A professional-grade image generation skill that leverages the ZhiPu GLM-Image API to transform text descriptions into high-definition visual assets.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install glm-image-gen
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install glm-image-gen using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is GLM-Image Generation Skill?
The GLM-Image Generation skill is a powerful tool designed for the Openclaw Skills ecosystem, enabling developers and creators to generate high-quality images from natural language prompts. It excels at creating scientific illustrations, high-quality portraits, and commercial-grade posters with precise control over quality and style. By integrating the latest GLM-Image and CogView models, it provides a robust solution for automated visual content creation.
This skill is built for performance and reliability, offering HD generation modes and built-in content safety filtering. Whether you are building an automated marketing pipeline or an AI-powered creative assistant, this skill provides the necessary programmatic interface to generate, filter, and save visual media efficiently within your Openclaw Skills environment.
GLM-Image Generation Skill Use Cases
- Creating high-fidelity AI art, concept sketches, and digital illustrations from text prompts.
- Generating technical and scientific illustrations for documentation and presentations.
- Designing commercial posters and social media graphics with specific aspect ratios.
- Automating the creation of visual assets for web and mobile applications using Openclaw Skills.
- Visualizing user descriptions in real-time during AI-assisted chat sessions.
How GLM-Image Generation Skill Works
- The user provides a text description and optional configuration parameters like resolution or model type.
- The skill invokes a specialized Python CLI script that interfaces with the ZhiPu BigModel platform.
- The request is authenticated using the ZHIPU_API_KEY configured in the Openclaw Skills environment.
- The GLM-Image model processes the prompt to generate an image, applying content safety checks simultaneously.
- The API returns a temporary URL for the image, which remains valid for 30 days.
- If a local path is specified, the skill automatically downloads and saves the generated file for permanent storage.
GLM-Image Generation Skill Setup
To get started with this skill in Openclaw Skills, ensure you have a valid ZhiPu API key. You can configure the environment variable in your shell or within your global configuration file.
Set the environment variable:
export ZHIPU_API_KEY="your_secret_api_key"
Run a test generation via the CLI:
python scripts/glm_image_cli.py --prompt "A futuristic city in cyberpunk style" --size 1728x960 --quality hd
GLM-Image Generation Skill Data Schema & Taxonomy
The skill returns a structured JSON object containing all relevant metadata for the generated asset. This ensures seamless integration with other tools in the Openclaw Skills library.
| Attribute | Type | Description |
|---|---|---|
ok |
Boolean | Indicates if the generation was successful. |
image_url |
String | The temporary cloud URL for the generated image. |
saved_file |
String | The absolute local path where the image is stored (if requested). |
content_filter |
Array | Safety analysis results showing potential violations. |
size |
String | The dimensions of the generated image (e.g., 1280x1280). |
model |
String | The specific model used (glm-image, cogview-4, etc.). |
GLM-Image Generation Skill Advanced Features
- Support for multiple state-of-the-art models including glm-image, cogview-4, and cogview-3-flash.
- Advanced aspect ratio control with presets for square, landscape (3:2), portrait (2:3), and ultra-wide (16:9) formats.
- Professional watermark control for commercial use cases (requires signed disclaimer).
- Integrated content moderation with severity levels (0-3) to ensure compliant output within Openclaw Skills.
- Local file persistence options to bypass the 30-day expiration of temporary cloud URLs.
SKILL.md
Loading
Related Openclaw Skills
MIMO Systems & Wireless Architecture Specialist
A specialized technical skill for AI agents focusing on the physics, architecture, and deployment of Multiple-Input Multiple-Output wireless communication systems.
Find Partners
A strategic partnership discovery tool designed to identify, shortlist, and rank potential business allies for ecommerce and brand growth.
QSearch Agent
A versatile universal search assistant that delivers safe, filtered, and ad-free results for movies, stocks, news, and more using simple slash commands.
Find Creators
A high-signal discovery tool for identifying and shortlisting influencers, TikTok Shop partners, and content creators tailored to specific ecommerce goals.
Contract Guardian
A professional AI contract review assistant designed to identify legal risks, extract key information, and track expiration dates automatically.
IP Geolocation Toolkit
A zero-dependency Python toolkit for performing public IP discovery, geolocation lookups, and reverse DNS queries via free APIs.
Featured*
