feat(06-03): pass custom parameters through generation pipeline

- Include parameters in requestPayload for nanoBanana and generateVideo nodes
- Add logging for custom parameters in Replicate and fal.ai provider functions
- Parameters already spread via ...input.parameters in existing provider code

src/app/api/generate/route.ts

@@ -278,6 +278,11 @@ async function generateWithReplicate(
     ...input.parameters,
   };
 
+  // Log parameters being passed
+  if (input.parameters && Object.keys(input.parameters).length > 0) {
+    console.log(`[API:${requestId}] Custom parameters:`, JSON.stringify(input.parameters));
+  }
+
   // Add image input if provided (for img2img workflows)
   // Note: Different Replicate models use different parameter names
   // Using 'image' as it's most common for img2img models
@@ -461,6 +466,11 @@ async function generateWithFal(
     ...input.parameters,
   };
 
+  // Log parameters being passed
+  if (input.parameters && Object.keys(input.parameters).length > 0) {
+    console.log(`[API:${requestId}] Custom parameters:`, JSON.stringify(input.parameters));
+  }
+
   // Add image_url if provided (for img2img workflows)
   // fal.ai accepts both URLs and data URIs in this field
   if (input.images && input.images.length > 0) {

src/store/workflowStore.ts

@@ -1072,6 +1072,7 @@ export const useWorkflowStore = create<WorkflowStore>((set, get) => ({
                 model: nodeData.model,
                 useGoogleSearch: nodeData.useGoogleSearch,
                 selectedModel: nodeData.selectedModel,
+                parameters: nodeData.parameters,
               };
 
               // Build headers with API keys for external providers
@@ -1277,6 +1278,7 @@ export const useWorkflowStore = create<WorkflowStore>((set, get) => ({
                 images,
                 prompt: text,
                 selectedModel: nodeData.selectedModel,
+                parameters: nodeData.parameters,
               };
 
               // Build headers with API keys for external providers