oliverv/caro / .agents/skills/oma-image
oma-image
VerifiedGenerate raster images or reference-guided variations through the OMA image CLI. Use for image assets or explicit vendor comparisons; pixel editing uses an editor.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 7.8 KB |
| resources/checklist.md | markdown | 1.1 KB |
| resources/execution-protocol.md | markdown | 7.5 KB |
| resources/invocation.md | markdown | 6.8 KB |
| resources/prompt-tips.md | markdown | 4.6 KB |
| resources/vendor-matrix.md | markdown | 6.0 KB |
resources/prompt-tips.md
View on GitHub# Prompt Tips
Good prompts are specific about scene, subject, lighting, and style. `gpt-image-2` (codex), the Gemini-family "nano-banana" models (antigravity, chosen internally by agy), and flux/zimage (pollinations) all respond well to the same structural cues.
## Structure
```
Scene/backdrop → Subject → Details → Constraints
```
Example: `minimalist product photography: single white ceramic coffee cup on dark marble surface, steam rising softly, dramatic side lighting`
## Do's
- **Lighting**: "warm golden hour side light", "overcast diffused", "backlit with rim light"
- **Camera**: "shallow depth of field", "aerial view", "close-up macro", "35mm film grain"
- **Style**: "photorealistic", "oil painting", "3D render", "concept art", "isometric vector"
- **Mood**: "serene", "dramatic", "moody", "vibrant", "washed-out"
- **Resolution cue**: "ultra detailed", "8K", "high fidelity"
## Don'ts
- Avoid vague prompts like "a nice picture"; both models produce generic output.
- Don't stack contradicting styles ("photorealistic cel-shaded 3D").
- Skip negative prompts; none of the routed models treats them as first-class.
## Examples
| Category | Prompt |
|----------|--------|
| Product | `Elegant perfume bottle on reflective black surface, studio lighting, luxury brand catalog style` |
| Landscape | `Aerial drone shot of Jeju coastline, turquoise water meeting volcanic rock, golden hour` |
| Food | `Overhead flat-lay of Korean bibimbap in stone pot, steam rising, vibrant vegetables, dark wood table` |
| Architecture | `Modern minimalist house with floor-to-ceiling windows overlooking misty mountain valley` |
| Portrait | `Professional headshot, soft natural window light, shallow depth of field, neutral background` |
| UI Mockup | `iPhone 15 Pro mockup showing a fitness app dashboard, clean UI, dark mode, floating on gradient background` |
| Concept | `Lone astronaut on crater edge of Mars, looking at Earth rising on the horizon, cinematic, volumetric dust` |
## Vendor Nuances
- **Codex (`gpt-image-2`)**: prefers slightly longer, descriptive prompts. Quality flag `high` noticeably sharpens fine detail but doubles generation time.
- **Antigravity (Gemini-family "nano-banana" models, chosen internally by agy)**: robust on CJK-composed scenes ("hanok", "cherry blossom"); slightly stronger on illustrated/painterly styles. Driven via the `agy` CLI's agentic loop — output is JPEG by default even if `.png` is requested (the provider sniffs magic bytes and renames the extension accordingly).
## Comparing Output
`--vendor all` generates the same prompt on every registered provider and writes each result (PNG or JPEG, per provider) into a `…-compare/` folder with a single `manifest.json`. Use it for A/B picking when starting a new visual style.
## External Prompt Libraries (agent reference only)
When the **Clarification Protocol** (see `SKILL.md`) reaches the *amplify* step, these galleries provide structural references.
| Source | Vendor match | Prompts | Categories |
|--------|--------------|---------|-----------|
| [awesome-gpt-image-2](https://github.com/YouMind-OpenLab/awesome-gpt-image-2) | `codex` (gpt-image-2) | ~100 | Profile/Avatar · Social Post · Infographic · YouTube Thumbnail · Comic/Storyboard · Poster/Flyer · App/Web Design |
| [awesome-nano-banana-pro-prompts](https://github.com/YouMind-OpenLab/awesome-nano-banana-pro-prompts) | `antigravity` (nano-banana-pro) | 10,000+ | Same taxonomy; 16 localized READMEs (en, ko-KR, ja-JP, zh-TW, de-DE, …) |
For `pollinations` (flux / zimage), patterns from either library transfer; use the subject framing, lighting cues, and style-keyword structure, not the literal wording.
**How to use (agent-side):**
1. Classify the user's intent into one of the 7 categories. If no match, fall back to the *Scene/backdrop → Subject → Details → Constraints* template at the top of this file.
2. Fetch the relevant README section via `gh api`, e.g. Korean user + nano-banana:
```
gh api repos/YouMind-OpenLab/awesome-nano-banana-pro-prompts/contents/README_ko-KR.md \
--jq .content | base64 -d | less
```
Scan headings like `### No. N: <Category> - <Title>` to locate 1–2 analogous entries.
3. **Internalize the pattern, do not reproduce the text.** Extract lighting / framing / camera / style-keyword choices. The structural approach is not copyrightable; the expression is.
4. Compose your amplified prompt by applying that pattern to the user's subject in your own wording. Preserve the requested subject and constraints; proceed under existing generation authorization unless a material decision remains.
**Fallback:** If no category matches, use the structural template at the top of this file.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |