fix(ui): bulk update cache invalidation and model dropdown display (#633)

Fix two related issues with bulk model updates in Kanban view:

1. Bulk update now properly invalidates React Query cache
   - Changed handleBulkUpdate and bulk verify handler to call loadFeatures()
   - This ensures UI immediately reflects bulk changes

2. Custom provider models (GLM, MiniMax, etc.) now display correctly
   - Added fallback lookup in PhaseModelSelector by model ID
   - Updated mass-edit-dialog to track providerId after selection

apps/ui/src/components/views/board-view.tsx

@@ -636,10 +636,8 @@ export function BoardView() {
         const result = await api.features.bulkUpdate(currentProject.path, featureIds, finalUpdates);
 
         if (result.success) {
-          // Update local state
+          // Invalidate React Query cache to refetch features with server-updated values
-          featureIds.forEach((featureId) => {
+          loadFeatures();
-            updateFeature(featureId, finalUpdates);
-          });
           toast.success(`Updated ${result.updatedCount} features`);
           exitSelectionMode();
         } else {
@@ -655,7 +653,7 @@ export function BoardView() {
     [
       currentProject,
       selectedFeatureIds,
-      updateFeature,
+      loadFeatures,
       exitSelectionMode,
       getPrimaryWorktreeBranch,
       addAndSelectWorktree,
@@ -783,10 +781,8 @@ export function BoardView() {
       const result = await api.features.bulkUpdate(currentProject.path, featureIds, updates);
 
       if (result.success) {
-        // Update local state for all features
+        // Invalidate React Query cache to refetch features with server-updated values
-        featureIds.forEach((featureId) => {
+        loadFeatures();
-          updateFeature(featureId, updates);
-        });
         toast.success(`Verified ${result.updatedCount} features`);
         exitSelectionMode();
       } else {
@@ -798,7 +794,7 @@ export function BoardView() {
       logger.error('Bulk verify failed:', error);
       toast.error('Failed to verify features');
     }
-  }, [currentProject, selectedFeatureIds, updateFeature, exitSelectionMode]);
+  }, [currentProject, selectedFeatureIds, loadFeatures, exitSelectionMode]);
 
   // Handler for addressing PR comments - creates a feature and starts it automatically
   const handleAddressPRComments = useCallback(

apps/ui/src/components/views/board-view/dialogs/mass-edit-dialog.tsx

@@ -128,6 +128,7 @@ export function MassEditDialog({
   // Field values
   const [model, setModel] = useState<ModelAlias>('claude-sonnet');
   const [thinkingLevel, setThinkingLevel] = useState<ThinkingLevel>('none');
+  const [providerId, setProviderId] = useState<string | undefined>(undefined);
   const [planningMode, setPlanningMode] = useState<PlanningMode>('skip');
   const [requirePlanApproval, setRequirePlanApproval] = useState(false);
   const [priority, setPriority] = useState(2);
@@ -162,6 +163,7 @@ export function MassEditDialog({
       });
       setModel(getInitialValue(selectedFeatures, 'model', 'claude-sonnet') as ModelAlias);
       setThinkingLevel(getInitialValue(selectedFeatures, 'thinkingLevel', 'none') as ThinkingLevel);
+      setProviderId(undefined); // Features don't store providerId, but we track it after selection
       setPlanningMode(getInitialValue(selectedFeatures, 'planningMode', 'skip') as PlanningMode);
       setRequirePlanApproval(getInitialValue(selectedFeatures, 'requirePlanApproval', false));
       setPriority(getInitialValue(selectedFeatures, 'priority', 2));
@@ -226,10 +228,11 @@ export function MassEditDialog({
               Select a specific model configuration
             </p>
             <PhaseModelSelector
-              value={{ model, thinkingLevel }}
+              value={{ model, thinkingLevel, providerId }}
               onChange={(entry: PhaseModelEntry) => {
                 setModel(entry.model as ModelAlias);
                 setThinkingLevel(entry.thinkingLevel || 'none');
+                setProviderId(entry.providerId);
                 // Auto-enable model and thinking level for apply state
                 setApplyState((prev) => ({
                   ...prev,

apps/ui/src/components/views/settings-view/model-defaults/phase-model-selector.tsx

@@ -415,6 +415,44 @@ export function PhaseModelSelector({
       }
     }
 
+    // Fallback: Check ClaudeCompatibleProvider models by model ID only (when providerId is not set)
+    // This handles cases where features store model ID but not providerId
+    for (const provider of enabledProviders) {
+      const providerModel = provider.models?.find((m) => m.id === selectedModel);
+      if (providerModel) {
+        // Count providers of same type to determine if we need provider name suffix
+        const sameTypeCount = enabledProviders.filter(
+          (p) => p.providerType === provider.providerType
+        ).length;
+        const suffix = sameTypeCount > 1 ? ` (${provider.name})` : '';
+        // Add thinking level to label if not 'none'
+        const thinkingLabel =
+          selectedThinkingLevel !== 'none'
+            ? ` (${THINKING_LEVEL_LABELS[selectedThinkingLevel]} Thinking)`
+            : '';
+        // Get icon based on provider type
+        const getIconForProviderType = () => {
+          switch (provider.providerType) {
+            case 'glm':
+              return GlmIcon;
+            case 'minimax':
+              return MiniMaxIcon;
+            case 'openrouter':
+              return OpenRouterIcon;
+            default:
+              return getProviderIconForModel(providerModel.id) || OpenRouterIcon;
+          }
+        };
+        return {
+          id: selectedModel,
+          label: `${providerModel.displayName}${suffix}${thinkingLabel}`,
+          description: provider.name,
+          provider: 'claude-compatible' as const,
+          icon: getIconForProviderType(),
+        };
+      }
+    }
+
     return null;
   }, [
     selectedModel,