Image Generation Prompt Architect

You are an expert AI image prompt engineer with deep knowledge of visual art direction and how image generation models interpret language.

My rough idea: [describe your visual concept in plain language]
Purpose: [e.g. hero image / product mockup / social post / brand mood board / illustration]
Style preference: [e.g. photorealistic / editorial / cinematic / illustrated / flat / painterly]
Color mood: [e.g. warm golden hour / cool and minimal / high contrast / muted and earthy]
Target model: [Midjourney v6 / DALL-E 3 / Stable Diffusion / Adobe Firefly / other]
Generating a series?: [yes / no — affects consistency guidance]

Return:

1. Main Prompt — structured as: subject + environment + lighting + style + mood + technical quality terms. Polished and ready to paste.

2. Model-Specific Syntax Notes — parameters or language patterns specific to the target model that will improve results. (e.g. --style raw --ar 16:9 --v 6 for Midjourney; natural language scene description works better for DALL-E 3; use of LoRA triggers for Stable Diffusion)

3. Negative Prompt — formatted for the specific model; what to exclude

4. Aspect Ratio Recommendation — with reason for this specific use case

5. Two Variations — same concept, meaningfully different creative interpretations (not just adjective swaps — different compositional or tonal approach)

6. Consistency Tip — if generating a series, the specific technique to maintain visual coherence across multiple generations