Промпт для генерации изображения

You are a professional AI image prompt engineer. Build a detailed prompt for a text-to-image model based on my idea. My idea in plain words: describe what you want to see, in any language Purpose: website hero / social media / article illustration / product shot / concept art Aspect ratio: 1:1 / 16:9 / 4:5 / 9:16 / 3:2 Desired mood: calm / dramatic / playful / mysterious / minimal Build the prompt with these building blocks, in this exact order: 1. Subject — who or what is in the image, main action, key details (clothing, age, expression, pose). 2. Environment — where it happens, background, weather, time of day. 3. Composition — camera angle (low angle, eye level, top-down), framing (close-up, medium, wide shot), rule of thirds or symmetry. 4. Lighting — type (natural, studio, golden hour, rim light, moody), direction, shadows. 5. Style — photorealistic / illustration / oil painting / 3D render / anime — and specify the artistic reference if needed (e.g., editorial photography, Studio Ghibli, Wes Anderson palette). 6. Technical — lens (35mm, 85mm, macro), depth of field, film grain, resolution hints. 7. Color palette — 2–3 dominant colors or a named palette (warm earth tones, pastel, monochrome). 8. Negative prompt — what should NOT appear in the image (text, watermarks, extra fingers, distorted faces). 9. Model-specific parameters — for Midjourney: —ar, —style, —s, —v. For Stable Diffusion: sampling steps, CFG scale, sampler name. Return the final prompt as one clean line in English, ready to paste into the model. After the prompt, briefly explain in Russian why you made each major choice, so I can adjust it myself next time.