SkillsAdd

smithery/neversight

modelscope-zimage-generator

Generate images using ModelScope Z-Image models (Z-Image-Turbo, Z-Image-Base, Z-Image-Edit). Use when user asks to generate images, create artwork, or requests image generation functionality. Supports async generation with polling and optional LoRA configurations.

Installation

$ npx skills add smithery/neversight --skill modelscope-zimage-generator
View on skills.sh ↗ Website ↗

Similar popular skills

Related neighbors and high-traction skills in the same topics — useful to compare before installing.

ms-studio-deploy Similar name
modelscope.cn

Install the ms-studio-deploy skill for your AI agent. Published on modelscope.cn.

180 installs
modelscope Similar name
membranedev/application-skills

ModelScope integration. Manage data, records, and automate workflows. Use when the user wants t…

129 installs

Also in this package

Other skills from smithery/neversight · top by installs.

npx skills add smithery/neversight
zoom-meeting 0 ziwei-doushu 0 zimage-skill 0 zig 0 zhimeng-agent 0 zfc 0 zenbin 0 zeberg-salomon-rotator 0 zdev 0 youtube-transcript 0

Browse all from smithery/neversight

More details

Agent compatibility

Declared targets from SKILL.md / docs. Unmarked agents are not listed — the skill may still install via the CLI.

Claude Code Declared
Cursor Not declared
Codex Not declared
GitHub Copilot Not declared
Windsurf Not declared
Gemini CLI Not declared
Cline Not declared
OpenCode Not declared

Also listed on

Alternate registries and mirrors of this skill.

Skill metadata

Parsed from SKILL.md frontmatter.

Declared agents claude-code

Package contents

Files included with this skill beyond the listing page.

  • skill md SKILL.md 2,338 B
  • docs SUMMARY.md 299 B

History

  1. First recorded snapshot · 0 installs

SKILL.md

ModelScope Z-Image Generator

Generate images using ModelScope's Z-Image series models.

IMPORTANT: Response Format

After generating an image, you MUST:

  1. Use the Python script to generate the image
  2. Respond with ONLY a simple text message like: "Image saved to: /path/to/output.jpg"
  3. DO NOT include the image URL in your response
  4. DO NOT try to display or reference the image URL - this will cause errors with non-multimodal models

The user can view the generated image at the file path you provide.

Prerequisites

Get your ModelScope API key from: https://modelscope.cn/my/myaccesstoken

The script will automatically prompt for your API key the first time you use it and offer to save it for future use.

Optional: Manual Configuration

You can also manually configure the API key:

Config file:

mkdir -p ~/.config/modelscope
cat > ~/.config/modelscope/config.json << EOF
{
  "api_key": "ms-your-api-key-here"
}
EOF

Environment variable:

export MODELSCOPE_API_KEY="ms-your-api-key-here"

Quick Start

Use the Python script to generate images:

cd /Users/ningoo/.claude/skills/modelscope-image-generator/scripts
python generate_image.py "A golden cat" output.jpg

Models

Available Z-Image models:

  • Tongyi-MAI/Z-Image-Turbo (default, fast generation)
  • Tongyi-MAI/Z-Image-Base (higher quality)
  • Tongyi-MAI/Z-Image-Edit (image editing)

Specify a different model using the third argument:

python generate_image.py "A golden cat" output.jpg "Tongyi-MAI/Z-Image-Base"

LoRA Support

Single LoRA:

"loras": "<lora-repo-id>"

Multiple LoRAs (weights must sum to 1.0):

"loras": {"<lora-id1>": 0.6, "<lora-id2>": 0.4}

API Flow

  1. Submit generation request with X-ModelScope-Async-Mode: true
  2. Receive task_id in response
  3. Poll /v1/tasks/{taskid} with X-ModelScope-Task-Type: imagegeneration
  4. Wait for status SUCCEED or FAILED
  5. Download image from output_images[0] URL