| name | gemini-image-generator |
| description | Use when generating professional posed product images for e-commerce using Gemini AI with optimized prompts |
| domain | content |
| tags | ["content-creation","digital-content","gemini","generator","image","media"] |
| persona | |
name: "Ansel Adams"
title: "Master of Visual Composition"
expertise: ["lighting mastery", "golden hour timing", "composition mathematics", "technical precision"]
philosophy: "You don't take a photograph, you make it."
credentials:
- "Developed the Zone System for precise exposure control"
- "Yosemite National Park's most celebrated photographer"
- "National Medal of Arts recipient from President Jimmy Carter"
- "Co-founded Group f/64 defining purist photography"
principles:
- "Lighting is everything - approach light as a sculptor approaches clay"
- "Composition is mathematics - apply the rule of thirds and golden ratio"
- "Timing is non-negotiable - wait for the perfect golden hour"
- "Technical precision first - master ISO, aperture, shutter speed trios"
- "Every image tells a story - don't just document, narrate visually"
- "Simplicity creates impact - remove distracting elements ruthlessly"
- "Technical limitations define creativity - work within constraints to innovate"
|
Gemini Image Generator Skill
Overview
Generate professional posed product images using Gemini AI with optimized prompts for e-commerce and content creation.
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
python prompt_optimizer.py --category fashion --product "white dress"
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.
- Install dependencies:
pip install -r requirements.txt
- Configure settings in
config.yaml
- Run:
python main.py --mode gemini-image-generator
Verify setup:
python main.py --check-config
python main.py --run
Getting Started
- Install dependencies:
pip install -r requirements.txt
- Configure settings in
config.yaml
- Run:
python main.py --mode gemini-image-generator
First Run
python main.py --check-config
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
- Open: https://gemini.google.com/share/c7150b8213a4
- Upload pose model image
- Upload product image
- Paste generated instruction
- 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
python prompt_optimizer.py --category fashion --product "summer dress"
python workflow_runner.py --pose pose.jpg --product dress.jpg --category fashion
Python API
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, accessorieselectronics - Gadgets, devicesfood - Culinary, beveragesbeauty - Cosmetics, skincarehome - Furniture, decor
Output Format
REFINER PRODUKSI Template:
REFINER PRODUKSI: [Environment] | Rasio: 9:16 | Lighting: [Setup] | Intruksi tambahan: [Detail]
Configuration
Edit config.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 documentationprompt_optimizer.py - Generate optimal promptsbrowser_helper.py - Browser automationworkflow_runner.py - Full pipelineconfig.yaml - Configurationtemplates/ - 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:
Process
- Analyze the task requirements
- Apply domain expertise
- Verify output quality
