER NOT NULL DEFAULT 0,
            usage_control TEXT,
            PRIMARY KEY (model_id, version_id),
            FOREIGN KEY(model_id) REFERENCES model_update_status(model_id) ON DELETE CASCADE
        );
@@ -465,6 +483,10 @@ class ModelUpdateService:
                "ALTER TABLE model_update_versions "
                "ADD COLUMN is_early_access INTEGER NOT NULL DEFAULT 0"
            ),
            "usage_control": (
                "ALTER TABLE model_update_versions "
                "ADD COLUMN usage_control TEXT"
            ),
        }
        for column, statement in migrations.items():
@@ -667,6 +689,7 @@ class ModelUpdateService:
        *,
        force_refresh: bool = False,
        target_model_ids: Optional[Sequence[int]] = None,
        folder_path: Optional[str] = None,
    ) -> Dict[int, ModelUpdateRecord]:
        """Refresh update information for every model present in the cache."""
        scanner.reset_cancellation()
@@ -681,6 +704,7 @@ class ModelUpdateService:
        local_versions = await self._collect_local_versions(
            scanner,
            target_model_ids=target_filter,
            folder_path=folder_path,
        )
        total_models = len(local_versions)
        if total_models == 0:
@@ -967,18 +991,22 @@ class ModelUpdateService:
                fallback_attempted = True
                try:
                    response = await metadata_provider.get_model_versions(model_id)
                    if response is not None:
                        await self._enrich_version_entries(
                            metadata_provider,
                            {model_id: response},
                        )
                except RateLimitError:
                    raise
                except ResourceNotFoundError as exc:
                    fallback_error_message = str(exc) or "resource not found"
                    mark_model_as_ignored = True
                except Exception as exc:  # pragma: no cover - defensive log
-                    logger.error(
+                    logger.warning(
                        "Failed to fetch versions for model %s (%s): %s",
                        model_id,
                        model_type,
                        exc,
                        exc_info=True,
                    )
                    fallback_error_message = str(exc)
            if response is not None:
@@ -1061,6 +1089,136 @@ class ModelUpdateService:
            self._upsert_record(record)
            return record
    async def _enrich_version_entries(
        self,
        metadata_provider,
        responses_by_model_id: Dict[int, Mapping],
    ) -> None:
        """Enrich version entries with ``usageControl`` via batch hash endpoint.
        The model-level API does not include ``usageControl`` on version
        entries. This method collects SHA256 hashes from every version's
        primary model file, calls ``POST /api/v1/model-versions/by-hash``
        (up to 100 hashes per request), and injects ``usageControl`` +
        ``earlyAccessEndsAt`` into each version entry dict in-place.
        """
        if not metadata_provider or not responses_by_model_id:
            return
        hashes_by_version: Dict[int, str] = {}
        for response in responses_by_model_id.values():
            hashes_by_version.update(
                self._collect_hashes_from_response(response)
            )
        if not hashes_by_version:
            return
        version_ids_by_hash: Dict[str, List[int]] = {}
        for version_id, sha256 in hashes_by_version.items():
            version_ids_by_hash.setdefault(sha256, []).append(version_id)
        all_hashes = list(version_ids_by_hash.keys())
        BATCH_SIZE = 100
        enrichment: Dict[int, Dict] = {}
        try:
            for start in range(0, len(all_hashes), BATCH_SIZE):
                batch = all_hashes[start : start + BATCH_SIZE]
                try:
                    enriched = await metadata_provider.get_model_versions_by_hashes(
                        batch
                    )
                except NotImplementedError:
                    return
                except RateLimitError:
                    raise
                except Exception:
                    continue
                if not enriched:
                    continue
                for entry in enriched:
                    if not isinstance(entry, dict):
                        continue
                    version_id = entry.get("id")
                    if version_id is None:
                        continue
                    enrichment[version_id] = {
                        "usageControl": _normalize_string(
                            entry.get("usageControl")
                        ),
                        "earlyAccessEndsAt": _normalize_string(
                            entry.get("earlyAccessEndsAt")
                        ),
                    }
        except RateLimitError:
            raise
        if not enrichment:
            return
        for response in responses_by_model_id.values():
            versions = response.get("modelVersions")
            if not isinstance(versions, list):
                continue
            for version in versions:
                if not isinstance(version, dict):
                    continue
                version_id = version.get("id")
                if version_id not in enrichment:
                    continue
                extra = enrichment[version_id]
                if extra.get("usageControl") and not version.get("usageControl"):
                    version["usageControl"] = extra["usageControl"]
                if extra.get("earlyAccessEndsAt") and not version.get(
                    "earlyAccessEndsAt"
                ):
                    version["earlyAccessEndsAt"] = extra["earlyAccessEndsAt"]
    @staticmethod
    def _collect_hashes_from_response(response: Mapping) -> Dict[int, str]:
        """Extract ``{version_id: sha256}`` from a model-level API response.
        Returns an empty dict if the response structure is unexpected.
        """
        result: Dict[int, str] = {}
        versions = response.get("modelVersions")
        if not isinstance(versions, list):
            return result
        for entry in versions:
            if not isinstance(entry, dict):
                continue
            version_id = _normalize_int(entry.get("id"))
            if version_id is None:
                continue
            sha256 = ModelUpdateService._extract_sha256_from_version_entry(entry)
            if sha256:
                result[version_id] = sha256
        return result
    @staticmethod
    def _extract_sha256_from_version_entry(entry: Mapping) -> Optional[str]:
        """Return the SHA256 hash from the primary model file of a version entry."""
        files = entry.get("files")
        if not isinstance(files, list):
            return None
        for file_info in files:
            if not isinstance(file_info, dict):
                continue
            if file_info.get("type") != "Model":
                continue
            primary = file_info.get("primary")
            if primary is not True and str(primary).strip().lower() != "true":
                continue
            hashes = file_info.get("hashes")
            if isinstance(hashes, dict):
                sha256 = hashes.get("SHA256")
                if sha256:
                    return sha256
        return None
    async def _fetch_model_versions_bulk(
        self,
        metadata_provider,
@@ -1112,6 +1270,7 @@ class ModelUpdateService:
            len(aggregated),
            provider_name,
        )
        await self._enrich_version_entries(metadata_provider, aggregated)
        return aggregated
    async def _collect_local_versions(
@@ -1119,6 +1278,7 @@ class ModelUpdateService:
        scanner,
        *,
        target_model_ids: Optional[Sequence[int]] = None,
        folder_path: Optional[str] = None,
    ) -> Dict[int, List[int]]:
        cache = await scanner.get_cached_data()
        mapping: Dict[int, set[int]] = {}
@@ -1131,7 +1291,19 @@ class ModelUpdateService:
            if not target_set:
                return {}
        normalized_folder = None
        if folder_path is not None:
            normalized_folder = folder_path.replace("\\", "/").strip("/")
        for item in cache.raw_data:
            # Apply folder filter first (cheapest check)
            if normalized_folder is not None:
                if not isinstance(item, dict):
                    continue
                item_folder = (item.get("folder") or "").replace("\\", "/").strip("/")
                if item_folder != normalized_folder and not item_folder.startswith(normalized_folder + "/"):
                    continue
            civitai = item.get("civitai") if isinstance(item, dict) else None
            if not isinstance(civitai, dict):
                continue
@@ -1239,6 +1411,7 @@ class ModelUpdateService:
                    sort_index=sort_map.get(version_id, index),
                    early_access_ends_at=remote_version.early_access_ends_at,
                    is_early_access=remote_version.is_early_access,
                    usage_control=remote_version.usage_control,
                )
            )
@@ -1337,6 +1510,7 @@ class ModelUpdateService:
        # Check availability field from bulk API for basic EA detection
        availability = _normalize_string(entry.get("availability"))
        is_early_access = availability == "EarlyAccess"
        usage_control = _normalize_string(entry.get("usageControl"))
        return ModelVersionRecord(
            version_id=version_id,
@@ -1350,6 +1524,7 @@ class ModelUpdateService:
            early_access_ends_at=early_access_ends_at,
            sort_index=index,
            is_early_access=is_early_access,
            usage_control=usage_control,
        )
    def _extract_size_bytes(self, files) -> Optional[int]:
@@ -1464,7 +1639,7 @@ class ModelUpdateService:
                    f"""
                    SELECT model_id, version_id, sort_index, name, base_model, released_at,
                           size_bytes, preview_url, is_in_library, should_ignore, early_access_ends_at,
-                           is_early_access
+                           is_early_access, usage_control
                    FROM model_update_versions
                    WHERE model_id IN ({placeholders})
                    ORDER BY model_id ASC, sort_index ASC, version_id ASC
@@ -1492,6 +1667,7 @@ class ModelUpdateService:
                    early_access_ends_at=row["early_access_ends_at"],
                    sort_index=_normalize_int(row["sort_index"]) or 0,
                    is_early_access=bool(row["is_early_access"]),
                    usage_control=row["usage_control"],
                )
            )
@@ -1548,8 +1724,8 @@ class ModelUpdateService:
                    INSERT INTO model_update_versions (
                        version_id, model_id, sort_index, name, base_model, released_at,
                        size_bytes, preview_url, is_in_library, should_ignore, early_access_ends_at,
-                        is_early_access
+                        is_early_access, usage_control
-                    ) VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?)
+                    ) VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?)
                    """,
                    (
                        version.version_id,
@@ -1564,6 +1740,7 @@ class ModelUpdateService:
                        1 if version.should_ignore else 0,
                        version.early_access_ends_at,
                        1 if version.is_early_access else 0,
                        version.usage_control,
                    ),
                )
            conn.commit()
--- a/py/services/persistent_recipe_cache.py
+++ b/py/services/persistent_recipe_cache.py
@@ -38,6 +38,7 @@ class PersistentRecipeCache:
        "json_path",
        "title",
        "folder",
        "source_path",
        "base_model",
        "fingerprint",
        "created_date",
@@ -334,6 +335,7 @@ class PersistentRecipeCache:
                            json_path TEXT,
                            title TEXT,
                            folder TEXT,
                            source_path TEXT,
                            base_model TEXT,
                            fingerprint TEXT,
                            created_date REAL,
@@ -358,6 +360,13 @@ class PersistentRecipeCache:
                        );
                        """
                    )
                    # Migration: add source_path column to existing databases
                    try:
                        conn.execute(
                            "ALTER TABLE recipes ADD COLUMN source_path TEXT"
                        )
                    except Exception:
                        pass  # column already exists
                    conn.commit()
                self._schema_initialized = True
            except Exception as exc:
@@ -406,6 +415,7 @@ class PersistentRecipeCache:
            json_path,
            recipe.get("title"),
            recipe.get("folder"),
            recipe.get("source_path"),
            recipe.get("base_model"),
            recipe.get("fingerprint"),
            float(recipe.get("created_date") or 0.0),
@@ -456,6 +466,7 @@ class PersistentRecipeCache:
            "file_path": row["file_path"] or "",
            "title": row["title"] or "",
            "folder": row["folder"] or "",
            "source_path": row["source_path"] or "",
            "base_model": row["base_model"] or "",
            "fingerprint": row["fingerprint"] or "",
            "created_date": row["created_date"] or 0.0,
--- a/py/services/recipe_scanner.py
+++ b/py/services/recipe_scanner.py
@@ -65,7 +65,7 @@ class RecipeScanner:
            cls._instance._civitai_client = None  # Will be lazily initialized
        return cls._instance
-    REPAIR_VERSION = 3
+    REPAIR_VERSION = 4
    def __init__(
        self,
@@ -292,6 +292,32 @@ class RecipeScanner:
        if recipe.get("repair_version", 0) >= self.REPAIR_VERSION:
            return False
        # 1.5 Detect and clear corrupted checkpoint (LoRA data saved as checkpoint).
        #     A checkpoint whose modelVersionId also appears in a LoRA entry is
        #     definitely wrong — the CivitAI import code used to pick
        #     modelVersionIds[0] as the checkpoint, which was often a LoRA.
        #     Clearing it lets the enrichment flow re-resolve the correct
        #     checkpoint from CivitAI image metadata.
        cp = recipe.get("checkpoint")
        lora_mvids = {
            l.get("modelVersionId")
            for l in recipe.get("loras", [])
            if l.get("modelVersionId")
        }
        if cp and cp.get("modelVersionId") and cp["modelVersionId"] in lora_mvids:
            cp_mvid = cp["modelVersionId"]
            logger.info(
                "Recipe %s: checkpoint modelVersionId %s matches a LoRA — "
                "clearing corrupted checkpoint and removing matching LoRA entry",
                recipe.get("id"),
                cp_mvid,
            )
            recipe["checkpoint"] = None
            recipe["loras"] = [
                l for l in recipe.get("loras", [])
                if l.get("modelVersionId") != cp_mvid
            ]
        # 2. Identification: Is repair needed?
        has_checkpoint = (
            "checkpoint" in recipe
@@ -504,6 +530,9 @@ class RecipeScanner:
                    self._cache.raw_data = recipes
                    self._update_folder_metadata(self._cache)
                    self._sort_cache_sync()
                    # Backfill source_path from JSON files if missing (schema migration)
                    if self._backfill_source_path_if_needed(recipes, json_paths):
                        self._persistent_cache.save_cache(recipes, json_paths)
                    return self._cache
                else:
                    # Partial update: some files changed
@@ -514,6 +543,8 @@ class RecipeScanner:
                    self._cache.raw_data = recipes
                    self._update_folder_metadata(self._cache)
                    self._sort_cache_sync()
                    # Backfill source_path from JSON files if missing (schema migration)
                    self._backfill_source_path_if_needed(recipes, json_paths)
                    # Persist updated cache
                    self._persistent_cache.save_cache(recipes, json_paths)
                    return self._cache
@@ -642,6 +673,34 @@ class RecipeScanner:
        return recipes, changed, json_paths
    def _backfill_source_path_if_needed(
        self,
        recipes: List[Dict],
        json_paths: Dict[str, str],
    ) -> bool:
        """Backfill source_path from recipe JSON files if missing from cache.
        Returns True if any recipes were updated (caller should persist cache).
        """
        updated = False
        for recipe in recipes:
            if recipe.get("source_path"):
                continue
            recipe_id = str(recipe.get("id", ""))
            json_path = json_paths.get(recipe_id)
            if not json_path or not os.path.exists(json_path):
                continue
            try:
                with open(json_path, "r", encoding="utf-8") as f:
                    json_data = json.load(f)
                file_source_path = json_data.get("source_path")
                if file_source_path:
                    recipe["source_path"] = file_source_path
                    updated = True
            except Exception:
                pass
        return updated
    def _full_directory_scan_sync(
        self, recipes_dir: str
    ) -> Tuple[List[Dict], Dict[str, str]]:
@@ -2484,6 +2543,7 @@ class RecipeScanner:
                continue
            file_name = None
            folder = ""
            hash_value = (lora.get("hash") or "").lower()
            if (
                hash_value
@@ -2493,6 +2553,11 @@ class RecipeScanner:
                file_path = self._lora_scanner._hash_index.get_path(hash_value)
                if file_path:
                    file_name = os.path.splitext(os.path.basename(file_path))[0]
                    if lora_cache is not None:
                        for cached_lora in getattr(lora_cache, "raw_data", []):
                            if cached_lora.get("file_path") == file_path:
                                folder = cached_lora.get("folder", "")
                                break
            if not file_name and lora.get("modelVersionId") and lora_cache is not None:
                for cached_lora in getattr(lora_cache, "raw_data", []):
@@ -2507,13 +2572,16 @@ class RecipeScanner:
                            file_name = os.path.splitext(os.path.basename(cached_path))[
                                0
                            ]
                            folder = cached_lora.get("folder", "")
                        break
            if not file_name:
                file_name = lora.get("file_name", "unknown-lora")
                folder = lora.get("folder", "")
            lora_name = f"{folder}/{file_name}" if folder else file_name
            strength = lora.get("strength", 1.0)
-            syntax_parts.append(f"<lora:{file_name}:{strength}>")
+            syntax_parts.append(f"<lora:{lora_name}:{strength}>")
        return syntax_parts
--- a/py/services/recipes/analysis_service.py
+++ b/py/services/recipes/analysis_service.py
@@ -2,6 +2,7 @@
 from __future__ import annotations
 import asyncio
 import base64
 import io
 import os
@@ -14,6 +15,7 @@ from PIL import Image
 from ...utils.utils import calculate_recipe_fingerprint
 from ...utils.civitai_utils import extract_civitai_image_id, rewrite_preview_url
 from ...recipes.enrichment import RecipeEnricher
 from .errors import (
    RecipeDownloadError,
    RecipeNotFoundError,
@@ -170,9 +172,29 @@ class RecipeAnalysisService:
                await self._download_image(url, temp_path)
            if metadata is None and not is_video:
-                metadata = self._exif_utils.extract_image_metadata(temp_path)
+                metadata = await asyncio.to_thread(
                    self._exif_utils.extract_image_metadata, temp_path
                )
-            return await self._parse_metadata(
+            if not metadata and civitai_image_id and image_info:
                original_url = image_info.get("url")
                if original_url:
                    self._logger.debug(
                        "Optimized image lacks embedded metadata, "
                        "falling back to original image: %s",
                        original_url,
                    )
                    orig_temp_path = self._create_temp_path(suffix=".png")
                    try:
                        await self._download_image(original_url, orig_temp_path)
                        metadata = await asyncio.to_thread(
                            self._exif_utils.extract_image_metadata,
                            orig_temp_path,
                        )
                    finally:
                        self._safe_cleanup(orig_temp_path)
            result = await self._parse_metadata(
                metadata or {},
                recipe_scanner=recipe_scanner,
                image_path=temp_path,
@@ -180,6 +202,37 @@ class RecipeAnalysisService:
                is_video=is_video,
                extension=extension,
            )
            if civitai_image_id and image_info and not result.payload.get("error"):
                mvid = image_info.get("modelVersionId")
                if not mvid:
                    mvids = image_info.get("modelVersionIds")
                    if isinstance(mvids, list) and mvids:
                        mvid = mvids[0]
                recipe_for_enrich = {
                    "gen_params": result.payload.get("gen_params", {}),
                    "loras": result.payload.get("loras", []),
                    "base_model": result.payload.get("base_model", "") or "",
                    "checkpoint": result.payload.get("checkpoint") or result.payload.get("model"),
                    "source_path": url,
                }
                await RecipeEnricher.enrich_recipe(
                    recipe=recipe_for_enrich,
                    civitai_client=civitai_client,
                    request_params=None,
                    prefetched_civitai_meta_raw=image_info.get("meta"),
                    prefetched_model_version_id=mvid,
                )
                result.payload["gen_params"] = recipe_for_enrich["gen_params"]
                if recipe_for_enrich.get("checkpoint"):
                    result.payload["checkpoint"] = recipe_for_enrich["checkpoint"]
                if recipe_for_enrich.get("base_model"):
                    result.payload["base_model"] = recipe_for_enrich["base_model"]
            return result
        finally:
            if temp_path:
                self._safe_cleanup(temp_path)
@@ -199,7 +252,9 @@ class RecipeAnalysisService:
        if not os.path.isfile(normalized_path):
            raise RecipeNotFoundError("File not found")
-        metadata = self._exif_utils.extract_image_metadata(normalized_path)
+        metadata = await asyncio.to_thread(
            self._exif_utils.extract_image_metadata, normalized_path
        )
        if not metadata:
            return self._metadata_not_found_response(normalized_path)
--- a/py/services/recipes/persistence_service.py
+++ b/py/services/recipes/persistence_service.py
@@ -115,6 +115,10 @@ class RecipePersistenceService:
        if metadata.get("source_path"):
            recipe_data["source_path"] = metadata.get("source_path")
        nsfw_level = metadata.get("preview_nsfw_level")
        if nsfw_level is not None and isinstance(nsfw_level, int):
            recipe_data["preview_nsfw_level"] = nsfw_level
        json_filename = f"{recipe_id}.recipe.json"
        json_path = os.path.join(recipes_dir, json_filename)
        json_path = os.path.normpath(json_path)
--- a/py/services/settings_manager.py
+++ b/py/services/settings_manager.py
@@ -96,6 +96,7 @@ DEFAULT_SETTINGS: Dict[str, Any] = {
    "compact_mode": False,
    "priority_tags": DEFAULT_PRIORITY_TAG_CONFIG.copy(),
    "model_name_display": "model_name",
    "lora_syntax_format": "legacy",
    "model_card_footer_action": "replace_preview",
    "show_version_on_card": True,
    "update_flag_strategy": "same_base",
--- a/py/services/tag_update_service.py
+++ b/py/services/tag_update_service.py
@@ -4,7 +4,9 @@ from __future__ import annotations
 import os
-from typing import Awaitable, Callable, Dict, List, Sequence
+from typing import Awaitable, Callable, Dict, List, Sequence, Tuple
 from .auto_tag_service import extract_auto_tags
 class TagUpdateService:
@@ -20,9 +22,8 @@ class TagUpdateService:
        new_tags: Sequence[str],
        metadata_loader: Callable[[str], Awaitable[Dict[str, object]]],
        update_cache: Callable[[str, str, Dict[str, object]], Awaitable[bool]],
-    ) -> List[str]:
+    ) -> Tuple[List[str], List[str]]:
-        """Add tags to a metadata entry while keeping case-insensitive uniqueness."""
+        """Add tags to a metadata entry and return updated tags and auto_tags."""
        base, _ = os.path.splitext(file_path)
        metadata_path = f"{base}.metadata.json"
        metadata = await metadata_loader(metadata_path)
@@ -44,5 +45,6 @@ class TagUpdateService:
        await self._metadata_manager.save_metadata(file_path, metadata)
        await update_cache(file_path, file_path, metadata)
-        return existing_tags
+        auto_tags = extract_auto_tags(metadata)
        return existing_tags, auto_tags
--- a/py/utils/civitai_utils.py
+++ b/py/utils/civitai_utils.py
@@ -7,7 +7,7 @@ from typing import Any, Dict, Iterable, Mapping, Sequence
 from urllib.parse import parse_qs, urlparse, urlunparse
-_SUPPORTED_CIVITAI_PAGE_HOSTS = frozenset({"civitai.com", "civitai.red"})
+_SUPPORTED_CIVITAI_PAGE_HOSTS = frozenset({"civitai.com", "civitai.red", "civitai.green"})
 DEFAULT_CIVITAI_PAGE_HOST = "civitai.com"
 _DEFAULT_ALLOW_COMMERCIAL_USE: Sequence[str] = ("Sell",)
 _LICENSE_DEFAULTS: Dict[str, Any] = {
@@ -66,6 +66,46 @@ def build_civitai_model_page_url(
    return None
 _RE_CDN_IMAGE_ID = re.compile(r"/(\d+)\.(?:jpeg|jpg|png|webp|gif)(?:\?|#|$)")
 def extract_civitai_image_id_from_cdn_url(url: str | None) -> str | None:
    """Extract the numeric image ID from a Cloudflare CDN image URL.
    CivitAI image CDN URLs follow the pattern::
        https://image.civitai.com/{cf_uuid}/{params}/{image_id}.{ext}
    The image database ID is always the last path segment (minus extension)
    because ``getEdgeUrl(…, name=id.toString())`` embeds it explicitly
    in the model-versions REST API response.
    """
    if not url:
        return None
    match = _RE_CDN_IMAGE_ID.search(url)
    return match.group(1) if match else None
 def build_civitai_image_page_url(
    image_id: str | int | None,
    *,
    host: str | None = None,
 ) -> str | None:
    """Build a Civitai image page URL.
    Returns something like ``https://civitai.com/images/12345``.
    The host is resolved through :func:`normalize_civitai_page_host` and
    therefore respects the user's ``civitai_host`` setting.
    """
    if not image_id:
        return None
    normalized_host = normalize_civitai_page_host(host)
    normalized_id = str(image_id).strip()
    if not normalized_id:
        return None
    return urlunparse(("https", normalized_host, f"/images/{normalized_id}", "", "", ""))
 def _parse_supported_civitai_page_url(url: str | None):
    if not url:
        return None
@@ -239,9 +279,9 @@ def _resolve_commercial_bits(values: Sequence[str]) -> int:
            normalized_values.add(normalized)
    has_sell = "sell" in normalized_values
-    has_rent = has_sell or "rent" in normalized_values
+    has_rent = "rent" in normalized_values
-    has_rentcivit = has_rent or "rentcivit" in normalized_values
+    has_rentcivit = "rentcivit" in normalized_values
-    has_image = has_sell or "image" in normalized_values
+    has_image = "image" in normalized_values
    commercial_bits = (
        (1 if has_sell else 0) << 3
@@ -328,8 +368,10 @@ def rewrite_preview_url(
 __all__ = [
    "build_civitai_image_page_url",
    "build_license_flags",
    "extract_civitai_image_id",
    "extract_civitai_image_id_from_cdn_url",
    "extract_civitai_page_host",
    "extract_civitai_model_url_parts",
    "is_supported_civitai_page_host",
--- a/py/utils/constants.py
+++ b/py/utils/constants.py
@@ -101,8 +101,34 @@ DEFAULT_PRIORITY_TAG_CONFIG = {
 DIFFUSION_MODEL_BASE_MODELS = frozenset(
    [
        "Anima",
-        "ZImageTurbo",
+        # Flux series — DiT architecture, loaded via UNETLoader in ComfyUI
-        "ZImageBase",
+        "Flux.1 D",
        "Flux.1 S",
        "Flux.1 Krea",
        "Flux.1 Kontext",
        "Flux.2 D",
        "Flux.2 Klein 9B",
        "Flux.2 Klein 9B-base",
        "Flux.2 Klein 4B",
        "Flux.2 Klein 4B-base",
        # Non-UNet / DiT image diffusion models
        "AuraFlow",
        "Chroma",
        "HiDream",
        "Hunyuan 1",
        "Kolors",
        "Lumina",
        "PixArt a",
        "PixArt E",
        # Video diffusion models
        "CogVideoX",
        "Hunyuan Video",
        "LTXV",
        "LTXV2",
        "LTXV 2.3",
        "Mochi",
        "SVD",
        "Wan Video",
        "Wan Video 1.3B t2v",
        "Wan Video 14B t2v",
        "Wan Video 14B i2v 480p",
@@ -112,9 +138,13 @@ DIFFUSION_MODEL_BASE_MODELS = frozenset(
        "Wan Video 2.2 T2V-A14B",
        "Wan Video 2.5 T2V",
        "Wan Video 2.5 I2V",
-        "CogVideoX",
+        # Other diffusion models
-        "Mochi",
+        "Ernie",
        "Ernie Turbo",
        "Nucleus",
        "Qwen",
        "ZImageBase",
        "ZImageTurbo",
    ]
 )
@@ -178,5 +208,8 @@ SUPPORTED_DOWNLOAD_SKIP_BASE_MODELS = frozenset(
        "Wan Video 2.5 I2V",
        "Hunyuan Video",
        "Anima",
        "Ernie",
        "Ernie Turbo",
        "Nucleus",
    ]
 )
--- a/py/utils/example_images_download_manager.py
+++ b/py/utils/example_images_download_manager.py
@@ -397,13 +397,12 @@ class DownloadManager:
            models_with_hash = len(all_models_with_hash)
-            # Calculate pending count: check which models actually need processing
+            # Calculate pending count: check which models actually need processing.
-            # A model is pending if it has a hash, is not in processed_models,
+            # A model is pending if it has a hash, is not already processed or known-failed,
-            # and its folder doesn't exist or is empty
+            # and its folder doesn't exist or is empty.
            pending_hashes = set()
            for model_hash, model_name in all_models_with_hash:
-                if model_hash not in processed_models:
+                if model_hash not in processed_models and model_hash not in failed_models:
                    # Check if model folder exists with files
                    model_dir = ExampleImagePathResolver.get_model_folder(
                        model_hash, active_library
                    )
--- a/py/utils/example_images_metadata.py
+++ b/py/utils/example_images_metadata.py
@@ -452,3 +452,111 @@ class MetadataUpdater:
        except Exception as e:
            logger.error(f"Error parsing image metadata: {e}", exc_info=True)
            return None
    @staticmethod
    async def prune_stale_example_images(metadata) -> bool:
        """Remove example-image metadata entries whose files no longer exist on disk.
        Checks ``civitai.customImages`` (by ``id``) and ``civitai.images`` entries
        that have an empty ``url`` (no remote fallback) against actual files in
        the model's example-image folder.  Stale entries are removed in-place so
        the caller can persist the cleaned metadata afterwards.
        Args:
            metadata: A ``BaseModelMetadata`` instance (modified in place).
        Returns:
            True if at least one entry was removed.
        """
        from ..utils.example_images_paths import get_model_folder
        model_hash = getattr(metadata, "sha256", None)
        if not model_hash:
            return False
        model_folder = get_model_folder(model_hash)
        if not model_folder:
            return False
        civitai = getattr(metadata, "civitai", None)
        if not isinstance(civitai, dict):
            return False
        has_changes = False
        custom_images = civitai.get("customImages")
        if isinstance(custom_images, list) and custom_images:
            stale: list[int] = []
            for idx, img in enumerate(custom_images):
                img_id = img.get("id", "")
                if not img_id:
                    continue
                if not os.path.isdir(model_folder):
                    stale.append(idx)
                else:
                    found = False
                    try:
                        prefix = f"custom_{img_id}"
                        for fname in os.listdir(model_folder):
                            if fname.startswith(prefix) and os.path.isfile(
                                os.path.join(model_folder, fname)
                            ):
                                found = True
                                break
                    except OSError:
                        stale.append(idx)
                        continue
                    if not found:
                        stale.append(idx)
            if stale:
                for idx in reversed(stale):
                    custom_images.pop(idx)
                has_changes = True
                logger.info(
                    "Pruned %d stale custom image(s) for %s",
                    len(stale),
                    getattr(metadata, "model_name", model_hash),
                )
        images = civitai.get("images")
        if isinstance(images, list) and images:
            stale: list[int] = []
            for idx, img in enumerate(images):
                if img.get("url", ""):
                    # Has a remote fallback – keep it even if the local copy
                    # is gone.
                    continue
                if not os.path.isdir(model_folder):
                    stale.append(idx)
                else:
                    found = False
                    try:
                        prefix = f"image_{idx}."
                        for fname in os.listdir(model_folder):
                            if fname.startswith(prefix):
                                found = True
                                break
                    except OSError:
                        stale.append(idx)
                        continue
                    if not found:
                        stale.append(idx)
            if stale:
                for idx in reversed(stale):
                    images.pop(idx)
                has_changes = True
                logger.info(
                    "Pruned %d stale image entry(ies) for %s",
                    len(stale),
                    getattr(metadata, "model_name", model_hash),
                )
        return has_changes
--- a/py/utils/file_utils.py
+++ b/py/utils/file_utils.py
@@ -1,7 +1,10 @@
 import hashlib
 import json
 import logging
 import os
 import struct
 from typing import Any
 from .constants import (
    CARD_PREVIEW_WIDTH,
@@ -31,7 +34,7 @@ def _get_hash_chunk_size_bytes() -> int:
 async def calculate_sha256(file_path: str) -> str:
-    """Calculate SHA256 hash of a file"""
+    """Calculate SHA256 hash of a file (full file content)."""
    sha256_hash = hashlib.sha256()
    chunk_size = _get_hash_chunk_size_bytes()
    with open(file_path, "rb") as f:
@@ -39,6 +42,79 @@ async def calculate_sha256(file_path: str) -> str:
            sha256_hash.update(byte_block)
    return sha256_hash.hexdigest()
 def calculate_autov2(file_path: str) -> str:
    """Calculate CivitAI AutoV2 hash.
    AutoV2 is the first 10 characters of the full file SHA256.
    Used by CivitAI as a shortened file identifier.
    Reference: https://developer.civitai.com/site/reference/model-versions
    """
    full_hash = hashlib.sha256()
    chunk_size = _get_hash_chunk_size_bytes()
    with open(file_path, "rb") as f:
        for byte_block in iter(lambda: f.read(chunk_size), b""):
            full_hash.update(byte_block)
    return full_hash.hexdigest()[:10]
 def read_safetensors_metadata(file_path: str) -> dict[str, Any]:
    """Read the ``__metadata__`` dict from a safetensors file header.
    Safetensors file format:
      - 8 bytes: header length (little-endian 64-bit)
      - N bytes: UTF-8 JSON header
      - The header JSON contains a ``__metadata__`` key holding arbitrary metadata.
    Returns an empty dict if the file is not a valid safetensors file or has no
    metadata.
    """
    try:
        with open(file_path, "rb") as f:
            header_len_bytes = f.read(8)
            if len(header_len_bytes) < 8:
                return {}
            header_len = struct.unpack("<Q", header_len_bytes)[0]
            header_bytes = f.read(header_len)
            if len(header_bytes) < header_len:
                return {}
            header = json.loads(header_bytes.decode("utf-8"))
            return header.get("__metadata__", {})
    except (OSError, json.JSONDecodeError, UnicodeDecodeError, struct.error, MemoryError, Exception):
        return {}
 def calculate_autov3(file_path: str) -> str | None:
    """Calculate CivitAI AutoV3 hash from a safetensors file.
    AutoV3 is extracted from the safetensors file's embedded metadata, not
    computed from the file bytes directly. The orchestrator reads the
    ``sshs_model_hash`` (kohya-ss format) or ``modelspec.hash_sha256`` field
    from the safetensors header and stores the first 12 characters.
    The embedded hash itself is the SHA256 of the file after skipping the
    8-byte header length + JSON header (a.k.a. the addnet hash / tensor-only
    hash).
    Reference:
      - CivitAI DB trigger: ``SUBSTRING(NEW.hash FROM 1 FOR 12)``
      - https://developer.civitai.com/site/reference/model-versions
    Returns ``None`` when no AutoV3 hash can be determined (e.g. the file is
    not safetensors, or the metadata doesn't contain a recognised hash field).
    """
    metadata = read_safetensors_metadata(file_path)
    if not metadata:
        return None
    embedded_hash = metadata.get("sshs_model_hash") or metadata.get("modelspec.hash_sha256")
    if embedded_hash and isinstance(embedded_hash, str) and len(embedded_hash) >= 12:
        return embedded_hash[:12]
    return None
 def find_preview_file(base_name: str, dir_path: str) -> str:
    """Find preview file for given base name in directory.
--- a/py/utils/session_logging.py
+++ b/py/utils/session_logging.py
@@ -64,6 +64,27 @@ def _build_log_file_path(settings_file: str | None, started_at: datetime) -> str
    return os.path.join(log_dir, f"standalone-session-{timestamp}.log")
 _KEEP_LOG_COUNT = 3
 def _prune_old_logs(log_dir: str) -> None:
    """Remove older session log files, keeping only the ``_KEEP_LOG_COUNT`` newest."""
    try:
        files = [
            os.path.join(log_dir, name)
            for name in os.listdir(log_dir)
            if name.startswith("standalone-session-") and name.endswith(".log")
        ]
    except OSError:
        return
    files.sort(key=os.path.getmtime, reverse=True)
    for path in files[_KEEP_LOG_COUNT:]:
        try:
            os.remove(path)
        except OSError:
            pass
 def setup_standalone_session_logging(settings_file: str | None) -> StandaloneSessionLogState:
    global _session_state
@@ -90,6 +111,7 @@ def setup_standalone_session_logging(settings_file: str | None) -> StandaloneSes
            file_handler.set_name(_FILE_HANDLER_NAME)
            file_handler.setFormatter(formatter)
            root_logger.addHandler(file_handler)
            _prune_old_logs(os.path.dirname(log_file_path))
        _session_state = StandaloneSessionLogState(
            started_at=started_at,
--- a/py/utils/utils.py
+++ b/py/utils/utils.py
@@ -15,30 +15,64 @@ def get_lora_info(lora_name):
        scanner = await ServiceRegistry.get_lora_scanner()
        cache = await scanner.get_cached_data()
        lora_name_normalized = lora_name.replace("\\", "/")
        lora_name_no_ext = lora_name_normalized
        for ext in (".safetensors", ".ckpt", ".pt", ".bin"):
            if lora_name_no_ext.lower().endswith(ext):
                lora_name_no_ext = lora_name_no_ext[: -len(ext)]
                break
        has_path = "/" in lora_name_no_ext
        basename = os.path.basename(lora_name_no_ext) if has_path else lora_name_no_ext
        best_fallback = None
        for item in cache.raw_data:
-            if item.get("file_name") == lora_name:
+            file_name = item.get("file_name", "")
-                file_path = item.get("file_path")
+            folder = item.get("folder", "")
-                if file_path:
+            file_name_no_ext = file_name
-                    # Check all lora roots including extra paths
+            for ext in (".safetensors", ".ckpt", ".pt", ".bin"):
-                    all_roots = list(config.loras_roots or []) + list(
+                if file_name_no_ext.lower().endswith(ext):
-                        config.extra_loras_roots or []
+                    file_name_no_ext = file_name_no_ext[: -len(ext)]
                    break
            path_name = f"{folder}/{file_name_no_ext}".replace("\\", "/") if folder else file_name_no_ext
            if lora_name_no_ext not in (file_name_no_ext, path_name):
                if has_path and file_name_no_ext == basename:
                    if folder and lora_name_no_ext.startswith(folder.replace("\\", "/") + "/"):
                        best_fallback = item
                    elif best_fallback is None:
                        best_fallback = item
                continue
            file_path = item.get("file_path")
            if not file_path:
                continue
            all_roots = list(config.loras_roots or []) + list(
                config.extra_loras_roots or []
            )
            for root in all_roots:
                root = root.replace(os.sep, "/")
                if file_path.startswith(root):
                    relative_path = os.path.relpath(file_path, root).replace(
                        os.sep, "/"
                    )
                    for root in all_roots:
                        root = root.replace(os.sep, "/")
                        if file_path.startswith(root):
                            relative_path = os.path.relpath(file_path, root).replace(
                                os.sep, "/"
                            )
                            # Get trigger words from civitai metadata
                            civitai = item.get("civitai", {})
                            trigger_words = (
                                civitai.get("trainedWords", []) if civitai else []
                            )
                            return relative_path, trigger_words
                    # If not found in any root, return path with trigger words from cache
                    civitai = item.get("civitai", {})
-                    trigger_words = civitai.get("trainedWords", []) if civitai else []
+                    trigger_words = (
-                    return file_path, trigger_words
+                        civitai.get("trainedWords", []) if civitai else []
                    )
                    return relative_path, trigger_words
            civitai = item.get("civitai", {})
            trigger_words = civitai.get("trainedWords", []) if civitai else []
            return file_path, trigger_words
        if best_fallback:
            file_path = best_fallback.get("file_path")
            if file_path:
                civitai = best_fallback.get("civitai", {})
                trigger_words = civitai.get("trainedWords", []) if civitai else []
                return file_path, trigger_words
        return lora_name, []
    try:
@@ -77,15 +111,54 @@ def get_lora_info_absolute(lora_name):
        scanner = await ServiceRegistry.get_lora_scanner()
        cache = await scanner.get_cached_data()
        lora_name_normalized = lora_name.replace("\\", "/")
        lora_name_no_ext = lora_name_normalized
        for ext in (".safetensors", ".ckpt", ".pt", ".bin"):
            if lora_name_no_ext.lower().endswith(ext):
                lora_name_no_ext = lora_name_no_ext[: -len(ext)]
                break
        has_path = "/" in lora_name_no_ext
        basename = os.path.basename(lora_name_no_ext) if has_path else lora_name_no_ext
        best_fallback = None
        for item in cache.raw_data:
-            if item.get("file_name") == lora_name:
+            file_name = item.get("file_name", "")
            folder = item.get("folder", "")
            file_name_no_ext = file_name
            for ext in (".safetensors", ".ckpt", ".pt", ".bin"):
                if file_name_no_ext.lower().endswith(ext):
                    file_name_no_ext = file_name_no_ext[: -len(ext)]
                    break
            path_name = f"{folder}/{file_name_no_ext}".replace("\\", "/") if folder else file_name_no_ext
            if lora_name_no_ext == file_name_no_ext:
                file_path = item.get("file_path")
                if file_path:
                    # Return absolute path directly
                    # Get trigger words from civitai metadata
                    civitai = item.get("civitai", {})
                    trigger_words = civitai.get("trainedWords", []) if civitai else []
                    return file_path, trigger_words
            if lora_name_no_ext == path_name:
                file_path = item.get("file_path")
                if file_path:
                    civitai = item.get("civitai", {})
                    trigger_words = civitai.get("trainedWords", []) if civitai else []
                    return file_path, trigger_words
            if has_path and file_name_no_ext == basename:
                if folder and lora_name_no_ext.startswith(folder.replace("\\", "/") + "/"):
                    best_fallback = item
                elif best_fallback is None:
                    best_fallback = item
        if best_fallback:
            file_path = best_fallback.get("file_path")
            if file_path:
                civitai = best_fallback.get("civitai", {})
                trigger_words = civitai.get("trainedWords", []) if civitai else []
                return file_path, trigger_words
        return lora_name, []
    try:
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -1,7 +1,7 @@
 [project]
 name = "comfyui-lora-manager"
 description = "Revolutionize your workflow with the ultimate LoRA companion for ComfyUI!"
-version = "1.0.5"
+version = "1.0.11"
 license = {file = "LICENSE"}
 dependencies = [
    "aiohttp",
--- a/scripts/migrate_legacy_metadata.py
+++ b/scripts/migrate_legacy_metadata.py
@@ -0,0 +1,353 @@
 #!/usr/bin/env python3
 """
 Migrate metadata from old sidecar JSON format to LoRA Manager's metadata.json format.
 This script automatically discovers model folders from LoRA Manager's settings.json,
 finds JSON files with the same basename as model files (e.g., `model.json` for
 `model.safetensors`), and migrates their content to the corresponding `.metadata.json` files.
 Fields migrated:
 - "activation text" → civitai.trainedWords (array of trigger words)
 - "preferred weight" → usage_tips.strength (LoRA only, skipped for Checkpoint)
 - "notes" → notes (user-defined notes)
 Supported model types: LoRA, Checkpoint
 Usage:
    python scripts/migrate_legacy_metadata.py [--dry-run] [--verbose]
 The script will:
 1. Read settings.json to find all configured model folders
 2. Recursively scan for model files (.safetensors, .ckpt, .pt, .pth, .bin)
 3. Find corresponding legacy metadata JSON files
 4. Migrate data to .metadata.json files
 """
 from __future__ import annotations
 import argparse
 import json
 import logging
 import os
 import re
 import sys
 from pathlib import Path
 from typing import Any
 from platformdirs import user_config_dir
 logging.basicConfig(
    level=logging.INFO,
    format="%(asctime)s - %(levelname)s - %(message)s",
 )
 logger = logging.getLogger(__name__)
 APP_NAME = "ComfyUI-LoRA-Manager"
 MODEL_EXTENSIONS = {".safetensors", ".ckpt", ".pt", ".pth", ".bin"}
 SECRET_PATTERN = re.compile(r"(key|token|secret|password|auth|credential)", re.IGNORECASE)
 def resolve_settings_path() -> Path:
    repo_root = Path(__file__).parent.parent.resolve()
    portable = repo_root / "settings.json"
    if portable.exists():
        payload = load_json(portable)
        if isinstance(payload, dict) and payload.get("use_portable_settings") is True:
            return portable
    return Path(user_config_dir(APP_NAME, appauthor=False)) / "settings.json"
 def load_json(path: Path) -> dict[str, Any]:
    try:
        with path.open("r", encoding="utf-8") as f:
            return json.load(f)
    except FileNotFoundError:
        return {}
    except json.JSONDecodeError as exc:
        logger.error(f"Invalid JSON in {path}: {exc}")
        return {}
    except OSError as exc:
        logger.error(f"Cannot read {path}: {exc}")
        return {}
 def expand_path(value: str) -> str:
    return str(Path(value).expanduser().resolve(strict=False))
 def normalize_path_list(value: Any) -> list[str]:
    if isinstance(value, str):
        return [expand_path(value)] if value else []
    if isinstance(value, list):
        return [expand_path(item) for item in value if isinstance(item, str) and item]
    return []
 def dedupe(values: list[str]) -> list[str]:
    seen: set[str] = set()
    result: list[str] = []
    for value in values:
        if value not in seen:
            result.append(value)
            seen.add(value)
    return result
 def get_model_roots(settings: dict[str, Any]) -> dict[str, list[str]]:
    roots: dict[str, list[str]] = {}
    active_library = settings.get("active_library") or "default"
    sources = [settings]
    library = settings.get("libraries", {}).get(active_library)
    if isinstance(library, dict):
        sources.insert(0, library)
    for source in sources:
        folder_paths = source.get("folder_paths")
        if isinstance(folder_paths, dict):
            for key, value in folder_paths.items():
                roots.setdefault(key, []).extend(normalize_path_list(value))
    for default_key, folder_key in (
        ("default_lora_root", "loras"),
        ("default_checkpoint_root", "checkpoints"),
        ("default_embedding_root", "embeddings"),
        ("default_unet_root", "unet"),
    ):
        value = settings.get(default_key)
        if isinstance(value, str) and value:
            roots.setdefault(folder_key, []).append(expand_path(value))
    return {key: dedupe(values) for key, values in roots.items()}
 def find_model_files(directory: Path) -> list[Path]:
    model_files = []
    for ext in MODEL_EXTENSIONS:
        model_files.extend(directory.rglob(f"*{ext}"))
    return model_files
 def find_legacy_metadata(model_path: Path) -> Path | None:
    base_name = model_path.stem
    legacy_path = model_path.with_name(f"{base_name}.json")
    if legacy_path.exists() and legacy_path.is_file():
        return legacy_path
    return None
 def load_legacy_metadata(legacy_path: Path) -> dict[str, Any] | None:
    try:
        with open(legacy_path, "r", encoding="utf-8") as f:
            return json.load(f)
    except json.JSONDecodeError as e:
        logger.error(f"Invalid JSON in legacy file {legacy_path}: {e}")
        return None
    except Exception as e:
        logger.error(f"Error reading legacy file {legacy_path}: {e}")
        return None
 def load_metadata(metadata_path: Path) -> dict[str, Any]:
    if not metadata_path.exists():
        return {}
    try:
        with open(metadata_path, "r", encoding="utf-8") as f:
            return json.load(f)
    except json.JSONDecodeError as e:
        logger.warning(f"Invalid JSON in metadata file {metadata_path}: {e}. Starting fresh.")
        return {}
    except Exception as e:
        logger.error(f"Error reading metadata file {metadata_path}: {e}")
        return {}
 def save_metadata(metadata_path: Path, data: dict[str, Any], dry_run: bool = False) -> bool:
    if dry_run:
        logger.info(f"[DRY RUN] Would save metadata to: {metadata_path}")
        return True
    temp_path = metadata_path.with_suffix(".tmp")
    try:
        with open(temp_path, "w", encoding="utf-8") as f:
            json.dump(data, f, indent=2, ensure_ascii=False)
        os.replace(temp_path, metadata_path)
        return True
    except Exception as e:
        logger.error(f"Error saving metadata to {metadata_path}: {e}")
        if temp_path.exists():
            try:
                temp_path.unlink()
            except:
                pass
        return False
 def migrate_metadata(
    legacy_data: dict[str, Any],
    existing_metadata: dict[str, Any],
    model_type: str
 ) -> dict[str, Any] | None:
    metadata = existing_metadata.copy()
    changes_made = False
    if "civitai" not in metadata:
        metadata["civitai"] = {}
    activation_text = legacy_data.get("activation text")
    if activation_text and isinstance(activation_text, str):
        trigger_words = [
            word.strip()
            for word in activation_text.replace("\n", ",").split(",")
            if word.strip()
        ]
        if trigger_words:
            existing_trained = metadata["civitai"].get("trainedWords", [])
            if not isinstance(existing_trained, list):
                existing_trained = []
            merged = list(dict.fromkeys(existing_trained + trigger_words))
            if merged != existing_trained:
                metadata["civitai"]["trainedWords"] = merged
                changes_made = True
                logger.debug(f"  Migrated activation text: {trigger_words}")
    if model_type == "lora":
        preferred_weight = legacy_data.get("preferred weight")
        if preferred_weight is not None:
            try:
                weight_value = float(preferred_weight)
                usage_tips_str = metadata.get("usage_tips", "{}")
                if isinstance(usage_tips_str, str):
                    try:
                        usage_tips = json.loads(usage_tips_str)
                    except json.JSONDecodeError:
                        usage_tips = {}
                elif isinstance(usage_tips_str, dict):
                    usage_tips = usage_tips_str
                else:
                    usage_tips = {}
                if "strength" not in usage_tips:
                    usage_tips["strength"] = weight_value
                    metadata["usage_tips"] = json.dumps(usage_tips, ensure_ascii=False)
                    changes_made = True
                    logger.debug(f"  Migrated preferred weight: {weight_value}")
            except (ValueError, TypeError) as e:
                logger.warning(f"  Could not parse preferred weight '{preferred_weight}': {e}")
    else:
        if legacy_data.get("preferred weight") is not None:
            logger.debug("  Skipping 'preferred weight' for non-LoRA model")
    notes = legacy_data.get("notes")
    if notes and isinstance(notes, str) and notes.strip():
        existing_notes = metadata.get("notes", "")
        if not existing_notes:
            metadata["notes"] = notes.strip()
            changes_made = True
            logger.debug("  Migrated notes")
        elif notes.strip() not in existing_notes:
            metadata["notes"] = f"{existing_notes}\n\n{notes.strip()}".strip()
            changes_made = True
            logger.debug("  Appended notes")
    return metadata if changes_made else None
 def process_model(model_path: Path, model_type: str, dry_run: bool = False) -> bool:
    legacy_path = find_legacy_metadata(model_path)
    if not legacy_path:
        return True
    logger.info(f"Processing: {model_path.name} ({model_type})")
    logger.info(f"  Found legacy metadata: {legacy_path.name}")
    legacy_data = load_legacy_metadata(legacy_path)
    if legacy_data is None:
        return False
    metadata_path = model_path.with_suffix(".metadata.json")
    existing_metadata = load_metadata(metadata_path)
    migrated = migrate_metadata(legacy_data, existing_metadata, model_type)
    if migrated is None:
        logger.info("  No changes needed (fields already exist or no migratable data)")
        return True
    if save_metadata(metadata_path, migrated, dry_run):
        logger.info(f"  ✓ Successfully migrated metadata to: {metadata_path.name}")
        return True
    else:
        logger.error("  ✗ Failed to save metadata")
        return False
 def main() -> int:
    parser = argparse.ArgumentParser(
        description="Migrate legacy metadata JSON files to LoRA Manager's metadata.json format.",
        formatter_class=argparse.RawDescriptionHelpFormatter,
        epilog="""
 Examples:
  python scripts/migrate_legacy_metadata.py
  python scripts/migrate_legacy_metadata.py --dry-run
  python scripts/migrate_legacy_metadata.py --verbose
        """
    )
    parser.add_argument(
        "--dry-run",
        action="store_true",
        help="Preview changes without modifying any files"
    )
    parser.add_argument(
        "-v", "--verbose",
        action="store_true",
        help="Enable verbose output"
    )
    args = parser.parse_args()
    if args.verbose:
        logging.getLogger().setLevel(logging.DEBUG)
    settings_path = resolve_settings_path()
    logger.info(f"Using settings: {settings_path}")
    settings = load_json(settings_path)
    if not settings:
        logger.error("Could not load settings.json. Please ensure LoRA Manager is configured.")
        return 1
    roots = get_model_roots(settings)
    if not roots:
        logger.error("No model folders configured in settings.json.")
        return 1
    lora_roots = roots.get("loras", [])
    checkpoint_roots = roots.get("checkpoints", []) + roots.get("unet", [])
    all_roots = []
    for root_list in [lora_roots, checkpoint_roots]:
        for root in root_list:
            path = Path(root)
            if path.exists() and path.is_dir():
                all_roots.append((path, "lora" if root in lora_roots else "checkpoint"))
    if not all_roots:
        logger.error("No valid model folders found.")
        return 1
    logger.info(f"Found {len(lora_roots)} LoRA root(s), {len(checkpoint_roots)} Checkpoint root(s)")
    processed = 0
    migrated = 0
    errors = 0
    skipped = 0
    lora_count = 0
    checkpoint_count = 0
    for root_path, model_type in all_roots:
        logger.info(f"Scanning: {root_path} ({model_type})")
        model_files = find_model_files(root_path)
        logger.debug(f"  Found {len(model_files)} model files")
        for model_path in model_files:
            legacy_path = find_legacy_metadata(model_path)
            if not legacy_path:
                skipped += 1
                continue
            processed += 1
            if process_model(model_path, model_type, dry_run=args.dry_run):
                migrated += 1
                if model_type == "lora":
                    lora_count += 1
                else:
                    checkpoint_count += 1
            else:
                errors += 1
    logger.info("\n" + "=" * 50)
    logger.info("Migration Summary:")
    logger.info(f"  Models with legacy metadata: {processed}")
    logger.info(f"  Successfully migrated: {migrated}")
    logger.info(f"    - LoRA models: {lora_count}")
    logger.info(f"    - Checkpoint models: {checkpoint_count}")
    logger.info(f"  Errors: {errors}")
    logger.info(f"  Skipped (no legacy file): {skipped}")
    if args.dry_run:
        logger.info("\n  [DRY RUN MODE - No files were modified]")
    return 0 if errors == 0 else 1
 if __name__ == "__main__":
    sys.exit(main())
--- a/scripts/restore_suffixed_filenames.py
+++ b/scripts/restore_suffixed_filenames.py
@@ -0,0 +1,403 @@
 #!/usr/bin/env python3
 """
 Restore original filenames by removing leftover 4-char hash suffixes.
 When LoRA Manager's old duplicate filename resolver ran, it appended
 ``-{first4ofSHA256}`` to duplicate filenames, e.g.::
    my_lora.safetensors  →  my_lora-a3f7.safetensors
 With full-path LoRA syntax now available (``<lora:subfolder/name:1.0>``),
 these suffixes are unnecessary.  This script detects such files and, with
 your confirmation, restores their original names.
 The same suffix pattern is also used by the download conflict handler
 (``{name}-{hash}.{ext}``).  To avoid false positives, this script skips
 any file whose original name already exists in the same directory — those
 were likely added by a download conflict, not the old resolver.
 Usage::
    # Detect only (dry-run, default)
    python scripts/restore_suffixed_filenames.py
    # Detect + restore (with confirmation prompt)
    python scripts/restore_suffixed_filenames.py --apply
 After restoring filenames, run **Rebuild Cache** in the LoRA Manager
 Doctor panel to refresh the model cache.
 """
 from __future__ import annotations
 import argparse
 import json
 import logging
 import os
 import re
 import sys
 from pathlib import Path
 from typing import Any
 from platformdirs import user_config_dir
 logging.basicConfig(
    level=logging.INFO,
    format="%(message)s",
 )
 logger = logging.getLogger(__name__)
 APP_NAME = "ComfyUI-LoRA-Manager"
 MODEL_EXTENSIONS = {".safetensors", ".ckpt", ".pt", ".pth", ".bin"}
 PREVIEW_EXTENSIONS = {
    ".png", ".jpg", ".jpeg", ".webp", ".gif", ".bmp",
    ".mp4", ".webm", ".mov",
 }
 # Matches filenames like "my_lora-a3f7.safetensors"
 # Groups: (base_name, 4-char-hex, extension)
 _SUFFIX_RE = re.compile(r"^(.+)-([0-9a-f]{4})(\.[^.]+)$")
 # ── helpers (copied from migrate_legacy_metadata.py for consistency) ──────────
 def resolve_settings_path() -> Path:
    repo_root = Path(__file__).parent.parent.resolve()
    portable = repo_root / "settings.json"
    if portable.exists():
        payload = _load_json(portable)
        if isinstance(payload, dict) and payload.get("use_portable_settings") is True:
            return portable
    return Path(user_config_dir(APP_NAME, appauthor=False)) / "settings.json"
 def _load_json(path: Path) -> dict[str, Any]:
    try:
        with path.open("r", encoding="utf-8") as f:
            return json.load(f)
    except (FileNotFoundError, json.JSONDecodeError, OSError):
        return {}
 def _expand_path(value: str) -> str:
    return str(Path(value).expanduser().resolve(strict=False))
 def _normalize_path_list(value: Any) -> list[str]:
    if isinstance(value, str):
        return [_expand_path(value)] if value else []
    if isinstance(value, list):
        return [_expand_path(item) for item in value if isinstance(item, str) and item]
    return []
 def _dedupe(values: list[str]) -> list[str]:
    seen: set[str] = set()
    result: list[str] = []
    for value in values:
        if value not in seen:
            result.append(value)
            seen.add(value)
    return result
 def get_model_roots(settings: dict[str, Any]) -> dict[str, list[str]]:
    """Extract model folder roots from LoRA Manager settings.
    Returns ``{model_type: [path, ...]}`` where *model_type* is one of
    ``loras``, ``checkpoints``, ``embeddings``, ``unet``, etc.
    Both primary (``folder_paths``) and extra (``extra_folder_paths``)
    paths are included.  Extra paths can be configured via the UI at
    Settings → Model Libraries → Extra Folder Paths.
    """
    roots: dict[str, list[str]] = {}
    active_library = settings.get("active_library") or "default"
    sources = [settings]
    library = settings.get("libraries", {}).get(active_library)
    if isinstance(library, dict):
        sources.insert(0, library)
    for source in sources:
        # Primary folder paths.
        folder_paths = source.get("folder_paths")
        if isinstance(folder_paths, dict):
            for key, value in folder_paths.items():
                roots.setdefault(key, []).extend(_normalize_path_list(value))
        # Extra folder paths (Settings → Model Libraries → Extra Folder Paths).
        extra_folder_paths = source.get("extra_folder_paths")
        if isinstance(extra_folder_paths, dict):
            for key, value in extra_folder_paths.items():
                roots.setdefault(key, []).extend(_normalize_path_list(value))
    for default_key, folder_key in (
        ("default_lora_root", "loras"),
        ("default_checkpoint_root", "checkpoints"),
        ("default_unet_root", "unet"),
        ("default_embedding_root", "embeddings"),
    ):
        value = settings.get(default_key)
        if isinstance(value, str) and value:
            roots.setdefault(folder_key, []).append(_expand_path(value))
    return {key: _dedupe(values) for key, values in roots.items()}
 def find_model_files(directory: Path) -> list[Path]:
    """Recursively find all model files in *directory*."""
    files: list[Path] = []
    for ext in MODEL_EXTENSIONS:
        files.extend(directory.rglob(f"*{ext}"))
    return files
 # ── core detection logic ──────────────────────────────────────────────────────
 def check_file(path: Path) -> tuple[str, str, str] | None:
    """If *path* matches the suffix pattern, return ``(base_name, hex, ext)``.
    Returns ``None`` when:
    * The filename does not match the pattern, or
    * The original name (without the suffix) already exists in the same
      directory (likely a download-conflict rename, not a doctor rename).
    """
    match = _SUFFIX_RE.match(path.name)
    if not match:
        return None
    base_name = match.group(1)
    hex_part = match.group(2)
    extension = match.group(3)
    orig_name = base_name + extension
    orig_path = path.with_name(orig_name)
    # Safety: skip if the original name already exists.
    if orig_path.exists():
        return None
    return base_name, hex_part, extension
 def scan_roots(
    roots: dict[str, list[str]],
 ) -> dict[str, list[tuple[Path, str, str, str]]]:
    """Scan all model roots and return detected files grouped by model type.
    Returns ``{model_type: [(full_path, base_name, hex, ext), ...]}``.
    """
    results: dict[str, list[tuple[Path, str, str, str]]] = {}
    for model_type, root_list in roots.items():
        type_results: list[tuple[Path, str, str, str]] = []
        for root in root_list:
            root_path = Path(root)
            if not root_path.is_dir():
                continue
            for model_file in find_model_files(root_path):
                match = check_file(model_file)
                if match:
                    type_results.append((model_file, *match))
        if type_results:
            results[model_type] = type_results
    return results
 def rename_file(
    path: Path, base_name: str, extension: str, dry_run: bool
 ) -> bool:
    """Rename *path* to ``{base_name}{extension}``.
    Also renames sidecar files (``.metadata.json``, ``.civitai.info``) and
    preview images.  Returns ``True`` on success.
    """
    new_path = path.with_name(base_name + extension)
    old_stem = path.with_suffix("")  # /dir/base_name-hex  (no ext)
    new_stem = new_path.with_suffix("")  # /dir/base_name  (no ext)
    if dry_run:
        logger.info("  would rename:  %s", path.name)
        logger.info("             ->  %s", new_path.name)
        return True
    try:
        os.rename(path, new_path)
    except OSError as exc:
        logger.error("  FAILED to rename %s: %s", path.name, exc)
        return False
    # Rename sidecar metadata files.
    for suffix in (".metadata.json", ".civitai.info"):
        old_sidecar = old_stem.with_name(old_stem.name + suffix)
        new_sidecar = new_stem.with_name(new_stem.name + suffix)
        if old_sidecar.exists():
            try:
                os.rename(old_sidecar, new_sidecar)
            except OSError as exc:
                logger.warning("  could not rename sidecar %s: %s", old_sidecar.name, exc)
    # Rename preview images.
    for preview_ext in PREVIEW_EXTENSIONS:
        old_preview = old_stem.with_name(old_stem.name + preview_ext)
        new_preview = new_stem.with_name(new_stem.name + preview_ext)
        if old_preview.exists():
            try:
                os.rename(old_preview, new_preview)
            except OSError as exc:
                logger.warning("  could not rename preview %s: %s", old_preview.name, exc)
    logger.info("  renamed: %s  ->  %s", path.name, new_path.name)
    return True
 # ── report helpers ────────────────────────────────────────────────────────────
 def print_report(results: dict[str, list[tuple[Path, str, str, str]]]) -> int:
    """Print a human-readable report of detected files.  Returns total count."""
    if not results:
        logger.info("No leftover suffixed filenames detected.")
        return 0
    total = 0
    for model_type in sorted(results):
        entries = results[model_type]
        total += len(entries)
        label = model_type.capitalize()
        logger.info("")
        logger.info("─" * 50)
        logger.info("  %s  (%d file(s))", label, len(entries))
        logger.info("─" * 50)
        for path, base_name, hex_part, ext in sorted(entries):
            logger.info("  %s  →  %s%s", path.name, base_name, ext)
    logger.info("")
    logger.info("=" * 50)
    logger.info("  Total: %d file(s) with leftover suffixes.", total)
    logger.info("=" * 50)
    return total
 def prompt_user(count: int) -> bool:
    """Ask the user whether to proceed with the rename."""
    try:
        answer = input(
            f"\nRestore {count} file(s) to their original names? [y/N] "
        ).strip().lower()
    except (EOFError, KeyboardInterrupt):
        print()
        return False
    return answer in ("y", "yes")
 # ── main ──────────────────────────────────────────────────────────────────────
 def main() -> int:
    parser = argparse.ArgumentParser(
        description=(
            "Detect and restore model filenames that have leftover "
            "4-character hash suffixes from the old conflict resolver."
        ),
        formatter_class=argparse.RawDescriptionHelpFormatter,
        epilog=(
            "Examples:\n"
            "  python scripts/restore_suffixed_filenames.py\n"
            "  python scripts/restore_suffixed_filenames.py --apply\n"
            "  python scripts/restore_suffixed_filenames.py --apply --yes\n"
        ),
    )
    parser.add_argument(
        "--apply",
        action="store_true",
        help="Actually rename files (with confirmation prompt unless --yes is given)",
    )
    parser.add_argument(
        "--yes", "-y",
        action="store_true",
        help="Skip confirmation prompt (implies --apply)",
    )
    parser.add_argument(
        "--dry-run",
        action="store_true",
        help="Detect only — show what would be renamed without making changes",
    )
    parser.add_argument(
        "-v", "--verbose",
        action="store_true",
        help="Enable debug-level logging",
    )
    args = parser.parse_args()
    if args.verbose:
        logging.getLogger().setLevel(logging.DEBUG)
    # Resolve settings.
    settings_path = resolve_settings_path()
    logger.info("Settings: %s", settings_path)
    settings = _load_json(settings_path)
    if not settings:
        logger.error("Could not load settings.json.  Is LoRA Manager configured?")
        return 1
    roots = get_model_roots(settings)
    if not roots:
        logger.error("No model folders found in settings.")
        return 1
    # Log which roots are being scanned.
    for model_type, root_list in roots.items():
        for root in root_list:
            logger.info("Scanning %s: %s", model_type, root)
    # Detect.
    results = scan_roots(roots)
    total = print_report(results)
    if total == 0:
        return 0
    # Determine mode.
    dry_run = not args.apply and not args.yes
    if dry_run:
        logger.info("\n[Dry-run mode — no files modified]")
        logger.info("Run with --apply to restore filenames.")
        return 0
    # Confirm unless --yes.
    if not args.yes:
        if not prompt_user(total):
            logger.info("Aborted.")
            return 0
    # Rename.
    logger.info("")
    success = 0
    fail = 0
    for model_type in sorted(results):
        entries = results[model_type]
        logger.info("")
        logger.info("─" * 50)
        logger.info("  Restoring %s  (%d file(s))", model_type, len(entries))
        logger.info("─" * 50)
        for path, base_name, hex_part, ext in sorted(entries):
            ok = rename_file(path, base_name, ext, dry_run=False)
            if ok:
                success += 1
            else:
                fail += 1
    logger.info("")
    logger.info("=" * 50)
    logger.info("  Done: %d restored, %d failed.", success, fail)
    logger.info("=" * 50)
    logger.info("")
    logger.info("  ⚠  Please run Rebuild Cache in the LoRA Manager")
    logger.info("     Doctor panel to refresh the model cache.")
    return 0 if fail == 0 else 1
 if __name__ == "__main__":
    sys.exit(main())
--- a/settings.json.example
+++ b/settings.json.example
@@ -10,13 +10,14 @@
      "C:/path/to/your/checkpoints_folder",
      "C:/path/to/another/checkpoints_folder"
    ],
    "unet": [
      "C:/path/to/your/diffusion_models_folder",
      "C:/path/to/another/diffusion_models_folder"
    ],
    "embeddings": [
      "C:/path/to/your/embeddings_folder",
      "C:/path/to/another/embeddings_folder"
    ]
  },
  "example_images_open_mode": "system",
  "example_images_local_root": "",
  "example_images_open_uri_template": "",
  "auto_organize_exclusions": []
 }
--- a/static/css/base.css
+++ b/static/css/base.css
@@ -1,21 +1,20 @@
@import 'tokens/index.css';
 html,
 body {
    margin: 0;
    padding: 0;
    height: 100%;
    overflow: hidden;
    /* Disable default scrolling */
 }
 /* 针对Firefox */
 * {
    scrollbar-width: thin;
-    scrollbar-color: var(--border-color) transparent;
+    scrollbar-color: var(--border-base) transparent;
 }
 /* 针对Webkit browsers (Chrome, Safari等) */
 ::-webkit-scrollbar {
-    width: 8px;
+    width: var(--scrollbar-width, 8px);
 }
 ::-webkit-scrollbar-track {
@@ -24,116 +23,128 @@ body {
 }
 ::-webkit-scrollbar-thumb {
-    background-color: var(--border-color);
+    background-color: var(--border-base);
-    border-radius: 4px;
+    border-radius: var(--radius-xs);
 }
 :root {
    --bg-color: #ffffff;
    --text-color: #333333;
    --text-muted: #6c757d;
    --card-bg: #ffffff;
    --border-color: #e0e0e0;
    --header-height: 48px;
    /* Color Components */
    --lora-accent-l: 68%;
    --lora-accent-c: 0.28;
    --lora-accent-h: 256;
    --lora-warning-l: 75%;
    --lora-warning-c: 0.25;
    --lora-warning-h: 80;
    --lora-success-l: 70%;
    --lora-success-c: 0.2;
    --lora-success-h: 140;
    /* Composed Colors */
    --lora-accent: oklch(var(--lora-accent-l) var(--lora-accent-c) var(--lora-accent-h));
    --lora-surface: oklch(97% 0 0 / 0.95);
    --lora-border: oklch(72% 0.03 256 / 0.45);
    --lora-text: oklch(95% 0.02 256);
    --lora-error: oklch(75% 0.32 29);
    --lora-error-bg: color-mix(in oklch, var(--lora-error) 20%, transparent);
    --lora-error-border: color-mix(in oklch, var(--lora-error) 50%, transparent);
    --lora-warning: oklch(var(--lora-warning-l) var(--lora-warning-c) var(--lora-warning-h));
    --lora-success: oklch(var(--lora-success-l) var(--lora-success-c) var(--lora-success-h));
    --badge-update-bg: oklch(72% 0.2 220);
    --badge-update-text: oklch(28% 0.03 220);
    --badge-update-glow: oklch(72% 0.2 220 / 0.28);
    --badge-skip-refresh-bg: oklch(82% 0.12 45);
    --badge-skip-refresh-text: oklch(35% 0.02 45);
    --badge-skip-refresh-glow: oklch(82% 0.12 45 / 0.15);
    /* Spacing Scale */
    --space-1: calc(8px * 1);
    --space-2: calc(8px * 2);
    --space-3: calc(8px * 3);
    --space-4: calc(8px * 4);
    /* Z-index Scale */
    --z-base: 10;
    --z-header: 100;
    --z-modal: 1000;
    --z-overlay: 2000;
    /* Border Radius */
    --border-radius-base: 12px;
    --border-radius-md: 12px;
    --border-radius-sm: 8px;
    --border-radius-xs: 4px;
    --scrollbar-width: 8px;
    /* 添加滚动条宽度变量 */
-    /* Shortcut styles */
+    --shortcut-bg: var(--color-accent-subtle);
-    --shortcut-bg: oklch(var(--lora-accent-l) var(--lora-accent-c) var(--lora-accent-h) / 0.12);
+    --shortcut-border: var(--color-accent-border);
-    --shortcut-border: oklch(var(--lora-accent-l) var(--lora-accent-c) var(--lora-accent-h) / 0.25);
+    --shortcut-text: var(--text-primary);
-    --shortcut-text: var(--text-color);
+
    --lora-accent-transparent: var(--color-accent-transparent);
    /* Legacy spacing aliases: 8px base grid to match existing component usage */
    --space-1: 8px;
    --space-2: 16px;
    --space-3: 24px;
    --space-4: 32px;
    /* Legacy border-radius aliases to match existing component usage */
    --border-radius-xs: 4px;
    --border-radius-sm: 6px;
    --border-radius-base: 8px;
    --border-radius-md: 12px;
    --border-radius-lg: 16px;
 }
 :root {
    --bg-color: var(--bg-base);
    --text-color: var(--text-primary);
    --text-muted: var(--text-secondary);
    --card-bg: var(--surface-base);
    --border-color: var(--border-base);
    --lora-accent: var(--color-accent);
    --lora-surface: var(--bg-elevated);
    --lora-border: var(--border-subtle);
    --lora-text: var(--text-primary);
    --lora-error: var(--color-error);
    --lora-error-bg: var(--color-error-bg);
    --lora-error-border: var(--color-error-border);
    --lora-warning: var(--color-warning);
    --lora-success: var(--color-success);
    --badge-update-bg: var(--color-info-bg);
    --badge-update-text: var(--color-info-text);
    --badge-update-glow: var(--color-info-glow);
    --badge-skip-refresh-bg: var(--color-skip-refresh-bg);
    --badge-skip-refresh-text: var(--color-skip-refresh-text);
    --badge-skip-refresh-glow: var(--color-skip-refresh-glow);
 }
 [data-theme="dark"] {
    --bg-color: var(--bg-base);
    --text-color: var(--text-primary);
    --text-muted: var(--text-secondary);
    --card-bg: var(--surface-base);
    --border-color: var(--border-base);
    --lora-accent: var(--color-accent);
    --lora-surface: var(--bg-elevated);
    --lora-border: var(--border-subtle);
    --lora-text: var(--text-primary);
    --lora-error: var(--color-error);
    --lora-error-bg: var(--color-error-bg);
    --lora-error-border: var(--color-error-border);
    --lora-warning: var(--color-warning);
    --lora-success: var(--color-success);
    --badge-update-bg: var(--color-info-bg);
    --badge-update-text: var(--color-info-text);
    --badge-update-glow: var(--color-info-glow);
    --badge-skip-refresh-bg: var(--color-skip-refresh-bg);
    --badge-skip-refresh-text: var(--color-skip-refresh-text);
    --badge-skip-refresh-glow: var(--color-skip-refresh-glow);
 }
 html[data-theme="dark"] {
-    background-color: #1a1a1a !important;
+    background-color: var(--bg-base) !important;
    color-scheme: dark;
 }
 html[data-theme="light"] {
-    background-color: #ffffff !important;
+    background-color: var(--bg-base) !important;
    color-scheme: light;
 }
 [data-theme="dark"] {
    --bg-color: #1a1a1a;
    --text-color: #e0e0e0;
    --text-muted: #a0a0a0;
    --card-bg: #2d2d2d;
    --border-color: #404040;
    --lora-accent: oklch(68% 0.28 256);
    --lora-surface: oklch(25% 0.02 256 / 0.98);
    --lora-border: oklch(90% 0.02 256 / 0.15);
    --lora-text: oklch(98% 0.02 256);
    --lora-warning: oklch(75% 0.25 80);
    /* Modified to be used with oklch() */
    --lora-error-bg: color-mix(in oklch, var(--lora-error) 15%, transparent);
    --lora-error-border: color-mix(in oklch, var(--lora-error) 40%, transparent);
    --badge-update-bg: oklch(62% 0.18 220);
    --badge-update-text: oklch(98% 0.02 240);
    --badge-update-glow: oklch(62% 0.18 220 / 0.4);
    --badge-skip-refresh-bg: oklch(82% 0.12 45);
    --badge-skip-refresh-text: oklch(98% 0.02 45);
    --badge-skip-refresh-glow: oklch(82% 0.12 45 / 0.15);
 }
 body {
-    font-family: 'Segoe UI', sans-serif;
+    font-family: var(--font-body);
-    background: var(--bg-color);
+    background: var(--bg-base);
-    color: var(--text-color);
+    color: var(--text-primary);
    display: flex;
    flex-direction: column;
    padding-top: 0;
    /* Remove the padding-top */
 }
 .hidden {
    display: none !important;
-}
+}
 :focus-visible {
    outline: 2px solid var(--color-accent);
    outline-offset: 2px;
 }
 button:focus:not(:focus-visible),
 input:focus:not(:focus-visible),
 select:focus:not(:focus-visible) {
    outline: none;
 }
@media (prefers-reduced-motion: reduce) {
    *,
    *::before,
    *::after {
        animation-duration: 0.01ms !important;
        animation-iteration-count: 1 !important;
        transition-duration: 0.01ms !important;
    }
    html {
        scroll-behavior: auto !important;
    }
 }
--- a/static/css/components/alphabet-bar.css
+++ b/static/css/components/alphabet-bar.css
@@ -46,7 +46,7 @@
    flex-direction: column;
    gap: 6px;
    align-items: center;
-    box-shadow: 2px 0 8px rgba(0, 0, 0, 0.1);
+    box-shadow: var(--shadow-side);
    max-height: 80vh;
    overflow-y: auto;
    scrollbar-width: thin;
@@ -75,7 +75,7 @@
    width: 20px;
    height: 40px;
    align-self: center;
-    box-shadow: 2px 0 8px rgba(0, 0, 0, 0.1);
+    box-shadow: var(--shadow-side);
 }
 .toggle-alphabet-bar:hover {
@@ -99,7 +99,7 @@
    min-width: 24px;
    text-align: center;
    font-size: 0.85em;
-    transition: all 0.2s ease;
+    transition: var(--transition-base);
    border: 1px solid var(--border-color);
 }
@@ -107,7 +107,7 @@
    background: var(--lora-accent);
    color: white;
    transform: scale(1.1);
-    box-shadow: 0 2px 4px rgba(0, 0, 0, 0.1);
+    box-shadow: var(--shadow-md);
 }
 .letter-chip.active {
--- a/static/css/components/banner.css
+++ b/static/css/components/banner.css
@@ -68,7 +68,7 @@
    text-decoration: none;
    font-size: 0.85em;
    font-weight: 500;
-    transition: all 0.2s ease;
+    transition: var(--transition-base);
    white-space: nowrap;
    border: 1px solid transparent;
 }
@@ -102,7 +102,7 @@
    color: white;
    border-color: var(--lora-accent);
    transform: translateY(-1px);
-    box-shadow: 0 2px 4px rgba(0, 0, 0, 0.1);
+    box-shadow: var(--shadow-md);
 }
 /* Tertiary Action Button */
@@ -133,7 +133,7 @@
    display: flex;
    align-items: center;
    justify-content: center;
-    transition: all 0.2s ease;
+    transition: var(--transition-base);
    font-size: 0.8em;
 }
--- a/static/css/components/batch-import-modal.css
+++ b/static/css/components/batch-import-modal.css
@@ -76,7 +76,7 @@
    display: flex;
    align-items: center;
    justify-content: center;
-    transition: all 0.2s;
+    transition: var(--transition-base);
    background: var(--bg-color);
 }
@@ -87,7 +87,7 @@
 .checkbox-label input[type="checkbox"]:checked + .checkmark::after {
    content: '\f00c';
-    font-family: 'Font Awesome 6 Free';
+    font-family: 'Font Awesome 6 Free', sans-serif;
    font-weight: 900;
    color: var(--lora-text);
    font-size: 12px;
@@ -166,7 +166,7 @@
    background: var(--card-bg);
    color: var(--text-color);
    cursor: pointer;
-    transition: all 0.2s;
+    transition: var(--transition-base);
 }
 .back-btn:hover {
@@ -237,7 +237,7 @@
    padding: 8px 10px;
    border-radius: var(--border-radius-xs);
    cursor: pointer;
-    transition: all 0.2s;
+    transition: var(--transition-base);
    border: 1px solid transparent;
 }
--- a/static/css/components/card.css
+++ b/static/css/components/card.css
@@ -1,12 +1,12 @@
-/* 卡片网格布局 */
+/* Card grid layout */
 .card-grid {
    display: grid;
    grid-template-columns: repeat(auto-fill, minmax(260px, 1fr)); /* Base size */
    gap: 12px; /* Consistent gap for both row and column spacing */
    row-gap: 20px; /* Increase vertical spacing between rows */
    margin-top: var(--space-2);
-    padding-top: 4px; /* 添加顶部内边距，为悬停动画提供空间 */
+    padding-top: 4px;
-    padding-bottom: 4px; /* 添加底部内边距，为悬停动画提供空间 */
+    padding-bottom: 4px;
    width: 100%; /* Ensure it takes full width of container */
    max-width: 1400px; /* Base container width */
    margin-left: auto;
@@ -19,9 +19,10 @@
    border: 1px solid var(--lora-border);
    border-radius: var(--border-radius-base);
    backdrop-filter: blur(16px);
-    transition: transform 160ms ease-out;
+    transition: transform var(--transition-fast) ease-out, box-shadow var(--transition-fast) ease-out, border-color var(--transition-fast) ease-out;
    aspect-ratio: 896/1152; /* Preserve aspect ratio */
    max-width: 260px; /* Base size */
    min-width: 200px; /* Prevent cards from becoming too narrow */
    width: 100%;
    margin: 0 auto;
    cursor: pointer;
@@ -32,7 +33,8 @@
 .model-card:hover {
    transform: translateY(-2px);
-    background: oklch(100% 0 0 / 0.6);
+    box-shadow: var(--shadow-md);
    border-color: var(--lora-accent);
 }
 .model-card:focus-visible {
@@ -328,7 +330,6 @@
 }
 .card-actions i {
    margin-left: var(--space-1);
    cursor: pointer;
    color: white;
    transition: opacity 0.2s, transform 0.15s ease;
@@ -353,24 +354,38 @@
 }
 .card-actions {
    flex-shrink: 0;
    display: flex;
-    gap: var(--space-1); /* Use gap instead of margin for spacing between icons */
+    gap: var(--space-1);
-    align-items: center;
+    align-items: flex-end;
    align-self: flex-end;
 }
-.card-actions i:hover {
+.card-actions i:hover,
 .card-actions i:focus-visible {
    opacity: 0.9;
    transform: scale(1.1);
    background-color: rgba(255, 255, 255, 0.1);
    outline: 2px solid var(--lora-accent);
    outline-offset: 2px;
    border-radius: var(--border-radius-xs);
 }
 /* Style for active favorites */
 .favorite-active {
-    color: #ffc107 !important; /* Gold color for favorites */
+    color: var(--favorite-color) !important;
-    text-shadow: 0 0 5px rgba(255, 193, 7, 0.5);
+    text-shadow: 0 0 5px var(--favorite-glow);
 }
-/* 响应式设计 */
+@media (max-width: 1200px) {
    .card-grid {
        grid-template-columns: repeat(auto-fill, minmax(220px, 1fr));
    }
    .model-card {
        max-width: 240px;
        min-width: 180px;
    }
 }
@media (max-width: 768px) {
    .card-grid {
        grid-template-columns: minmax(260px, 1fr); /* Adjusted minimum size for mobile */
@@ -378,17 +393,10 @@
    .model-card {
        max-width: 100%; /* Allow cards to fill available space on mobile */
        min-width: 200px;
    }
 }
 .card-actions {
    flex-shrink: 0;  /* Prevent actions from shrinking */
    display: flex;
    gap: var(--space-1);
    align-items: flex-end; /* 将图标靠下对齐 */
    align-self: flex-end; /* 将整个actions容器靠下对齐 */
 }
 .model-link {
    margin-top: var(--space-1);
 }
@@ -401,9 +409,13 @@
    text-shadow: none;
 }
-.model-link a:hover {
+.model-link a:hover,
 .model-link a:focus-visible {
    opacity: 0.8;
    text-decoration: none;
    outline: 2px solid var(--lora-accent);
    outline-offset: 2px;
    border-radius: var(--border-radius-xs);
 }
 /* Updated model name to fix text cutoff issues */
@@ -428,7 +440,7 @@
 .base-model {
    display: inline-block;
-    background: #f0f0f0;
+    background: var(--surface-hover, oklch(95% 0 0));
    padding: 2px 6px;
    border-radius: var(--border-radius-xs);
    margin-right: 6px;
@@ -497,21 +509,96 @@
    background: rgba(0,0,0,0.18); /* Optional: subtle background for contrast */
 }
 /* Version row — flex container for badges + version names */
 .version-row {
    display: flex;
    flex-wrap: wrap;
    align-items: center;
    gap: 3px;
    margin-top: 2px;
 }
 /* Badge + version-name binding: they wrap as a single unit */
 .badge-version-unit {
    display: inline-flex;
    align-items: center;
    gap: 3px;
    min-width: 0;
    flex-shrink: 0;
 }
 /* Medium density adjustments for version name */
 .medium-density .version-name {
    font-size: 0.8em;
 }
 .medium-density .badge-version-unit .version-name {
    max-width: 90px;
    white-space: nowrap;
    overflow: hidden;
    text-overflow: ellipsis;
 }
 /* Compact density adjustments for version name */
 .compact-density .version-name {
    font-size: 0.75em;
 }
-/* Hide civitai version name when setting is disabled */
+.compact-density .badge-version-unit .version-name {
-body.hide-card-version .civitai-version {
+    max-width: 70px;
    white-space: nowrap;
    overflow: hidden;
    text-overflow: ellipsis;
 }
 .medium-density .version-row {
    gap: 2px;
 }
 /* HIGH / LOW badges — shown inline before version name in card footer */
 .hl-badge {
    display: inline-block;
    font-size: 0.7em;
    font-weight: 600;
    line-height: 1.1;
    padding: 1px 5px;
    border-radius: var(--border-radius-xs);
    border: 1px solid rgba(255, 255, 255, 0.2);
    white-space: nowrap;
 }
 .hl-badge--high {
    color: oklch(75% 0.12 230);
    background: oklch(55% 0.15 240 / 0.25);
    border-color: oklch(60% 0.18 250 / 0.3);
 }
 .hl-badge--low {
    color: oklch(78% 0.10 185);
    background: oklch(50% 0.10 190 / 0.25);
    border-color: oklch(55% 0.12 195 / 0.3);
 }
 .medium-density .hl-badge {
    font-size: 0.65em;
 }
 .compact-density .hl-badge {
    font-size: 0.62em;
    padding: 0px 4px;
 }
 /* Hide version-related elements when setting is disabled */
 body.hide-card-version .civitai-version,
 body.hide-card-version .hl-badge {
    display: none;
 }
 /* Compact density adjustments for version name */
 .compact-density .version-name {
    font-size: 0.75em;
 }
 /* Prevent text selection on cards and interactive elements */
 .model-card,
 .model-card *,
@@ -563,8 +650,13 @@ body.hide-card-version .civitai-version {
    position: absolute;
    box-sizing: border-box;
    transition: transform 160ms ease-out;
-    margin: 0; /* Remove margins, positioning is handled by VirtualScroller */
+    margin: 0;
-    width: 100%; /* Allow width to be set by the VirtualScroller */
+    width: 100%;
 }
 /* Allow cards to grow beyond 260px in virtual scroll mode */
 .virtual-scroll-item.model-card {
    max-width: none;
 }
 .virtual-scroll-item:hover {
@@ -576,11 +668,11 @@ body.hide-card-version .civitai-version {
 .card-grid.virtual-scroll {
    display: block;
    position: relative;
-    margin: 0 auto;
+    margin: 0; /* Remove auto margins - positioning handled by VirtualScroller leftOffset */
    padding: 4px 0; /* Add top/bottom padding equivalent to card padding */
    height: auto;
    width: 100%;
-    max-width: 1400px; /* Keep the max-width from original grid */
+    max-width: none; /* Remove max-width constraint - handled by VirtualScroller */
    box-sizing: border-box; /* Include padding in width calculation */
    overflow-x: hidden; /* Prevent horizontal overflow */
 }
--- a/static/css/components/duplicates.css
+++ b/static/css/components/duplicates.css
@@ -11,8 +11,8 @@
    border-bottom: 1px solid oklch(var(--lora-accent-l) var(--lora-accent-c) var(--lora-accent-h) / 0.4); /* Make bottom border stronger */
    z-index: var(--z-overlay);
    padding: 12px 0;
-    box-shadow: 0 3px 10px rgba(0, 0, 0, 0.2); /* Stronger shadow */
+    box-shadow: var(--shadow-lg); /* Stronger shadow */
-    transition: all 0.3s ease;
+    transition: var(--transition-slow);
    margin-bottom: 20px;
 }
@@ -65,7 +65,7 @@
    align-items: center;
    justify-content: center;
    gap: 6px;
-    transition: all 0.2s ease;
+    transition: var(--transition-base);
 }
 .duplicates-banner button.btn-exit-mode:hover {
@@ -86,16 +86,16 @@
    background: var(--card-bg);
    color: var(--text-color);
    font-size: 0.85em;
-    transition: all 0.2s ease;
+    transition: var(--transition-base);
    cursor: pointer;
-    box-shadow: 0 1px 2px rgba(0, 0, 0, 0.05);
+    box-shadow: var(--shadow-xs);
 }
 .duplicates-banner button:hover {
    border-color: var(--lora-accent-l) var(--lora-accent-c) var(--lora-accent-h);
    background: var(--bg-color);
    transform: translateY(-1px);
-    box-shadow: 0 3px 5px rgba(0, 0, 0, 0.08);
+    box-shadow: var(--shadow-sm);
 }
 .duplicates-banner button.btn-exit {
@@ -122,7 +122,7 @@
    padding: 16px;
    margin-bottom: 24px;
    background: var(--card-bg);
-    box-shadow: 0 2px 6px rgba(0, 0, 0, 0.12); /* Add subtle shadow to groups */
+    box-shadow: var(--shadow-md); /* Add subtle shadow to groups */
    /* Add responsive width settings to match banner */
    max-width: 1400px;
    margin-left: auto;
@@ -173,9 +173,9 @@
    background: var(--card-bg);
    color: var(--text-color);
    font-size: 0.85em;
-    transition: all 0.2s ease;
+    transition: var(--transition-base);
    cursor: pointer;
-    box-shadow: 0 1px 2px rgba(0, 0, 0, 0.05);
+    box-shadow: var(--shadow-xs);
    margin-left: 8px;
 }
@@ -183,7 +183,7 @@
    border-color: var(--lora-accent-l) var(--lora-accent-c) var(--lora-accent-h);
    background: var(--bg-color);
    transform: translateY(-1px);
-    box-shadow: 0 3px 5px rgba(0, 0, 0, 0.08);
+    box-shadow: var(--shadow-sm);
 }
 .card-group-container {
@@ -230,20 +230,20 @@
    justify-content: center;
    cursor: pointer;
    z-index: 1;
-    box-shadow: 0 1px 3px rgba(0, 0, 0, 0.1);
+    box-shadow: var(--shadow-sm);
-    transition: all 0.2s ease;
+    transition: var(--transition-base);
 }
 .group-toggle-btn:hover {
    border-color: var(--lora-accent-l) var(--lora-accent-c) var (--lora-accent-h);
    transform: translateY(-1px);
-    box-shadow: 0 3px 5px rgba(0, 0, 0, 0.08);
+    box-shadow: var(--shadow-sm);
 }
 /* Duplicate card styling */
 .model-card.duplicate {
    position: relative;
-    transition: all 0.2s ease;
+    transition: var(--transition-base);
 }
 .model-card.duplicate:hover {
@@ -257,7 +257,7 @@
 .model-card.duplicate-selected {
    border: 2px solid oklch(var(--lora-accent-l) var(--lora-accent-c) var(--lora-accent-h));
-    box-shadow: 0 0 8px rgba(0, 0, 0, 0.2);
+    box-shadow: var(--shadow-md);
 }
 .model-card .selector-checkbox {
@@ -290,7 +290,7 @@
    background-color: var(--card-bg);
    border: 1px solid var(--border-color);
    border-radius: var(--border-radius-sm);
-    box-shadow: 0 2px 10px rgba(0,0,0,0.2);
+    box-shadow: var(--shadow-lg);
    padding: 10px;
    z-index: 1000;
    max-width: 350px;
@@ -432,7 +432,7 @@
    border-radius: var(--border-radius-xs);
    font-size: 0.85em;
    cursor: pointer;
-    transition: all 0.2s ease;
+    transition: var(--transition-base);
 }
 .btn-verify-hashes:hover {
@@ -461,7 +461,7 @@
    position: absolute;
    top: -8px; /* Moved closer to button */
    right: -8px; /* Moved closer to button */
-    box-shadow: 0 1px 3px rgba(0, 0, 0, 0.15); /* Softer shadow */
+    box-shadow: var(--shadow-sm); /* Softer shadow */
    transition: transform 0.2s ease, opacity 0.2s ease;
 }
@@ -493,7 +493,7 @@
    cursor: help;
    font-size: 16px;
    margin-left: 8px;
-    transition: all 0.2s ease;
+    transition: var(--transition-base);
 }
 .help-icon:hover {
@@ -511,7 +511,7 @@
    border: 1px solid var(--border-color);
    border-radius: var(--border-radius-sm);
    padding: 12px 16px;
-    box-shadow: 0 4px 12px rgba(0, 0, 0, 0.15);
+    box-shadow: var(--shadow-elevated);
    z-index: var(--z-overlay);
    font-size: 0.9em;
    margin-top: 10px;
@@ -572,16 +572,16 @@
 /* In dark mode, add additional distinction */
 html[data-theme="dark"] .duplicates-banner {
-    box-shadow: 0 3px 12px rgba(0, 0, 0, 0.4); /* Stronger shadow in dark mode */
+    box-shadow: var(--shadow-dark-lg); /* Stronger shadow in dark mode */
    background-color: oklch(var(--lora-accent-l) var(--lora-accent-c) var(--lora-accent-h) / 0.15); /* Slightly stronger background in dark mode */
 }
 html[data-theme="dark"] .duplicate-group {
-    box-shadow: 0 2px 8px rgba(0, 0, 0, 0.25); /* Stronger shadow in dark mode */
+    box-shadow: var(--shadow-lg); /* Stronger shadow in dark mode */
 }
 html[data-theme="dark"] .help-tooltip {
-    box-shadow: 0 4px 12px rgba(0, 0, 0, 0.3);
+    box-shadow: var(--shadow-elevated);
 }
 /* Styles for disabled controls during duplicates mode */
--- a/static/css/components/filter-indicator.css
+++ b/static/css/components/filter-indicator.css
@@ -7,22 +7,22 @@
    color: white;
    border-radius: var(--border-radius-xs);
    padding: 4px 10px;
-    transition: all 0.2s ease;
+    transition: var(--transition-base);
    border: 1px solid var(--lora-accent);
    cursor: pointer;
-    box-shadow: 0 1px 3px rgba(0, 0, 0, 0.1);
+    box-shadow: var(--shadow-sm);
    font-size: 0.85em;
 }
 .control-group .filter-active:hover {
    opacity: 0.92;
    transform: translateY(-1px);
-    box-shadow: 0 3px 5px rgba(0, 0, 0, 0.15);
+    box-shadow: var(--shadow-md);
 }
 .control-group .filter-active:active {
    transform: translateY(0);
-    box-shadow: 0 1px 3px rgba(0, 0, 0, 0.1);
+    box-shadow: var(--shadow-sm);
 }
 .control-group .filter-active i.fa-filter {
@@ -59,9 +59,9 @@
 /* Animation for filter indicator */
@keyframes filterPulse {
-    0% { transform: scale(1); box-shadow: 0 1px 3px rgba(0, 0, 0, 0.1); }
+    0% { transform: scale(1); box-shadow: var(--shadow-sm); }
-    50% { transform: scale(1.03); box-shadow: 0 3px 8px rgba(0, 0, 0, 0.15); }
+    50% { transform: scale(1.03); box-shadow: var(--shadow-lg); }
-    100% { transform: scale(1); box-shadow: 0 1px 3px rgba(0, 0, 0, 0.1); }
+    100% { transform: scale(1); box-shadow: var(--shadow-sm); }
 }
 .filter-active.animate {
--- a/static/css/components/header.css
+++ b/static/css/components/header.css
@@ -7,7 +7,7 @@
  height: 48px;
  /* Reduced height */
  width: 100%;
-  box-shadow: 0 2px 4px rgba(0, 0, 0, 0.1);
+  box-shadow: var(--shadow-md);
  /* Slightly stronger shadow */
 }
@@ -22,6 +22,22 @@
  gap: 1rem;
 }
 /* Left section: Logo + Navigation */
 .header-left {
  display: flex;
  align-items: center;
  gap: 1rem;
  flex-shrink: 0;
 }
 /* Right section: Controls */
 .header-right {
  display: flex;
  align-items: center;
  gap: 1rem;
  flex-shrink: 0;
 }
 /* Responsive header container for larger screens */
@media (min-width: 2150px) {
  .header-container {
@@ -77,6 +93,7 @@
  align-items: center;
  gap: 0.5rem;
  font-size: 0.9rem;
  white-space: nowrap;
 }
 .nav-item:hover,
@@ -97,13 +114,101 @@
  color: white;
 }
-/* Header search */
+/* Header search - Centered with VS Code command palette style */
 .header-search {
  flex: 1;
-  max-width: 400px;
+  display: flex;
  justify-content: center;
  max-width: 600px;
  margin: 0 auto;
  transition: opacity 0.2s ease;
 }
 /* VS Code command palette style search container */
 .header-search .search-container {
  width: 100%;
  max-width: 600px;
  position: relative;
  display: flex;
  align-items: center;
  background: var(--input-bg, var(--card-bg));
  border: 1px solid var(--border-color);
  border-radius: var(--border-radius-sm, 6px);
  transition: border-color var(--transition-base), box-shadow var(--transition-base);
  box-shadow: var(--shadow-header);
  overflow: hidden;
 }
 .header-search .search-container:focus-within {
  border-color: var(--lora-accent);
  box-shadow: var(--shadow-header), 0 0 0 1px var(--lora-accent);
 }
 .header-search input {
  flex: 1;
  width: 100%;
  padding: 0.5rem 0.75rem;
  padding-left: 2.25rem !important;
  padding-right: 5rem !important;
  border: none;
  background: transparent;
  color: var(--text-color);
  font-size: 0.95rem;
  outline: none;
 }
 .header-search input::placeholder {
  color: var(--text-muted);
 }
 .header-search .search-icon {
  position: absolute;
  left: 0.75rem;
  color: var(--text-muted);
  font-size: 0.9rem;
  pointer-events: none;
 }
 .header-search .search-options-toggle,
 .header-search .search-filter-toggle {
  position: absolute;
  right: 0.5rem;
  width: 28px;
  height: 28px;
  display: flex;
  align-items: center;
  justify-content: center;
  background: transparent;
  border: none;
  color: var(--text-muted);
  cursor: pointer;
  border-radius: var(--border-radius-xs, 4px);
  transition: background-color var(--transition-base), color var(--transition-base);
 }
 .header-search .search-options-toggle {
  right: 2.25rem;
 }
 .header-search .search-options-toggle:hover,
 .header-search .search-filter-toggle:hover,
 .header-search .search-filter-toggle:focus-visible {
    background: var(--lora-surface-hover, oklch(95% 0.02 256));
    color: var(--lora-accent);
    outline: none;
 }
 .header-search .filter-badge {
  position: absolute;
  top: 2px;
  right: 2px;
  width: 8px;
  height: 8px;
  background: var(--lora-accent);
  border-radius: 50%;
  font-size: 0;
 }
 /* Disabled state for header search */
 .header-search.disabled {
  opacity: 0.5;
@@ -166,7 +271,7 @@
  align-items: center;
  justify-content: center;
  cursor: pointer;
-  transition: all 0.2s ease;
+  transition: background-color var(--transition-base), color var(--transition-base), transform var(--transition-base);
  position: relative;
 }
@@ -238,7 +343,7 @@
  background-color: var(--lora-error);
  border-radius: 50%;
  border: 2px solid var(--card-bg);
-  transition: all 0.2s ease;
+  transition: opacity var(--transition-base);
  pointer-events: none;
  opacity: 0;
 }
@@ -247,44 +352,216 @@
  opacity: 1;
 }
-/* Mobile adjustments */
+/* Hamburger menu button - hidden by default */
-@media (max-width: 768px) {
+.hamburger-menu-btn {
-  .app-title {
+  display: none;
-    display: none;
+  width: 32px;
-    /* Hide text title on mobile */
+  height: 32px;
  border-radius: 50%;
  background: var(--card-bg);
  border: 1px solid var(--border-color);
  color: var(--text-color);
  align-items: center;
  justify-content: center;
  cursor: pointer;
  transition: background-color var(--transition-base), color var(--transition-base);
  flex-shrink: 0;
 }
 .hamburger-menu-btn:hover,
 .hamburger-menu-btn:focus-visible {
    background: var(--lora-surface-hover, oklch(95% 0.02 256));
    color: var(--lora-accent);
    outline: none;
 }
 .hamburger-dropdown .dropdown-item:hover,
 .hamburger-dropdown .dropdown-item:focus-visible {
    background: var(--lora-surface-hover, oklch(95% 0.02 256));
    color: var(--lora-accent);
    outline: none;
 }
 /* Hamburger dropdown menu */
 .hamburger-dropdown {
  display: none;
  position: absolute;
  top: 100%;
  right: 0;
  margin-top: 8px;
  background: var(--card-bg);
  border: 1px solid var(--border-color);
  border-radius: var(--border-radius-sm, 6px);
  box-shadow: var(--shadow-toast);
  padding: 0.5rem;
  min-width: 160px;
  z-index: var(--z-dropdown, 200);
 }
 .hamburger-dropdown.active {
  display: flex;
  flex-direction: column;
  gap: 0.25rem;
 }
 .hamburger-dropdown .dropdown-item {
  display: flex;
  align-items: center;
  gap: 0.75rem;
  padding: 0.5rem 0.75rem;
  border-radius: var(--border-radius-xs, 4px);
  color: var(--text-color);
  cursor: pointer;
  transition: background-color var(--transition-base), color var(--transition-base);
  font-size: 0.9rem;
  white-space: nowrap;
 }
 .hamburger-dropdown .dropdown-item:hover {
  background: var(--lora-surface-hover, oklch(95% 0.02 256));
  color: var(--lora-accent);
 }
 .hamburger-dropdown .dropdown-item i {
  width: 20px;
  text-align: center;
 }
 .hamburger-dropdown .dropdown-divider {
  height: 1px;
  background: var(--border-color);
  margin: 0.25rem 0;
 }
 /* Responsive: Early optimization at 1200px - reduce gaps and padding */
@media (max-width: 1200px) {
  .header-container {
    gap: 0.75rem;
    padding: 0 12px;
  }
  .main-nav {
    gap: 0.25rem;
  }
  .nav-item {
    padding: 0.25rem 0.5rem;
    font-size: 0.85rem;
  }
  .header-controls {
-    gap: 4px;
+    gap: 6px;
  }
-  .header-controls>div {
+  .header-controls > div {
-    width: 28px;
+    width: 30px;
-    height: 28px;
+    height: 30px;
  }
 }
 /* Responsive: Hide nav icons at 1100px to save space */
@media (max-width: 1100px) {
  .nav-item {
    gap: 0;
    padding: 0.25rem 0.4rem;
  }
  .nav-item i {
    display: none;
  }
  .header-search {
    max-width: 450px;
  }
 }
@media (max-width: 950px) {
  .app-title {
    display: none !important;
  }
  .header-container {
    padding: 0 10px;
    gap: 0.5rem;
  }
  .header-controls {
    display: none !important;
  }
  .hamburger-menu-btn {
    display: flex !important;
  }
  .hamburger-dropdown {
    display: none;
  }
  .hamburger-dropdown.active {
    display: flex;
  }
  .header-search {
    max-width: none;
-    margin: 0 0.5rem;
+    margin: 0;
    flex: 1;
    min-width: 200px;
  }
  .main-nav {
-    margin-right: 0.5rem;
+    gap: 0.25rem;
    margin-right: 0;
  }
  .nav-item {
    padding: 0.25rem 0.35rem;
    font-size: 0.8rem;
  }
 }
-/* For very small screens */
+/* Responsive: Compact mode at 768px */
@media (max-width: 768px) {
  .header-search input {
    padding: 0.4rem 0.6rem;
    padding-left: 2rem !important;
    padding-right: 4.5rem !important;
    font-size: 0.9rem;
  }
  .header-search .search-container {
    border-radius: var(--border-radius-xs, 4px);
  }
 }
 /* For very small screens - switch nav to icons only */
@media (max-width: 600px) {
  .header-container {
    padding: 0 8px;
    gap: 0.4rem;
  }
  .main-nav {
-    display: none;
+    display: flex;
-    /* Hide navigation on very small screens */
+    gap: 0.15rem;
    margin-right: 0;
  }
-  .header-search {
+  .nav-item {
-    flex: 1;
+    padding: 0.25rem;
    font-size: 0.75rem;
  }
  .nav-item span {
    display: none;
  }
  .nav-item i {
    display: block;
    font-size: 1rem;
  }
 }
 /* Position relative for hamburger menu positioning */
 .header-right {
  position: relative;
 }
--- a/static/css/components/import-modal.css
+++ b/static/css/components/import-modal.css
@@ -757,7 +757,7 @@
    position: relative;
    border-radius: var(--border-radius-sm);
    overflow: hidden;
-    box-shadow: 0 2px 4px rgba(0, 0, 0, 0.1);
+    box-shadow: var(--shadow-md);
    transition: transform 0.2s ease;
 }
--- a/static/css/components/initialization.css
+++ b/static/css/components/initialization.css
@@ -176,7 +176,7 @@
  background: rgba(var(--lora-accent), 0.05);
  border-radius: var(--border-radius-base);
  padding: var(--space-2);
-  box-shadow: 0 2px 8px rgba(0, 0, 0, 0.05);
+  box-shadow: var(--shadow-sm);
 }
 .tips-header {
--- a/static/css/components/keyboard-nav.css
+++ b/static/css/components/keyboard-nav.css
@@ -11,7 +11,7 @@
    border: 1px solid var(--border-color);
    color: var(--text-color);
    cursor: help;
-    transition: all 0.2s ease;
+    transition: var(--transition-base);
    margin-left: 8px;
 }
@@ -19,7 +19,7 @@
    background: var(--lora-accent);
    color: white;
    transform: translateY(-2px);
-    box-shadow: 0 3px 5px rgba(0, 0, 0, 0.08);
+    box-shadow: var(--shadow-sm);
 }
 .keyboard-nav-hint i {
@@ -46,7 +46,7 @@
    transform: translateY(-15%); /* Vertically center */
    opacity: 0;
    transition: opacity 0.3s;
-    box-shadow: 0 3px 8px rgba(0, 0, 0, 0.15);
+    box-shadow: var(--shadow-lg);
    border: 1px solid var(--lora-border);
    font-size: 0.85em;
    line-height: 1.4;
@@ -92,5 +92,5 @@
    border-radius: 3px;
    padding: 1px 5px;
    font-size: 0.8em;
-    box-shadow: 0 1px 2px rgba(0, 0, 0, 0.08);
+    box-shadow: var(--shadow-xs);
 }
--- a/static/css/components/loading.css
+++ b/static/css/components/loading.css
@@ -18,7 +18,7 @@
    border-radius: var(--border-radius-base);
    text-align: center;
    border: 1px solid var(--lora-border);
-    width: min(400px, 90vw); /* 固定最大宽度，但保持响应式 */
+    width: min(400px, 90vw);
 }
 .loading-spinner {
@@ -33,7 +33,7 @@
 .loading-status {
    margin-bottom: 1rem;
-    color: var(--text-color); /* 使用主题文本颜色 */
+    color: var(--text-color);
    white-space: nowrap;
    overflow: hidden;
    text-overflow: ellipsis;
@@ -42,11 +42,11 @@
 }
 .progress-container {
-    width: 280px; /* 固定进度条宽度 */
+    width: 280px;
-    background-color: var(--lora-border); /* 使用主题边框颜色 */
+    background-color: var(--lora-border);
    border-radius: 4px;
    overflow: hidden;
-    margin: 0 auto; /* 居中显示 */
+    margin: 0 auto;
 }
 .progress-bar {
--- a/static/css/components/lora-modal/description.css
+++ b/static/css/components/lora-modal/description.css
@@ -62,7 +62,7 @@
 }
 .model-description-content code {
-  font-family: monospace;
+  font-family: var(--font-mono);
  font-size: 0.9em;
  background: rgba(0, 0, 0, 0.05);
  padding: 0.1em 0.3em;
--- a/static/css/components/lora-modal/lora-modal.css
+++ b/static/css/components/lora-modal/lora-modal.css
@@ -105,14 +105,14 @@
 .info-item {
    padding: var(--space-2);
-    background: rgba(0, 0, 0, 0.03);
+    background: var(--surface-subtle);
    border: 1px solid rgba(0, 0, 0, 0.1);
    border-radius: var(--border-radius-sm);
 }
-/* 调整深色主题下的样式 */
+/* Dark theme info item styles */
 [data-theme="dark"] .info-item {
-    background: rgba(255, 255, 255, 0.03);
+    background: var(--surface-subtle);
    border: 1px solid var(--lora-border);
 }
@@ -140,18 +140,70 @@
 /* Add specific styles for notes content */
 .info-item.notes .editable-field [contenteditable] {
-    height: 60px; /* Keep initial modal layout stable regardless of note length */
+    min-height: 60px;
-    min-height: 60px; /* Increase height for multiple lines */
+    white-space: pre-wrap;
-    max-height: 420px; /* Limit maximum height */
+    line-height: 1.5;
-    overflow: auto; /* Enable scrolling and resize handle for long content */
+    padding: 8px 12px;
-    resize: vertical; /* Allow manual vertical resizing */
+}
-    white-space: pre-wrap; /* Preserve line breaks */
+
-    line-height: 1.5; /* Improve readability */
+/* Notes expand/collapse — collapsed by default; only applies when JS detects long content */
-    padding: 8px 12px; /* Slightly increase padding */
+.info-item.notes .editable-field {
    position: relative;
    max-height: none;
    overflow: visible;
 }
 .info-item.notes .editable-field.collapsed {
    max-height: 80px;
    overflow: hidden;
 }
 /* Gradient fade overlay hint when collapsed */
 .info-item.notes .editable-field.collapsed::after {
    content: '';
    position: absolute;
    bottom: 0;
    left: 0;
    right: 0;
    height: 28px;
    background: linear-gradient(transparent, var(--bg-color));
    pointer-events: none;
 }
 /* Notes header row — label left, toggle button right */
 .notes-header {
    display: flex;
    align-items: center;
    justify-content: space-between;
 }
 /* Toggle button — icon only, inline with the label */
 .notes-toggle-btn {
    display: none;  /* shown by JS when content exceeds threshold */
    align-items: center;
    justify-content: center;
    width: 26px;
    height: 26px;
    padding: 0;
    border: none;
    background: none;
    color: var(--lora-accent);
    cursor: pointer;
    border-radius: 4px;
    transition: background 0.15s;
    flex-shrink: 0;
 }
 .notes-toggle-btn:hover {
    background: rgba(66, 153, 225, 0.1);
 }
 .notes-toggle-btn i {
    font-size: 0.85em;
 }
 .file-path {
-    font-family: monospace;
+    font-family: var(--font-mono);
    font-size: 0.9em;
 }
@@ -219,13 +271,13 @@
    }
 }
-/* 修改 back-to-top 按钮样式，使其固定在 modal 内部 */
+/* Back-to-top button pinned inside modal */
 .modal-content .back-to-top {
-    position: sticky; /* 改用 sticky 定位 */
+    position: sticky;
-    float: right; /* 使用 float 确保按钮在右侧 */
+    float: right;
-    bottom: 20px; /* 距离底部的距离 */
+    bottom: 20px;
-    margin-right: 20px; /* 右侧间距 */
+    margin-right: 20px;
-    margin-top: -56px; /* 负边距确保不占用额外空间 */
+    margin-top: -56px;
    width: 36px;
    height: 36px;
    border-radius: 50%;
@@ -239,7 +291,7 @@
    opacity: 0;
    visibility: hidden;
    transform: translateY(10px);
-    transition: all 0.3s ease;
+    transition: opacity var(--transition-slow), visibility var(--transition-slow), transform var(--transition-slow);
    z-index: 10;
 }
@@ -255,35 +307,39 @@
    transform: translateY(-2px);
 }
-/* File name copy styles */
+/* Editable inline field styles (file name, version name, etc.) */
-.file-name-wrapper {
+.file-name-wrapper,
 .version-name-wrapper {
    display: flex;
    align-items: center;
    gap: 8px;
-    padding: 4px;
+    padding: 4px 0;
    border-radius: var(--border-radius-xs);
    transition: background-color 0.2s;
    position: relative;
 }
-.file-name-content {
+.file-name-content,
-    padding: 2px 4px;
+.version-name-content {
    padding: 2px 4px 2px 0;
    border-radius: var(--border-radius-xs);
    border: 1px solid transparent;
    flex: 1;
 }
-.file-name-wrapper.editing .file-name-content {
+.file-name-wrapper.editing .file-name-content,
 .version-name-wrapper.editing .version-name-content {
    border: 1px solid var(--lora-accent);
    background: var(--bg-color);
    outline: none;
 }
-/* 合并编辑按钮样式 */
+/* Consolidated edit button styles */
 .edit-model-name-btn,
 .edit-file-name-btn,
 .edit-base-model-btn,
-.edit-model-description-btn {
+.edit-model-description-btn,
 .edit-version-name-btn {
    background: transparent;
    border: none;
    color: var(--text-color);
@@ -291,7 +347,7 @@
    cursor: pointer;
    padding: 2px 5px;
    border-radius: var(--border-radius-xs);
-    transition: all 0.2s ease;
+    transition: opacity var(--transition-base), background-color var(--transition-base);
    margin-left: var(--space-1);
 }
@@ -299,9 +355,11 @@
 .edit-file-name-btn.visible,
 .edit-base-model-btn.visible,
 .edit-model-description-btn.visible,
 .edit-version-name-btn.visible,
 .model-name-header:hover .edit-model-name-btn,
 .file-name-wrapper:hover .edit-file-name-btn,
 .base-model-display:hover .edit-base-model-btn,
 .version-name-wrapper:hover .edit-version-name-btn,
 .model-name-header:hover .edit-model-description-btn {
    opacity: 0.5;
 }
@@ -309,14 +367,16 @@
 .edit-model-name-btn:hover,
 .edit-file-name-btn:hover,
 .edit-base-model-btn:hover,
-.edit-model-description-btn:hover {
+.edit-model-description-btn:hover,
-    opacity: 0.8 !important;
+.edit-version-name-btn:hover {
    opacity: 0.8;
    background: rgba(0, 0, 0, 0.05);
 }
 [data-theme="dark"] .edit-model-name-btn:hover,
 [data-theme="dark"] .edit-file-name-btn:hover,
-[data-theme="dark"] .edit-base-model-btn:hover {
+[data-theme="dark"] .edit-base-model-btn:hover,
 [data-theme="dark"] .edit-version-name-btn:hover {
    background: rgba(255, 255, 255, 0.05);
 }
@@ -327,7 +387,7 @@
 }
 .base-wrapper {
-    flex: 2;  /* 分配更多空间给base model */
+    flex: 2;  /* Allocate more space to base model */
 }
 /* Base model display and editing styles */
@@ -338,7 +398,7 @@
 }
 .base-model-content {
-    padding: 2px 4px;
+    padding: 2px 4px 2px 0;
    border-radius: var(--border-radius-xs);
    border: 1px solid transparent;
    color: var(--text-color);
@@ -370,7 +430,7 @@
 }
 .size-wrapper span {
-    font-family: monospace;
+    font-family: var(--font-mono);
    font-size: 0.9em;
    opacity: 0.9;
 }
@@ -387,7 +447,7 @@
    margin: 0;
    padding: var(--space-1);
    border-radius: var(--border-radius-xs);
-    font-size: 1.5em !important;
+    font-size: 1.5em;
    font-weight: 600;
    line-height: 1.2;
    color: var(--text-color);
@@ -423,7 +483,7 @@
    color: var(--text-color);
    cursor: pointer;
    font-size: 0.95em;
-    transition: all 0.2s;
+    transition: var(--transition-base);
    opacity: 0.7;
    position: relative;
 }
@@ -828,18 +888,18 @@
    align-items: center;
    gap: 10px;
    padding: 2px 10px;
-    background: rgba(0, 0, 0, 0.03);
+    background: var(--surface-subtle);
    border: 1px solid rgba(0, 0, 0, 0.1);
    border-radius: var(--border-radius-sm);
    max-width: fit-content;
    cursor: pointer;
-    transition: all 0.2s;
+    transition: var(--transition-base);
 }
 [data-theme="dark"] .creator-info,
 [data-theme="dark"] .civitai-view,
 [data-theme="dark"] .modal-send-btn {
-    background: rgba(255, 255, 255, 0.03);
+    background: var(--surface-subtle);
    border: 1px solid var(--lora-border);
 }
@@ -898,14 +958,14 @@
    align-items: center;
    gap: 6px;
    padding: 6px 12px;
-    background: rgba(0, 0, 0, 0.03);
+    background: var(--surface-subtle);
    border: 1px solid rgba(0, 0, 0, 0.1);
    border-radius: var(--border-radius-sm);
    color: var(--text-color);
    cursor: pointer;
    font-weight: 500;
    font-size: 0.9em;
-    transition: all 0.2s;
+    transition: var(--transition-base);
 }
 .civitai-view i {
@@ -921,18 +981,18 @@
    align-items: center;
    gap: 6px;
    padding: 6px 12px;
-    background: rgba(0, 0, 0, 0.03);
+    background: var(--surface-subtle);
    border: 1px solid rgba(0, 0, 0, 0.1);
    border-radius: var(--border-radius-sm);
    color: var(--text-color);
    cursor: pointer;
    font-weight: 500;
    font-size: 0.9em;
-    transition: all 0.2s;
+    transition: var(--transition-base);
 }
 [data-theme="dark"] .modal-send-btn {
-    background: rgba(255, 255, 255, 0.03);
+    background: var(--surface-subtle);
    border: 1px solid var(--lora-border);
 }
--- a/static/css/components/lora-modal/preset-tags.css
+++ b/static/css/components/lora-modal/preset-tags.css
@@ -28,7 +28,7 @@
  border-radius: var(--border-radius-xs);
  padding: calc(var(--space-1) * 0.5) var(--space-1);
  gap: var(--space-1);
-  transition: all 0.2s ease;
+  transition: var(--transition-base);
 }
 .preset-tag span {
@@ -40,7 +40,7 @@
  color: var(--text-color);
  opacity: 0.5;
  cursor: pointer;
-  transition: all 0.2s ease;
+  transition: var(--transition-base);
 }
 .preset-tag:hover {
--- a/static/css/components/lora-modal/showcase.css
+++ b/static/css/components/lora-modal/showcase.css
@@ -111,8 +111,8 @@
  align-items: center;
  justify-content: center;
  cursor: pointer;
-  transition: all 0.2s ease;
+  transition: var(--transition-base);
-  box-shadow: 0 2px 5px rgba(0, 0, 0, 0.15);
+  box-shadow: var(--shadow-md);
  padding: 0;
  position: relative;
  overflow: hidden;
@@ -120,7 +120,7 @@
 .media-control-btn:hover {
  transform: translateY(-2px);
-  box-shadow: 0 3px 7px rgba(0, 0, 0, 0.2);
+  box-shadow: var(--shadow-lg);
 }
 .media-control-btn.set-preview-btn:hover {
@@ -141,8 +141,9 @@
  border-color: var(--lora-error);
 }
-/* Disabled state for delete button */
+/* Disabled state for delete and create-recipe buttons */
-.media-control-btn.example-delete-btn.disabled {
+.media-control-btn.example-delete-btn.disabled,
 .media-control-btn.create-recipe-btn.disabled {
  opacity: 0.5;
  cursor: not-allowed;
 }
@@ -204,7 +205,7 @@
  z-index: 5;
  max-height: 50%; /* Reduced to take less space */
  overflow-y: auto;
-  box-shadow: 0 -2px 8px rgba(0, 0, 0, 0.1);
+  box-shadow: var(--shadow-inset-top);
  opacity: 0;
  pointer-events: none;
 }
@@ -219,7 +220,7 @@
 /* Adjust to dark theme */
 [data-theme="dark"] .image-metadata-panel {
  background: var(--card-bg);
-  box-shadow: 0 -2px 8px rgba(0, 0, 0, 0.3);
+  box-shadow: var(--shadow-inset-top);
 }
 .metadata-content {
@@ -296,7 +297,7 @@
 .metadata-prompt {
  color: var(--text-color);
-  font-family: monospace;
+  font-family: var(--font-mono);
  font-size: 0.85em;
  white-space: pre-wrap;
 }
@@ -311,7 +312,7 @@
  opacity: 0.6;
  cursor: pointer;
  padding: 3px;
-  transition: all 0.2s ease;
+  transition: var(--transition-base);
 }
 .copy-prompt-btn:hover {
@@ -408,7 +409,7 @@
  border-radius: var(--border-radius-sm);
  padding: var(--space-4);
  text-align: center;
-  transition: all 0.3s ease;
+  transition: var(--transition-slow);
  background: var(--lora-surface);
  cursor: pointer;
 }
@@ -454,9 +455,9 @@
 }
 .import-formats {
-  font-size: 0.8em !important;
+  font-size: 0.8em;
-  opacity: 0.6 !important;
+  opacity: 0.6;
-  margin-top: var(--space-2) !important;
+  margin-top: var(--space-2);
 }
 .select-files-btn {
@@ -470,7 +471,7 @@
  display: flex;
  align-items: center;
  gap: 8px;
-  transition: all 0.2s;
+  transition: var(--transition-base);
 }
 .select-files-btn:hover {
@@ -480,7 +481,7 @@
 /* For dark theme */
 [data-theme="dark"] .import-container {
-  background: rgba(255, 255, 255, 0.03);
+  background: var(--surface-subtle);
 }
 /* Setup Guidance State - When example images path is not configured */
--- a/static/css/components/lora-modal/tag.css
+++ b/static/css/components/lora-modal/tag.css
@@ -21,7 +21,7 @@
 .model-tag-compact {
  /* Updated styles to match info-item appearance */
-  background: rgba(0, 0, 0, 0.03);
+  background: var(--surface-subtle);
  border: 1px solid rgba(0, 0, 0, 0.1);
  border-radius: var(--border-radius-xs);
  padding: 2px 8px;
@@ -45,7 +45,7 @@
 /* Adjust dark theme tag styles */
 [data-theme="dark"] .model-tag-compact {
-  background: rgba(255, 255, 255, 0.03);
+  background: var(--surface-subtle);
  border: 1px solid var(--lora-border);
 }
@@ -73,14 +73,14 @@
  background: var(--card-bg);
  border: 1px solid var(--border-color);
  border-radius: var(--border-radius-sm);
-  box-shadow: 0 3px 8px rgba(0, 0, 0, 0.15);
+  box-shadow: var(--shadow-lg);
  padding: 10px 14px;
  max-width: 400px;
  z-index: 10;
  opacity: 0;
  visibility: hidden;
  transform: translateY(-4px);
-  transition: all 0.2s ease;
+  transition: var(--transition-base);
  pointer-events: none;
 }
@@ -101,7 +101,7 @@
 .tooltip-tag {
  /* Updated styles to match info-item appearance */
-  background: rgba(0, 0, 0, 0.03);
+  background: var(--surface-hover);
  border: 1px solid rgba(0, 0, 0, 0.1);
  border-radius: var(--border-radius-xs);
  padding: 3px 8px;
@@ -111,7 +111,7 @@
 /* Adjust dark theme tooltip tag styles */
 [data-theme="dark"] .tooltip-tag {
-  background: rgba(255, 255, 255, 0.03);
+  background: var(--surface-hover);
  border: 1px solid var(--lora-border);
 }
@@ -130,7 +130,7 @@
  cursor: pointer;
  padding: 2px 5px;
  border-radius: var(--border-radius-xs);
-  transition: all 0.2s ease;
+  transition: var(--transition-base);
  margin-left: var(--space-1);
 }
--- a/static/css/components/lora-modal/triggerwords.css
+++ b/static/css/components/lora-modal/triggerwords.css
@@ -1,14 +1,14 @@
 /* Update Trigger Words styles */
 .info-item.trigger-words {
  padding: var(--space-2);
-  background: rgba(0, 0, 0, 0.03);
+  background: var(--surface-subtle);
  border: 1px solid rgba(0, 0, 0, 0.1);
  border-radius: var(--border-radius-sm);
 }
-/* 调整 trigger words 样式 */
+/* Trigger words styles */
 [data-theme="dark"] .info-item.trigger-words {
-  background: rgba(255, 255, 255, 0.03);
+  background: var(--surface-subtle);
  border: 1px solid var(--lora-border);
 }
@@ -48,7 +48,7 @@
  border-radius: var(--border-radius-xs);
  padding: 4px 8px;
  cursor: pointer;
-  transition: all 0.2s ease;
+  transition: var(--transition-base);
  gap: 6px;
  position: relative;
 }
--- a/static/css/components/lora-modal/versions.css
+++ b/static/css/components/lora-modal/versions.css
@@ -146,7 +146,7 @@
    background: color-mix(in oklch, var(--card-bg) 92%, var(--bg-color) 8%);
    border: 1px solid var(--lora-border);
    border-radius: var(--border-radius-sm);
-    box-shadow: 0 1px 2px rgba(0, 0, 0, 0.04);
+    box-shadow: var(--shadow-xs);
    transition: border-color 0.2s ease, box-shadow 0.2s ease, transform 0.2s ease;
 }
@@ -156,7 +156,7 @@
 .model-version-row:hover {
    transform: translateY(-1px);
-    box-shadow: 0 4px 16px rgba(0, 0, 0, 0.08);
+    box-shadow: var(--shadow-xl);
 }
 .model-version-row.is-clickable {
@@ -186,7 +186,7 @@
    height: 88px;
    border-radius: var(--border-radius-xs);
    overflow: hidden;
-    background: rgba(0, 0, 0, 0.03);
+    background: var(--surface-hover);
    display: flex;
    align-items: center;
    justify-content: center;
@@ -374,11 +374,23 @@
    background: color-mix(in oklch, var(--lora-surface) 35%, transparent);
 }
 .version-action-disabled {
    background: transparent;
    border-color: var(--border-color);
    color: var(--text-muted);
    opacity: 0.6;
    cursor: not-allowed;
 }
 .version-action:disabled {
    opacity: 0.6;
    cursor: not-allowed;
 }
 .version-action-disabled-wrapper {
    display: inline-flex;
 }
 .versions-loading-state,
 .versions-empty,
 .versions-error {
--- a/static/css/components/media-viewer.css
+++ b/static/css/components/media-viewer.css
@@ -0,0 +1,124 @@
 .media-viewer-overlay {
  position: fixed;
  top: 0;
  left: 0;
  right: 0;
  bottom: 0;
  background: rgba(0, 0, 0, 0);
  z-index: 10000;
  display: flex;
  align-items: center;
  justify-content: center;
  cursor: pointer;
  transition: background 0.3s ease;
 }
 .media-viewer-overlay.active {
  background: rgba(0, 0, 0, 0.92);
 }
 .media-viewer-close {
  position: fixed;
  top: 16px;
  right: 16px;
  width: 40px;
  height: 40px;
  border-radius: 50%;
  background: rgba(255, 255, 255, 0.1);
  border: none;
  color: #fff;
  font-size: 18px;
  display: flex;
  align-items: center;
  justify-content: center;
  cursor: pointer;
  z-index: 10001;
  transition: background 0.2s ease;
  opacity: 0;
 }
 .media-viewer-overlay.active .media-viewer-close {
  opacity: 1;
 }
 .media-viewer-close:hover {
  background: rgba(255, 255, 255, 0.25);
 }
 .media-viewer-content-container {
  display: flex;
  flex-direction: column;
  align-items: center;
  justify-content: center;
  max-width: 90vw;
  max-height: 95vh;
  cursor: default;
 }
 .media-viewer-media {
  display: block;
  max-width: 90vw;
  max-height: 85vh;
  object-fit: contain;
  border-radius: 4px;
  box-shadow: var(--shadow-dark-lg);
 }
 .media-viewer-video {
  max-height: 80vh;
 }
 .media-viewer-counter {
  margin-top: 8px;
  color: rgba(255, 255, 255, 0.5);
  font-size: 0.85em;
  text-align: center;
  min-height: 1.2em;
 }
 .media-viewer-title {
  margin-top: 4px;
  color: rgba(255, 255, 255, 0.7);
  font-size: 0.9em;
  text-align: center;
  max-width: 90vw;
  overflow: hidden;
  text-overflow: ellipsis;
  white-space: nowrap;
 }
 .media-viewer-nav {
  position: fixed;
  top: 50%;
  transform: translateY(-50%);
  width: 48px;
  height: 80px;
  border-radius: 4px;
  background: rgba(255, 255, 255, 0.06);
  border: none;
  color: #fff;
  font-size: 24px;
  display: flex;
  align-items: center;
  justify-content: center;
  cursor: pointer;
  z-index: 10001;
  opacity: 0;
  transition: opacity 0.2s ease, background 0.2s ease;
 }
 .media-viewer-overlay.active .media-viewer-nav {
  opacity: 1;
 }
 .media-viewer-nav:hover {
  background: rgba(255, 255, 255, 0.18);
 }
 .media-viewer-prev {
  left: 16px;
 }
 .media-viewer-next {
  right: 16px;
 }
--- a/static/css/components/menu.css
+++ b/static/css/components/menu.css
@@ -5,7 +5,7 @@
    border-radius: var(--border-radius-xs);
    padding: 4px 0;
    min-width: 180px;
-    box-shadow: 0 2px 10px rgba(0, 0, 0, 0.2);
+    box-shadow: var(--shadow-dropdown);
    z-index: 1000;
    display: none;
    backdrop-filter: blur(10px);
@@ -21,9 +21,11 @@
    background: var(--lora-surface);
 }
-.context-menu-item:hover {
+.context-menu-item:hover,
 .context-menu-item:focus-visible {
    background-color: var(--lora-accent);
    color: var(--lora-text);
    outline: none;
 }
 .context-menu-separator {
@@ -41,6 +43,63 @@
    text-align: center;
 }
 /* Section Headers */
 .context-menu-section-header {
    padding: 6px 12px 2px;
    font-size: 11px;
    font-weight: 600;
    text-transform: uppercase;
    letter-spacing: 0.5px;
    color: var(--text-muted);
    cursor: default;
    user-select: none;
 }
 /* Submenu */
 .context-menu-item.has-submenu {
    position: relative;
    justify-content: space-between;
 }
 .submenu-arrow {
    margin-left: auto;
    font-size: 10px;
    width: auto !important;
 }
 .context-submenu {
    position: absolute;
    left: calc(100% - 4px);
    top: -1px;
    display: none;
    background: var(--lora-surface);
    border: 1px solid var(--border-color);
    border-radius: var(--border-radius-xs);
    padding: 0;
    min-width: 200px;
    box-shadow: var(--shadow-dropdown);
    z-index: 1001;
    backdrop-filter: blur(10px);
 }
 .context-submenu .context-menu-item {
    white-space: nowrap;
    margin: 0;
 }
 .context-submenu .context-menu-item:first-child {
    padding-top: 9px;
 }
 .context-submenu .context-menu-item:last-child {
    padding-bottom: 9px;
 }
 .context-submenu.flip-left {
    left: auto;
    right: 100%;
 }
 /* NSFW Level Selector */
 .nsfw-level-selector {
    position: fixed;
@@ -51,7 +110,7 @@
    border: 1px solid var(--border-color);
    border-radius: var(--border-radius-base);
    padding: 16px;
-    box-shadow: 0 4px 20px rgba(0, 0, 0, 0.2);
+    box-shadow: var(--shadow-modal);
    z-index: var(--z-modal);
    width: 300px;
    display: none;
@@ -105,7 +164,7 @@
    border: 1px solid var(--border-color);
    color: var(--text-color);
    cursor: pointer;
-    transition: all 0.2s ease;
+    transition: var(--transition-base);
 }
 .nsfw-level-btn:hover {
@@ -129,7 +188,7 @@
    max-width: 350px;
    max-height: 400px;
    overflow-y: auto;
-    box-shadow: 0 2px 10px rgba(0, 0, 0, 0.2);
+    box-shadow: var(--shadow-dropdown);
    z-index: var(--z-overlay);
    display: none;
    backdrop-filter: blur(10px);
--- a/static/css/components/modal/_base.css
+++ b/static/css/components/modal/_base.css
@@ -1,4 +1,4 @@
-/* modal 基础样式 */
+/* Modal base styles */
 .modal {
    display: none;
    position: fixed;
@@ -6,19 +6,19 @@
    left: 0;
    width: 100%;
    height: calc(100% - var(--header-height, 48px)); /* Adjust height to exclude header */
-    background: rgba(0, 0, 0, 0.2); /* 调整为更淡的半透明黑色 */
+    background: rgba(0, 0, 0, 0.2);
    z-index: var(--z-modal);
    overflow: auto; /* Change from hidden to auto to allow scrolling */
 }
-/* 当模态窗口打开时，禁止body滚动 */
+/* Prevent body scroll when modal is open */
 body.modal-open {
    position: fixed;
    width: 100%;
-    padding-right: var(--scrollbar-width, 0px); /* 补偿滚动条消失导致的页面偏移 */
+    padding-right: var(--scrollbar-width, 0px);
 }
-/* modal-content 样式 */
+/* Modal content styles */
 .modal-content {
    position: relative;
    max-width: 800px;
@@ -29,12 +29,9 @@ body.modal-open {
    border-radius: var(--border-radius-base);
    padding: var(--space-3);
    border: 1px solid var(--lora-border);
-    box-shadow: 
+    box-shadow: var(--shadow-md);
        0 4px 6px -1px rgba(0, 0, 0, 0.1),
        0 2px 4px -1px rgba(0, 0, 0, 0.06),
        0 10px 15px -3px rgba(0, 0, 0, 0.05);
    overflow-y: auto;
-    overflow-x: hidden; /* 防止水平滚动条 */
+    overflow-x: hidden;
    scrollbar-gutter: stable both-edges; /* Reserve space to prevent layout shift when scrollbar toggles */
 }
@@ -42,10 +39,10 @@ body.modal-open {
    min-height: 480px;
 }
-/* 当 modal 打开时锁定 body */
+/* Lock body when modal is open */
 body.modal-open {
-    overflow: hidden !important; /* 覆盖 base.css 中的 scroll */
+    overflow: hidden !important;
-    padding-right: var(--scrollbar-width, 8px); /* 使用滚动条宽度作为补偿 */
+    padding-right: var(--scrollbar-width, 8px);
 }
@keyframes modalFadeIn {
@@ -67,12 +64,25 @@ body.modal-open {
 }
 .cancel-btn, .delete-btn, .exclude-btn, .confirm-btn {
-    padding: 8px var(--space-2);
+    display: flex;
-    border-radius: 6px;
+    align-items: center;
    justify-content: center;
    gap: var(--space-1);
    padding: var(--space-1) var(--space-2);
    border-radius: var(--border-radius-sm);
    border: none;
    cursor: pointer;
    font-weight: 500;
    font-size: 0.95em;
    min-width: 100px;
    transition: background-color var(--transition-base), opacity var(--transition-base), transform var(--transition-fast);
 }
 .cancel-btn:active,
 .delete-btn:active,
 .exclude-btn:active,
 .confirm-btn:active {
    transform: scale(0.98);
 }
 .cancel-btn {
@@ -92,16 +102,20 @@ body.modal-open {
    color: white;
 }
-.cancel-btn:hover {
+.cancel-btn:hover,
 .cancel-btn:focus-visible {
    background: var(--lora-border);
 }
-.delete-btn:hover {
+.delete-btn:hover,
-    opacity: 0.9;
+.delete-btn:focus-visible {
    background: oklch(from var(--lora-error) l c h / 85%);
 }
-.exclude-btn:hover, .confirm-btn:hover {
+.exclude-btn:hover,
-    opacity: 0.9;
+.exclude-btn:focus-visible,
 .confirm-btn:hover,
 .confirm-btn:focus-visible {
    background: oklch(from var(--lora-accent, #4f46e5) l c h / 85%);
 }
@@ -121,47 +135,41 @@ body.modal-open {
    font-size: 1.5em;
    cursor: pointer;
    opacity: 0.7;
-    transition: opacity 0.2s;
+    transition: opacity var(--transition-base);
    z-index: 10;
 }
-.close:hover {
+.close:hover,
 .close:focus-visible {
    opacity: 1;
    outline: 2px solid var(--lora-accent);
    outline-offset: 2px;
    border-radius: var(--border-radius-xs);
 }
-/* 统一各个 section 的样式 */
+/* Unified section styles */
 .support-section,
 .changelog-section,
 .update-info,
 .info-item,
 .path-preview {
-    background: rgba(0, 0, 0, 0.03);
+    background: var(--surface-subtle);
-    border: 1px solid rgba(0, 0, 0, 0.1);
+    border: 1px solid var(--lora-border);
    border-radius: var(--border-radius-sm);
    padding: var(--space-2);
 }
-/* 深色主题统一样式 */
+/* Dark theme unified styles */
 [data-theme="dark"] .modal-content {
    background: var(--lora-surface);
    border: 1px solid var(--lora-border);
 }
 [data-theme="dark"] .support-section,
 [data-theme="dark"] .changelog-section,
 [data-theme="dark"] .update-info,
 [data-theme="dark"] .info-item,
 [data-theme="dark"] .path-preview,
 [data-theme="dark"] #bulkDownloadMissingLorasModal .bulk-download-loras-preview {
    background: rgba(255, 255, 255, 0.03);
    border: 1px solid var(--lora-border);
 }
 .primary-btn {
    display: flex;
    align-items: center;
    gap: 8px;
-    padding: 8px 16px;
+    padding: var(--space-1) var(--space-2);
    background-color: var(--lora-accent);
    color: var(--lora-text);
    border: none;
@@ -171,9 +179,11 @@ body.modal-open {
    font-size: 0.95em;
 }
-.primary-btn:hover {
+.primary-btn:hover,
 .primary-btn:focus-visible {
    background-color: oklch(from var(--lora-accent) l c h / 85%);
    color: var(--lora-text);
    outline: none;
 }
 /* Secondary button styles */
@@ -181,19 +191,21 @@ body.modal-open {
    display: flex;
    align-items: center;
    gap: 8px;
-    padding: 8px 16px;
+    padding: var(--space-1) var(--space-2);
    background-color: var(--card-bg);
-    color: var (--text-color);
+    color: var(--text-color);
    border: 1px solid var(--border-color);
    border-radius: var(--border-radius-sm);
    cursor: pointer;
-    transition: all 0.2s;
+    transition: var(--transition-base);
    font-size: 0.95em;
 }
-.secondary-btn:hover {
+.secondary-btn:hover,
 .secondary-btn:focus-visible {
    background-color: var(--border-color);
    color: var(--text-color);
    outline: none;
 }
 /* Disabled button styles */
@@ -244,7 +256,7 @@ button:disabled,
    display: flex;
    align-items: center;
    gap: 8px;
-    padding: 8px 16px;
+    padding: var(--space-1) var(--space-2);
    background-color: var(--lora-error);
    color: white;
    border: none;
@@ -254,25 +266,22 @@ button:disabled,
    font-size: 0.95em;
 }
-.danger-btn:hover {
+.danger-btn:hover,
 .danger-btn:focus-visible {
    background-color: oklch(from var(--lora-error) l c h / 85%);
    color: white;
    outline: none;
 }
 /* Metadata archive status styles */
 .metadata-archive-status {
-    background: rgba(0, 0, 0, 0.03);
+    background: var(--surface-subtle);
-    border: 1px solid rgba(0, 0, 0, 0.1);
+    border: 1px solid var(--lora-border);
    border-radius: var(--border-radius-sm);
    padding: var(--space-2);
    margin-bottom: var(--space-2);
 }
 [data-theme="dark"] .metadata-archive-status {
    background: rgba(255, 255, 255, 0.03);
    border: 1px solid var(--lora-border);
 }
 .archive-status-item {
    display: flex;
    justify-content: space-between;
@@ -312,17 +321,12 @@ button:disabled,
 }
 .backup-status {
-    background: rgba(0, 0, 0, 0.03);
+    background: var(--surface-subtle);
-    border: 1px solid rgba(0, 0, 0, 0.1);
+    border: 1px solid var(--lora-border);
    border-radius: var(--border-radius-sm);
    padding: var(--space-3);
 }
 [data-theme="dark"] .backup-status {
    background: rgba(255, 255, 255, 0.03);
    border: 1px solid var(--lora-border);
 }
 .backup-summary-grid {
    display: grid;
    grid-template-columns: repeat(auto-fit, minmax(160px, 1fr));
@@ -331,17 +335,12 @@ button:disabled,
 }
 .backup-summary-card {
-    background: rgba(255, 255, 255, 0.5);
+    background: var(--lora-surface);
-    border: 1px solid rgba(0, 0, 0, 0.06);
+    border: 1px solid var(--lora-border);
    border-radius: var(--border-radius-sm);
    padding: var(--space-2);
 }
 [data-theme="dark"] .backup-summary-card {
    background: rgba(255, 255, 255, 0.02);
    border-color: rgba(255, 255, 255, 0.05);
 }
 .backup-summary-label {
    color: var(--text-color);
    font-size: 0.85rem;
@@ -404,14 +403,9 @@ button:disabled,
 }
 .backup-location-details {
-    border: 1px solid rgba(0, 0, 0, 0.1);
+    border: 1px solid var(--lora-border);
    border-radius: var(--border-radius-sm);
-    background: rgba(0, 0, 0, 0.02);
+    background: var(--surface-subtle);
 }
 [data-theme="dark"] .backup-location-details {
    border-color: var(--lora-border);
    background: rgba(255, 255, 255, 0.02);
 }
 .backup-location-details summary {
@@ -442,16 +436,12 @@ button:disabled,
    max-width: 100%;
    padding: 6px 8px;
    border-radius: var(--border-radius-sm);
-    background: rgba(0, 0, 0, 0.05);
+    background: var(--surface-subtle);
    color: var(--text-color);
    overflow-wrap: anywhere;
    word-break: break-word;
 }
 [data-theme="dark"] .backup-location-path {
    background: rgba(255, 255, 255, 0.05);
 }
@media (max-width: 768px) {
    .backup-status-row {
        grid-template-columns: 1fr;
@@ -519,8 +509,8 @@ button:disabled,
 }
 #bulkDownloadMissingLorasModal .bulk-download-loras-preview {
-    background: rgba(0, 0, 0, 0.03);
+    background: var(--surface-subtle);
-    border: 1px solid rgba(0, 0, 0, 0.1);
+    border: 1px solid var(--lora-border);
    border-radius: var(--border-radius-sm);
    padding: var(--space-3);
    margin-bottom: var(--space-3);
@@ -578,7 +568,7 @@ button:disabled,
    align-items: flex-start;
    gap: var(--space-2);
    padding: var(--space-2);
-    background: rgba(59, 130, 246, 0.1);
+    background: oklch(from var(--lora-accent) l c h / 0.1);
    border-radius: var(--border-radius-sm);
    font-size: 0.9em;
    color: var(--text-color);
--- a/static/css/components/modal/delete-modal.css
+++ b/static/css/components/modal/delete-modal.css
@@ -33,6 +33,39 @@
    animation: modalFadeIn 0.2s ease-out;
 }
 #resolveFilenameConflictsModal .confirmation-message {
    color: var(--text-color);
    margin: var(--space-2) 0;
    font-size: 1em;
    line-height: 1.5;
 }
 #resolveFilenameConflictsModal .resolve-conflicts-detail {
    color: var(--text-color);
    margin: var(--space-2) 0;
    font-size: 0.95em;
    line-height: 1.5;
 }
 #resolveFilenameConflictsModal .resolve-conflicts-detail code {
    background: var(--lora-surface);
    padding: 2px 6px;
    border-radius: 3px;
    font-family: var(--font-mono);
    border: 1px solid var(--lora-border);
 }
 #resolveFilenameConflictsModal .resolve-conflicts-impact {
    background: var(--lora-surface);
    border: 1px solid var(--lora-border);
    border-radius: var(--border-radius-sm);
    padding: var(--space-2);
    margin: var(--space-2) 0;
    color: var(--text-color);
    text-align: left;
    line-height: 1.5;
 }
 .delete-model-info,
 .exclude-model-info {
    /* Update info display styling */
--- a/static/css/components/modal/doctor-modal.css
+++ b/static/css/components/modal/doctor-modal.css
@@ -48,8 +48,7 @@
    padding: var(--space-3);
    border-radius: var(--border-radius-sm);
    border: 1px solid var(--lora-border);
-    background: rgba(0, 0, 0, 0.03);
+    background: var(--surface-subtle);
    box-shadow: inset 0 1px 0 rgba(255, 255, 255, 0.03);
 }
 .doctor-kicker {
@@ -128,7 +127,7 @@
 .doctor-issue-card {
    border: 1px solid rgba(0, 0, 0, 0.1);
-    background: rgba(0, 0, 0, 0.03);
+    background: var(--surface-subtle);
    border-radius: var(--border-radius-sm);
    padding: var(--space-3);
    box-shadow: none;
@@ -242,7 +241,7 @@
 [data-theme="dark"] .doctor-hero,
 [data-theme="dark"] .doctor-issue-card {
-    background: rgba(255, 255, 255, 0.03);
+    background: var(--surface-subtle);
    border-color: var(--lora-border);
    box-shadow: none;
 }
--- a/static/css/components/modal/download-modal.css
+++ b/static/css/components/modal/download-modal.css
@@ -37,7 +37,7 @@
    border: 1px solid var(--border-color);
    border-radius: var(--border-radius-sm);
    cursor: pointer;
-    transition: all 0.2s ease;
+    transition: var(--transition-base);
    background: var(--bg-color);
    margin: 1px;
    position: relative;
@@ -45,7 +45,7 @@
 .version-item:hover {
    border-color: var(--lora-accent);
-    box-shadow: 0 2px 8px rgba(0, 0, 0, 0.1);
+    box-shadow: var(--shadow-md);
    z-index: 1;
 }
@@ -156,7 +156,7 @@
    background: var(--bg-color);
    color: var(--text-color);
    cursor: pointer;
-    transition: all 0.2s ease;
+    transition: var(--transition-base);
    display: flex;
    align-items: center;
    justify-content: center;
@@ -225,7 +225,7 @@
    padding: 4px 8px;
    border-radius: var(--border-radius-xs);
    cursor: pointer;
-    transition: all 0.2s ease;
+    transition: var(--transition-base);
    color: var(--text-color);
    opacity: 0.7;
    text-decoration: none;
@@ -272,7 +272,7 @@
    padding: 4px 8px;
    cursor: pointer;
    border-radius: var(--border-radius-xs);
-    transition: all 0.2s ease;
+    transition: var(--transition-base);
    position: relative;
 }
@@ -293,7 +293,7 @@
    justify-content: center;
    cursor: pointer;
    border-radius: 2px;
-    transition: all 0.2s ease;
+    transition: var(--transition-base);
 }
 .tree-expand-icon:hover {
@@ -364,7 +364,7 @@
    color: var(--text-color);
    cursor: pointer;
    font-size: 0.8em;
-    transition: all 0.2s ease;
+    transition: var(--transition-base);
 }
 .create-folder-form button.confirm {
@@ -404,7 +404,7 @@
 .path-display {
    padding: var(--space-1);
    color: var(--text-color);
-    font-family: monospace;
+    font-family: var(--font-mono);
    font-size: 0.9em;
    line-height: 1.4;
    white-space: pre-wrap;
@@ -453,7 +453,7 @@
    right: 0;
    bottom: 0;
    background-color: var(--border-color);
-    transition: all 0.3s ease;
+    transition: var(--transition-slow);
    border-radius: 18px;
 }
@@ -465,9 +465,9 @@
    left: 3px;
    bottom: 3px;
    background-color: white;
-    transition: all 0.3s ease;
+    transition: var(--transition-slow);
    border-radius: 50%;
-    box-shadow: 0 1px 2px rgba(0, 0, 0, 0.2);
+    box-shadow: var(--shadow-xs);
 }
 .inline-toggle-container .toggle-switch input:checked+.toggle-slider {
@@ -502,4 +502,323 @@
    opacity: 0.5;
    pointer-events: none;
    user-select: none;
 }
 /* File Count Badge on Version Items */
 .file-select-badge {
    display: inline-flex;
    align-items: center;
    gap: 4px;
    padding: 2px 8px;
    border-radius: 10px;
    background: oklch(var(--lora-accent) / 0.18);
    color: var(--lora-accent);
    font-size: inherit;
    font-weight: 600;
    cursor: pointer;
    transition: var(--transition-base);
    border: 1px solid oklch(var(--lora-accent) / 0.35);
    user-select: none;
    box-shadow: 0 1px 2px oklch(var(--lora-accent) / 0.1);
 }
 .file-select-badge:hover {
    background: oklch(var(--lora-accent) / 0.3);
    border-color: var(--lora-accent);
    transform: scale(1.05);
    box-shadow: 0 2px 6px oklch(var(--lora-accent) / 0.2);
 }
 .file-select-badge:active {
    transform: scale(0.98);
 }
 .file-select-badge i {
    font-size: 0.9em;
 }
 .file-select-badge .badge-arrow {
    margin-left: 2px;
    font-size: 0.65em;
    opacity: 0.7;
 }
 /* File Selection Step */
 .file-selection-header {
    margin-bottom: var(--space-3);
 }
 .file-selection-header h3 {
    margin: 0 0 4px 0;
    font-size: 1.1em;
    color: var(--text-color);
 }
 .file-selection-version-name {
    font-size: 0.9em;
    color: var(--text-color);
    opacity: 0.7;
 }
 .file-selection-list {
    max-height: 360px;
    overflow-y: auto;
    margin: var(--space-2) 0;
    display: flex;
    flex-direction: column;
    gap: 8px;
 }
 .file-option {
    display: flex;
    align-items: center;
    gap: 12px;
    padding: 12px;
    border: 1px solid var(--border-color);
    border-radius: var(--border-radius-sm);
    cursor: pointer;
    transition: var(--transition-base);
    background: var(--bg-color);
 }
 .file-option:hover {
    border-color: var(--lora-accent);
    box-shadow: var(--shadow-sm);
 }
 .file-option.selected {
    border: 2px solid var(--lora-accent);
    background: oklch(var(--lora-accent) / 0.05);
 }
 .file-option-radio {
    flex-shrink: 0;
 }
 .file-option-radio input[type="radio"] {
    width: 16px;
    height: 16px;
    accent-color: var(--lora-accent);
    cursor: pointer;
 }
 .file-option-info {
    flex: 1;
    min-width: 0;
 }
 .file-option-tags {
    display: flex;
    flex-wrap: wrap;
    gap: 6px;
    margin-bottom: 4px;
 }
 .file-tag {
    display: inline-block;
    padding: 2px 7px;
    border-radius: 4px;
    font-size: 0.8em;
    font-weight: 500;
    line-height: 1.4;
 }
 .file-tag.format {
    background: oklch(var(--lora-accent) / 0.1);
    color: var(--lora-accent);
 }
 .file-tag.fp {
    background: oklch(0.6 0.15 250 / 0.1);
    color: oklch(0.55 0.15 250);
 }
 .file-tag.size {
    background: oklch(0.55 0.1 160 / 0.1);
    color: oklch(0.5 0.12 160);
 }
 .file-option-name {
    font-size: 0.8em;
    color: var(--text-color);
    opacity: 0.6;
    white-space: nowrap;
    overflow: hidden;
    text-overflow: ellipsis;
    margin-top: 2px;
 }
 .file-option-size {
    font-size: 0.9em;
    color: var(--text-color);
    white-space: nowrap;
    font-variant-numeric: tabular-nums;
 }
 /* Dark theme adjustments */
 [data-theme="dark"] .file-option {
    background: var(--lora-surface);
 }
 [data-theme="dark"] .file-tag.fp {
    background: oklch(0.55 0.12 250 / 0.15);
    color: oklch(0.7 0.12 250);
 }
 [data-theme="dark"] .file-tag.size {
    background: oklch(0.5 0.08 160 / 0.15);
    color: oklch(0.65 0.08 160);
 }
 /* Textarea for multi-URL input */
 #modelUrl {
    width: 100%;
    padding: 8px;
    border: 1px solid var(--border-color);
    border-radius: var(--border-radius-xs);
    background: var(--bg-color);
    color: var(--text-color);
    font-family: var(--font-mono);
    font-size: 0.9em;
    resize: vertical;
    line-height: 1.5;
 }
 .input-hint {
    display: flex;
    align-items: center;
    gap: 6px;
    color: var(--text-color);
    opacity: 0.7;
    font-size: 0.85em;
    margin-top: 6px;
 }
 .input-hint i {
    color: var(--lora-accent);
 }
 /* Batch Preview List */
 .batch-preview-list {
    max-height: 400px;
    overflow-y: auto;
    margin: var(--space-2) 0;
    display: flex;
    flex-direction: column;
    gap: 1px;
    background: var(--border-color);
    border: 1px solid var(--border-color);
    border-radius: var(--border-radius-sm);
 }
 .batch-preview-item {
    display: flex;
    align-items: center;
    gap: 12px;
    padding: 10px 12px;
    background: var(--bg-color);
 }
 .batch-preview-item:first-child {
    border-radius: var(--border-radius-sm) var(--border-radius-sm) 0 0;
 }
 .batch-preview-item:last-child {
    border-radius: 0 0 var(--border-radius-sm) var(--border-radius-sm);
 }
 .batch-preview-item:only-child {
    border-radius: var(--border-radius-sm);
 }
 .batch-preview-thumbnail {
    width: 48px;
    height: 48px;
    flex-shrink: 0;
    border-radius: var(--border-radius-xs);
    overflow: hidden;
    background: var(--lora-surface);
 }
 .batch-preview-thumbnail img {
    width: 100%;
    height: 100%;
    object-fit: cover;
 }
 .batch-preview-icon {
    width: 48px;
    height: 48px;
    flex-shrink: 0;
    display: flex;
    align-items: center;
    justify-content: center;
    color: var(--lora-error);
    font-size: 1.2em;
 }
 .batch-preview-info {
    flex: 1;
    min-width: 0;
 }
 .batch-preview-name {
    font-weight: 500;
    color: var(--text-color);
    white-space: nowrap;
    overflow: hidden;
    text-overflow: ellipsis;
 }
 .batch-preview-meta {
    display: flex;
    gap: 8px;
    font-size: 0.85em;
    color: var(--text-color);
    opacity: 0.7;
    margin-top: 2px;
 }
 .batch-preview-error-text {
    color: var(--lora-error);
    opacity: 1;
 }
 .batch-preview-local-badge {
    color: var(--lora-accent);
    opacity: 1;
 }
 .batch-preview-local {
    opacity: 0.6;
 }
 .batch-preview-change-version {
    flex-shrink: 0;
    font-size: 0.85em;
    padding: 4px 10px;
 }
 .batch-preview-remove {
    flex-shrink: 0;
    background: none;
    border: none;
    color: var(--text-color);
    opacity: 0.5;
    cursor: pointer;
    padding: 4px 8px;
    font-size: 1em;
 }
 .batch-preview-remove:hover {
    opacity: 1;
    color: var(--lora-error);
 }
 .batch-preview-error {
    background: oklch(0.5 0.15 25 / 0.05);
 }
 [data-theme="dark"] .batch-preview-item {
    background: var(--lora-surface);
 }
--- a/static/css/components/modal/example-access-modal.css
+++ b/static/css/components/modal/example-access-modal.css
@@ -20,12 +20,12 @@
    border: 1px solid var(--lora-border);
    background-color: var(--lora-surface);
    cursor: pointer;
-    transition: all 0.2s;
+    transition: var(--transition-base);
 }
 .example-option-btn:hover {
    transform: translateY(-2px);
-    box-shadow: 0 4px 8px rgba(0, 0, 0, 0.1);
+    box-shadow: var(--shadow-md);
    border-color: var(--lora-accent);
 }
@@ -68,5 +68,5 @@
 /* Dark theme adjustments */
 [data-theme="dark"] .example-option-btn:hover {
-    box-shadow: 0 4px 12px rgba(0, 0, 0, 0.25);
+    box-shadow: var(--shadow-elevated);
 }
--- a/static/css/components/modal/help-modal.css
+++ b/static/css/components/modal/help-modal.css
@@ -32,7 +32,7 @@
  color: var(--text-color);
  cursor: pointer;
  font-weight: 500;
-  transition: all 0.2s;
+  transition: var(--transition-base);
  opacity: 0.7;
 }
@@ -150,7 +150,7 @@
  margin-left: 8px;
  vertical-align: middle;
  animation: fadeIn 0.5s ease-in-out;
-  box-shadow: 0 1px 3px rgba(0, 0, 0, 0.2);
+  box-shadow: var(--shadow-sm);
  text-transform: uppercase;
  letter-spacing: 0.5px;
 }
@@ -164,7 +164,7 @@
 /* Dark theme adjustments for new content badge */
 [data-theme="dark"] .new-content-badge {
-  box-shadow: 0 1px 3px rgba(0, 0, 0, 0.4);
+  box-shadow: var(--shadow-lg);
 }
 /* Update video list styles */
@@ -210,7 +210,7 @@
  margin-left: 10px;
  vertical-align: middle;
  animation: fadeIn 0.5s ease-in-out;
-  box-shadow: 0 2px 4px rgba(0, 0, 0, 0.1);
+  box-shadow: var(--shadow-md);
 }
 .update-date-badge i {
@@ -225,7 +225,7 @@
 /* Dark theme adjustments */
 [data-theme="dark"] .update-date-badge {
-  box-shadow: 0 2px 4px rgba(0, 0, 0, 0.3);
+  box-shadow: var(--shadow-md);
 }
 /* Privacy-friendly video embed styles */
@@ -281,7 +281,7 @@
  border-radius: var(--border-radius-sm);
  font-weight: 500;
  cursor: pointer;
-  transition: all 0.2s ease;
+  transition: var(--transition-base);
  background-color: var(--lora-accent);
  color: white;
  text-decoration: none;
@@ -303,5 +303,5 @@
 /* Dark theme adjustments */
 [data-theme="dark"] .video-container {
-  background-color: rgba(255, 255, 255, 0.03);
+  background-color: var(--surface-hover);
 }
--- a/Show More
+++ b/Show More