diff --git a/optimized/image-edit/flux2klein-9b-character-sheet-4view/ui-workflow.json b/optimized/image-edit/flux2klein-9b-character-sheet-4view/ui-workflow.json
index 03ebdd5..af50d12 100644
--- a/optimized/image-edit/flux2klein-9b-character-sheet-4view/ui-workflow.json
+++ b/optimized/image-edit/flux2klein-9b-character-sheet-4view/ui-workflow.json
@@ -286,7 +286,9 @@
         "Node name for S&R": "CLIPTextEncode"
       },
       "widgets_values": [
-        "Use the provided reference image as the authoritative visual source. Create a professional multi-view reference sheet of the same subject. Infer the subject's identity, presentation, anatomy, proportions, facial features, hair, clothing or surface design, accessories, colors, materials, markings, pose, and every important visible detail directly from the reference image, regardless of the source image's gender, style, medium, or level of realism. Preserve those inferred attributes consistently across all views. Arrange exactly four equally sized full-subject views horizontally in one clean landscape sheet: front view, left side view, right side view, and back view. Use a neutral studio background, even soft lighting, consistent scale and alignment, realistic coherent anatomy or structure, natural hands when applicable, no extra subjects, no borders, no labels, no logos, no readable text, and no watermark. Do not redesign, simplify, gender-swap, restyle, or add or remove important features."
+
+
+        "Use the provided reference image as the authoritative visual source. Create a professional multi-view reference sheet of the same subject. Infer the subject's identity, presentation, anatomy, proportions, facial features, hair, clothing or surface design, accessories, colors, materials, markings, pose, and every important visible detail directly from the reference image, regardless of the source image's gender, style, medium, or level of realism. Preserve those inferred attributes consistently across all views. Arrange exactly four equally sized full-subject views horizontally in this exact order: panel 1 front view; panel 2 side view with the subject's nose and face pointing toward the RIGHT edge of the image; panel 3 side view with the subject's nose and face pointing toward the LEFT edge of the image; panel 4 back view. Panels 2 and 3 must face away from each other in opposite horizontal directions and must not be duplicates or the same profile mirrored incorrectly. Use a neutral studio background, even soft lighting, consistent scale and alignment, realistic coherent anatomy or structure, natural hands when applicable, no extra subjects, no borders, no labels, no logos, no readable text, and no watermark. Do not redesign, simplify, gender-swap, restyle, or add or remove important features."
       ]
     },
     {
diff --git a/optimized/image-edit/flux2klein-9b-character-sheet-4view/workflow.json b/optimized/image-edit/flux2klein-9b-character-sheet-4view/workflow.json
index 248a20e..c147221 100644
--- a/optimized/image-edit/flux2klein-9b-character-sheet-4view/workflow.json
+++ b/optimized/image-edit/flux2klein-9b-character-sheet-4view/workflow.json
@@ -123,7 +123,7 @@
   "74": {
     "class_type": "CLIPTextEncode",
     "inputs": {
-      "text": "Use the provided reference image as the authoritative visual source. Create a professional multi-view reference sheet of the same subject. Infer the subject's identity, presentation, anatomy, proportions, facial features, hair, clothing or surface design, accessories, colors, materials, markings, pose, and every important visible detail directly from the reference image, regardless of the source image's gender, style, medium, or level of realism. Preserve those inferred attributes consistently across all views. Arrange exactly four equally sized full-subject views horizontally in one clean landscape sheet: front view, left side view, right side view, and back view. Use a neutral studio background, even soft lighting, consistent scale and alignment, realistic coherent anatomy or structure, natural hands when applicable, no extra subjects, no borders, no labels, no logos, no readable text, and no watermark. Do not redesign, simplify, gender-swap, restyle, or add or remove important features.",
+      "text": "Use the provided reference image as the authoritative visual source. Create a professional multi-view reference sheet of the same subject. Infer the subject's identity, presentation, anatomy, proportions, facial features, hair, clothing or surface design, accessories, colors, materials, markings, pose, and every important visible detail directly from the reference image, regardless of the source image's gender, style, medium, or level of realism. Preserve those inferred attributes consistently across all views. Arrange exactly four equally sized full-subject views horizontally in this exact order: panel 1 front view; panel 2 side view with the subject's nose and face pointing toward the RIGHT edge of the image; panel 3 side view with the subject's nose and face pointing toward the LEFT edge of the image; panel 4 back view. Panels 2 and 3 must face away from each other in opposite horizontal directions and must not be duplicates or the same profile mirrored incorrectly. Use a neutral studio background, even soft lighting, consistent scale and alignment, realistic coherent anatomy or structure, natural hands when applicable, no extra subjects, no borders, no labels, no logos, no readable text, and no watermark. Do not redesign, simplify, gender-swap, restyle, or add or remove important features.",
       "clip": [
         "71",
         0