⚡ Master Prompt Directive

[COMFYUI AUTOMATED CHARACTER CONSISTENCY & LOOKBOOK PIPELINE]

NODE ENGINE: ComfyUI + LTX-2.5 Video/Image Enhancer LoRA V2 + IP-Adapter FaceID Plus v2
IMAGE RESOLUTION: 1080x1920 (Vertical Fashion) / 2048x1152 (Editorial)
DIFFUSION CHECKPOINT: Flux.1-Dev / SDXL Photoreal V2

NODE GRAPH CONFIGURATION:
1. IDENTITY ANCHOR NODE:
   - Load IP-Adapter FaceID Plus v2 (Weight: 0.85, Start Step: 0.0, End Step: 0.80)
   - Apply InsightFace vector extraction from high-res reference portrait
2. LTX-2.5 ENHANCER LORA V2 NODE:
   - Model Weight: 0.75, Clip Weight: 0.70
   - Purpose: Eliminates plastic surface smoothing, restores micro-pores, dermal subsurface scattering, and natural hair strands
3. BATCH CAMERA PROMPTING CYCLE:
   - Shot A (Hero Close-up): "Editorial fashion close-up, dramatic split lighting, high-fashion makeup, visible skin micro-texture, 85mm portrait lens."
   - Shot B (3/4 Profile Medium): "3/4 medium fashion shot, relaxed posture, silk designer blazer, studio seamless gray backdrop, rim lighting."
   - Shot C (Full Body Dynamic): "Full-length editorial walk, kinetic fabric flutter, natural gait, shallow depth of field, high-key studio lighting."

NEGATIVE PROMPT:
doll skin, airbrushed, cartoon, anime, plastic, duplicate face, deformed eyes, extra fingers, asymmetric face, bad anatomy, ghosting.