A high-performance local image generation skill that interfaces with DrawThings to provide Stable Diffusion capabilities via an API on macOS.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install drawthings
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 drawthings 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 DrawThings Image Generation?
DrawThings Image Generation is a specialized tool within the Openclaw Skills ecosystem designed for developers and creators who require privacy-focused, local AI image generation. By leveraging the DrawThings application, this skill provides a robust programmatic interface to Stable Diffusion, specifically optimized for Apple Silicon using MLX and CoreML acceleration. This ensures that users can generate high-quality visual assets directly on their hardware without the need for cloud-based GPU subscriptions.
This skill serves as a bridge between your AI agent and the local generation engine, supporting an Automatic1111-compatible API. It is ideal for those who need to integrate visual content creation into their automated workflows, offering full control over models, samplers, and fine-grained generation settings while maintaining a local-first data policy.
DrawThings Image Generation Use Cases
- Rapidly prototyping concept art and UI elements from simple text prompts.
- Generating large batches of character or landscape variations for creative projects.
- Testing and comparing different Stable Diffusion models and samplers locally.
- Creating high-resolution, reproducible images using specific seeds and parameters.
- Integrating automated visual asset generation into broader Openclaw Skills development pipelines.
How DrawThings Image Generation Works
- The Openclaw Skills agent receives a request containing a text prompt and optional generation parameters.
- The skill formats this request into an Automatic1111-compatible JSON payload.
- A POST request is sent to the local DrawThings API server running on the user's Mac.
- DrawThings processes the image using hardware-accelerated MLX or CoreML frameworks.
- The resulting image is saved to the local file system as a PNG with full generation metadata embedded.
- The skill returns the file path and status to the agent for further action.
DrawThings Image Generation Setup
To integrate this capability into your collection of Openclaw Skills, ensure DrawThings is installed and the API server is enabled in the app preferences. Configure your environment by setting the API URL:
export DRAWTHINGS_URL="http://127.0.0.1:7860"
You can also apply this configuration directly through the agent CLI:
openclaw config set env.DRAWTHINGS_URL "http://127.0.0.1:7860"
Verify the connection by running a simple generation script to ensure the local server is responding.
DrawThings Image Generation Data Schema & Taxonomy
The DrawThings skill organizes generation data to ensure clarity and traceability within Openclaw Skills projects:
| Component | Detail | Format |
|---|---|---|
| Image Files | Final visual output saved locally | PNG |
| Embedded Metadata | Contains prompt, negative prompt, steps, and seed | EXIF |
| Filename Schema | Timestamp-based naming (e.g., drawthings_output_20231027_120000.png) | String |
| API Parameters | Configuration for CFG scale, sampler, and dimensions | JSON |
All data remains on the local machine, providing a secure environment for sensitive design work.
DrawThings Image Generation Advanced Features
- Support for complex workflows using the Automatic1111-compatible /sdapi/v1/txt2img endpoint.
- Integrated presets including Fast, Quality, and NFT for optimized one-click generation.
- Batch generation support to produce multiple variations without repeated script execution.
- Fine-grained control over advanced parameters like CFG scale, DPM++ samplers, and custom resolutions.
- Seed-based generation for perfectly reproducible visual results across different sessions.
SKILL.md
Loading
Related Openclaw Skills
ProtonMail Skill
A specialized toolset for reading, searching, and scanning ProtonMail accounts via secure IMAP bridges like Proton Bridge or hydroxide.
Duola Lobster Agent
An institutional-grade execution playbook for the Duola Polymarket copy-trading CLI within agent workflows.
med-info
A specialized tool for fetching authoritative, label-backed medication answers with traceable identifiers and regulatory safety context.
ACE-Step Music Generation Skill
A professional-grade AI music generation tool that leverages the ACE-Step V1.5 API to create full songs, vocals, and instrumentals from text prompts.
Apify Web Scraping & Data Extraction
A powerful tool for discovering and executing Apify Actors to scrape structured data from social media and websites.
Intercom Conversations Read
A specialized tool for listing, finding, and searching Intercom customer conversations via a structured JSON interface.
Featured*
