Global-AI-Group-Inc/content-generation-skills / skills/mira-image-prompting
mira-image-prompting
VerifiedHow to write stills for Mira AI's image models: GPT Image 2, 2.5 and 2.6, Nano Banana Pro and 2.1, Seedream 4, 4K and 5 Pro, Muse Image, Ideogram 4.5, Flux. Which model holds text, which keeps a product identical across a series, how the craft banks for light, optics, composition, grade and materials are used, and the words that make an image worse. Use it before any generate_image call on Mira, or when the user says "image prompt", "промпт для картинки", "нарисуй", "сгенерируй картинку", poster, packshot, banner, cover. NOT for video (mira-video-prompting).
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 4.2 KB |
| references/composition.md | markdown | 2.3 KB |
| references/craft.md | markdown | 3.0 KB |
| references/grade.md | markdown | 2.3 KB |
| references/light.md | markdown | 3.7 KB |
| references/materiality.md | markdown | 8.0 KB |
| references/models.md | markdown | 7.0 KB |
| references/optics.md | markdown | 2.4 KB |
references/optics.md
View on GitHub<!-- generated from the registry by comfy_study/comfy_video/build_skills.py; edit the registry, not this file -->
# Optics and shot size
OPTICS - describe the lens in words, not numbers: these models read 'compressed background and shallow depth of field' reliably and 'f/1.8' barely at all. A focal length may appear only as a feel ('an 85mm feel'). Pick ONE entry and state what is sharp and what is not. RESOLUTION GOVERNS TEXTURE, AND IT IS DECIDED BEFORE THE PROMPT: pores, peach fuzz and thread-level weave are worth asking for only on a large native canvas AND a tight shot. At 1728x3072 a face in close-up is about 1200 pixels across and there is something to draw them with; in a mid shot on the same frame it is about 400 pixels, and at 1024x1536 in a mid shot there is nothing. Asking for pores where the pixels cannot hold them is exactly what produces the plastic, over-smoothed face this platform is trying to avoid - so at small canvas or wide framing ask instead for accurate skin tone, visible fine lines and a natural specular highlight.
### Optics and shot size
- **`clean_modern`** — Modern prime, neutral rendering, no distortion, background falling off gently. The default when nothing else is called for.
- **`macro_100`** — Macro on a small subject, extremely shallow focus, one detail sharp and everything else dissolving. Texture, ingredients, fabric weave, jewellery.
- **`wide_24`** — Wide lens close to the subject, expanded space, slight edge stretch. Interiors, full-body in a location, scale.
- **`anamorphic_35`** — Anamorphic character: oval highlights, gentle horizontal flare, compressed depth. Cinematic and premium; wrong for a clean e-commerce card.
- **`vintage_spherical`** — Older glass, softer contrast, mild edge falloff, slightly lifted blacks. Warm and human rather than clinical.
- **`shallow_dof`** — Subject sharp, background dissolved to soft tone and shape. Isolates a hero object from a busy place.
- **`deep_focus`** — Everything sharp front to back. Flatlays, layouts, anything where several objects must all be read.
- **`tilt_shift`** — A single narrow plane in focus across the frame. Deliberate and graphic; not a substitute for shallow depth of field.
- **`phone_camera`** — Small sensor look: wide, deep focus, straightforward exposure, no cinematic separation. This is what makes UGC read as real - a shallow cinematic background instantly breaks it.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |