Art Prompt Engine: CineRealism (Scripted)

[SYSTEM OVERRIDE: ROLEPLAY DISABLED] Transform AI Dungeon into a professional-grade Image Prompt Architect. This is not a game; it is a scripted engine designed to generate high-fidelity, photorealistic art prompts for Stable Diffusion, Midjourney, and Flux. // FEATURES: Automated Canon Research: Deep analysis of character lore. Dynamic Posing Engine: Enforces action-oriented, non-static composition. Lens & Lighting Logic: Auto-selects real-world camera lenses (e.g., Sigma 35mm). "Lottery" Command: Type lottery to retrieve alternative concept sketches. // USAGE: Simply enter a character name (e.g., "Batman") or concept. The engine will output a ready-to-paste prompt block. //BEST MODELS: // DeepSeek V3.2 is currently the most sophisticated model for descriptive prose. While Hermes 3 (405B) is smarter in raw logic, DeepSeek is superior in Aesthetic Narrative. It is specifically optimized to avoid clichés and "atmospheric clutter," which is vital for high-end Art Prompts where every word counts. // Muse is designed for character intelligence and nuanced storytelling. It uses Direct Preference Optimization (DPO) to aggressively reduce AI clichés—this is critical because clichés in a prompt (e.g., "glowing eyes," "epic pose") often lead to low-quality image generations.