# Gemini Image Generator

> Use when generating professional posed product images for e-commerce using Gemini AI with optimized prompts

- Skill: `oyi77/gemini-image-generator` (Agent Skill)
- Install (CLI): `npx skillmds add oyi77/gemini-image-generator`
- Raw SKILL.md: https://api.skillmd.com/api/skills/oyi77/gemini-image-generator/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- License: Apache-2.0
- Author: oyi77 (https://skillmd.com/u/oyi77)
- Updated: 2026-09-08
- Page: https://skillmd.com/skills/oyi77/gemini-image-generator

---





# Gemini Image Generator Skill

## Overview

Generate professional posed product images using Gemini AI with optimized prompts for e-commerce and content creation.


## Anti-Rationalization Table

| Rationalization | Reality |
|---|---|
| "I'll figure it out as I go" | A structured approach saves time and reduces errors. Follow the workflow in this skill rather than improvising. |
| "I already know this topic" | Familiarity breeds shortcuts. Use the checklist to verify you haven't missed critical steps. |
| "This doesn't apply to my situation" | The patterns here generalize across contexts. Adapt, don't skip — the underlying principles hold. |
| "One more tool will fix it" | Adding complexity rarely solves process gaps. Master the core workflow first. |

## When to Use

**Trigger phrases:**
- "gemini image generator"
- "When you need to create product images with model poses for e-commerce"
- "When preparing TikTok/social media content featuring products"
- "When you have product images and need them "modeled" without hiring photographer"


- When you need to create product images with model poses for e-commerce
- When preparing TikTok/social media content featuring products
- When you have product images and need them "modeled" without hiring photographers
- When batch-generating multiple product variations

## When NOT to Use

- When you need photorealistic human faces (AI limitations)
- When legal compliance requires real photography
- When the product requires accurate color representation (AI may shift colors)
- When generating trademarked brand items

## Quick Reference

```bash
# Generate prompt
python prompt_optimizer.py --category fashion --product "white dress"

# Full workflow
python workflow_runner.py --pose pose.jpg --product dress.jpg --category fashion
```

## Common Mistakes

- Using low-resolution input images (results will be blurry)
- Not specifying lighting setup in prompts (inconsistent results)
- Skipping the ChatGPT video step (misses viral content optimization)
- Ignoring TikTok AI disclosure requirements

## Workflow

```
Input: Pose Model + Product Image
    ↓
Gemini Link (c7150b8213a4) → AI Image Generation
    ↓
Output: Multiple Posed Product Images
    ↓
ChatGPT → Video Scene Generation
    ↓
TikTok Content (compliant)
```

## Quick Start

Get started with gemini-image-generator in three steps.

1. Install dependencies: `pip install -r requirements.txt`
2. Configure settings in `config.yaml`
3. Run: `python main.py --mode gemini-image-generator`

Verify setup:
```bash
python main.py --check-config
python main.py --run
```


### Getting Started
1. Install dependencies: `pip install -r requirements.txt`
2. Configure settings in `config.yaml`
3. Run: `python main.py --mode gemini-image-generator`

### First Run
```bash
# Verify setup
python main.py --check-config
# Execute
python main.py --run
```


### 1. Prepare Images
- **Pose Model**: Reference pose image (person/model)
- **Product**: Product image to integrate
- **Format**: JPG/PNG, any size (Rasio 9:16 recommended)

### 2. Generate Instruction

Use the prompt optimizer:
```
python prompt_optimizer.py --category fashion --style minimal --product "white dress"
```

### 3. Run in Gemini

1. Open: https://gemini.google.com/share/c7150b8213a4
2. Upload pose model image
3. Upload product image
4. Paste generated instruction
5. Generate images

### 4. Download Results
- Gemini generates 4-8 variations
- Download all to output folder
- Next: Send to ChatGPT for video scenes

## Usage

- Configure commerce, gemini, generating, generator, image settings before first use


### Command Line
```bash
# Generate instruction only
python prompt_optimizer.py --category fashion --product "summer dress"

# Full workflow (with browser automation)
python workflow_runner.py --pose pose.jpg --product dress.jpg --category fashion
```

### Python API
```python
from prompt_optimizer import generate_instruction

instruction = generate_instruction(
    category="fashion",
    product_name="summer dress",
    style="minimal",
    lighting="soft"
)
print(instruction)
```

## Categories

Available templates:
- `fashion` - Clothing, accessories
- `electronics` - Gadgets, devices
- `food` - Culinary, beverages
- `beauty` - Cosmetics, skincare
- `home` - Furniture, decor

## Output Format

**REFINER PRODUKSI Template:**
```
REFINER PRODUKSI: [Environment] | Rasio: 9:16 | Lighting: [Setup] | Intruksi tambahan: [Detail]
```

## Configuration

Edit `config.yaml`:
```yaml
gemini:
  shared_link: "https://gemini.google.com/share/c7150b8213a4"
  output_ratio: "9:16"
  
categories:
  fashion:
    environments: ["clean white studio", "minimal gray", "premium lifestyle"]
    lighting: ["soft diffused", "studio lighting", "natural daylight"]
    
defaults:
  ratio: "9:16"
  style: "premium"
  lighting: "soft diffused"
```

## Integration

- Configure commerce, gemini, generating, generator, image settings before first use


### Next Step: ChatGPT Video Generation
Send Gemini output to ChatGPT with:
```
Generate video scene descriptions from these images. 
Style: TikTok viral, 15-30 seconds, hook in first 3 seconds.
```

### TikTok Compliance
- Disclosure: "AI-generated content"
- Label videos appropriately
- Follow community guidelines

## Files

- `SKILL.md` - This documentation
- `prompt_optimizer.py` - Generate optimal prompts
- `browser_helper.py` - Browser automation
- `workflow_runner.py` - Full pipeline
- `config.yaml` - Configuration
- `templates/` - Category templates

## License

MIT - Free for personal and commercial use.

## Common Rationalizations

| Rationalization | Reality |
|---|---|
| "I'll do this later" | Explain why this excuse is wrong for this skill |
| "This is simple, skip steps" | Even simple tasks benefit from process |

## Red Flags

- Generated images have visible artifacts or distorted text
- Agent does not verify images are appropriate for the intended context
- Watch for shortcuts and skipped steps

## Verification

After completing this skill, confirm:

- [ ] Generated images are free of visible artifacts and distortions
- [ ] Images are appropriate for the intended use context
- [ ] All required outputs generated
- [ ] Success criteria met

## Process

1. Analyze the task requirements
2. Apply domain expertise
3. Verify output quality