Monggiiiii

comfyui-zimage

0
0
# Install this skill:
npx skills add Monggiiiii/comfyui-zimage-skill

Or install specific skill: npx add-skill https://github.com/Monggiiiii/comfyui-zimage-skill

# Description

|

# SKILL.md


name: comfyui-zimage
description: |
Z-Image Manga/Illustration Generation. Local scripts, no ToolSearch needed.
Trigger words: Hojo, Satoshi, manga, generate image, portrait.


Z-Image Image Generation System

⚠️ Do not use ToolSearch! Run scripts directly with Bash.


⚠️ Keyword Understanding (Must Read!)

User Says Meaning Script to Use
high quality, masterpiece, professional Good Prompt quality *_quick.py
high-res, upscale, 4K, large image High resolution *_highres.py or upscale_*.py
default (no specification) Quick preview *_quick.py

πŸ”€ Step 1: Identify Route (Important! Check Priority)

⚠️ Route Priority Rules

Priority Keyword Route Description
1 (Highest) Riko riko Always use riko route when "Riko" is mentioned!
2 Hojo, Tsukasa Hojo manga Pure manga style
3 Satoshi, Urushihara manga Pure manga style

Key Understanding

"Riko in Hojo style" = riko route + Hojo LoRA overlay

Riko's fixed appearance (hairstyle, facial features) and outfit definitions only exist in generate_riko_*.py.
If you use generate_manga_*.py, you'll lose Riko's characteristics!

⚠️ Prompt Extraction Rules When Using Riko

When user requests "Extract prompt from this image, then generate in Riko style":

Only extract these elements for --scene parameter:
- βœ… Actions, poses (sitting, standing, looking at viewer...)
- βœ… Emotions, expressions (smile, serious, melancholy...)
- βœ… Scene, background (classroom, window, sunset...)
- βœ… Lighting, atmosphere (soft lighting, dramatic shadow...)
- βœ… Composition, angle (close-up, full body, from below...)

Must discard these elements (Riko has fixed definitions):
- ❌ Hair color, hairstyle (blonde, ponytail, long hair...)
- ❌ Skin tone (tanned, pale...)
- ❌ Body type (slim, curvy...)
- ❌ Facial features (blue eyes, big eyes...)
- ❌ Clothing (unless user specifies maid instead of school_uniform)


🎨 Step 2: Execute Generation

DIR=/path/to/your/comfyui-zimage-skill

2.1 Riko Route (Use this whenever Riko is mentioned!)

# Pure Riko (realistic style)
python3 $DIR/generate_riko_quick.py \
  --outfit_type school_uniform \
  --scene "[scene description]"

# Riko + Hojo style (black and white manga-style Riko)
python3 $DIR/generate_riko_quick.py \
  --outfit_type school_uniform \
  --scene "[scene description]" \
  --lora "Z-image\\z-image-hojo.safetensors" \
  --style_trigger "Zanshou_kin_Hojo, monochrome, greyscale"

# Riko + Satoshi style (90s anime-style Riko)
python3 $DIR/generate_riko_quick.py \
  --outfit_type school_uniform \
  --scene "[scene description]" \
  --lora "Z-image\\z-image-satoshi.safetensors" \
  --style_trigger "sato, 90s anime style" \
  --style_tags "shiny skin, sparkle"

Outfit options: school_uniform (JK uniform) | maid (maid outfit)
View options: --view_type full (full body) | --view_type upper (upper body)

2.2 Pure Manga Route (When Riko is not involved)

# Hojo style
python3 $DIR/generate_manga_quick.py \
  --lora "Z-image\\z-image-hojo.safetensors" \
  --prompt "[full prompt]"

# Satoshi style
python3 $DIR/generate_manga_quick.py \
  --lora "Z-image\\z-image-satoshi.safetensors" \
  --prompt "[full prompt]"

2.3 High-Resolution Version (When user explicitly requests "high-res")

Replace *_quick.py with *_highres.py, requires timeout=300000


πŸ“€ Step 3: Display Results

<a href="[image_url]"><img src="[image_url]" width="350"></a>

Save image_url for later upscaling!


πŸ” Step 4: Upscale Existing Images

When user says "upscale" for an already generated image:

# Riko image upscale (no LoRA needed)
python3 $DIR/upscale_riko.py --image_url "[image_url]"

# Manga image upscale (needs original LoRA)
python3 $DIR/upscale_manga.py --image_url "[image_url]" --lora "[original LoRA]"

πŸ“‹ Quick Decision Table

User Request Route Script LoRA
Draw Riko riko generate_riko_quick.py None
Riko in Hojo style riko generate_riko_quick.py hojo + style_trigger
Riko in Satoshi style riko generate_riko_quick.py satoshi + style_trigger
Mecha girl in Hojo style manga generate_manga_quick.py hojo
Cyberpunk in Satoshi style manga generate_manga_quick.py satoshi

⚠️ Notes

  1. Riko first β€” Always use riko route when "Riko" is mentioned
  2. LoRA path uses double backslash \\
  3. highres/upscale needs timeout=300000
  4. Style guides are in data/ β€” Do not delete

# README.md

🎨 comfyui-zimage-skill - Create Stunning Manga and Illustrations

Download

πŸš€ Getting Started

Welcome to ComfyUI Z-Image Skill! This tool allows you to generate beautiful manga and illustration images using the ComfyUI API. Follow these steps to get started.

πŸ“‹ Prerequisites

Before you can run this application, make sure you have the following:

  • ComfyUI: You need to have ComfyUI running locally or on your server. You can download it from here.
  • Alma: This application requires Alma, which can be accessed from here.

πŸ“₯ Download & Install

  1. Visit the Releases Page: To download the application, go to the Releases page.

  2. Select the Latest Release: Look for the newest version at the top of the page. Click on it to view available downloads.

  3. Download the Files: Under the release, you will find various files. Choose the one that fits your operating system, usually labeled for Windows, macOS, or Linux.

  4. Run the Application: Once downloaded, locate the file in your downloads folder. Double-click to run it. Follow any instructions that may appear.

✨ Features

ComfyUI Z-Image Skill comes packed with features designed for your creative needs:

  • Text-to-Image Generation: Easily convert text into images using the ComfyUI API.
  • Multiple Art Styles: Choose from various styles, including:
  • πŸ–ŒοΈ Hojo Tsukasa: An 80s manga style, perfect for black and white artwork.
  • 🎨 Satoshi Urushihara: A colorful, 90s anime style.
  • Character Consistency: Create images with a character that has a fixed appearance across different artworks.
  • Quick & High-Res Modes: Get fast previews or choose production-quality images.
  • Upscaling Support: Enhance your generated images with Ultimate SD Upscale.

βš™οΈ How to Use

  1. Open the Application: After running the application, you will see its interface.

  2. Select Your Options: Choose the art style and input your text.

  3. Generate Your Image: Click the generate button to create your artwork.

  4. Save Your Creation: Once generated, you can save your image to your device.

πŸ”§ Troubleshooting

If you run into issues:

  • Check Requirements: Make sure ComfyUI and Alma are set up correctly.
  • Confirm Versions: Ensure you are using the latest version of ComfyUI to avoid compatibility problems.
  • Search for Solutions: Visit community forums or the GitHub issues page for help from other users.

πŸ“ž Support

For any questions or feedback, feel free to reach out via Twitter: @norsizu. You can also explore more on Civitai.

πŸ“„ License

This project is licensed under the MIT License. You can view the full license here.

🌐 Additional Resources

Download

# Supported AI Coding Agents

This skill is compatible with the SKILL.md standard and works with all major AI coding agents:

Learn more about the SKILL.md standard and how to use these skills with your preferred AI coding agent.