LaunchKit · 2026
Back to Skills

drawthings

Generate images with DrawThings (Stable Diffusion) via API.

1
1.2k downloads
by @dustinparsons

Setup & Installation

openclaw skills install @dustinparsons/drawthings

Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:

npx clawhub install drawthings

What This Skill Does

Connects to DrawThings, a local Stable Diffusion app for Mac, via its Automatic1111-compatible API. Generates images from text prompts with control over steps, samplers, dimensions, seeds, and batch size. Runs entirely on-device using MLX/CoreML acceleration.

Because DrawThings runs locally, generation is private and has no per-image API cost.

When to use it

  • Generate 10 character variations from a single prompt
  • Produce a reproducible image by fixing the seed
  • Batch-generate concept art for a game environment
  • Create a high-res portrait with a quality preset
  • Draft images quickly using 8-step fast preset

Example Workflow

Here's how your AI assistant might use this skill in practice.

INPUT

User asks: generate 5 cyberpunk cat variations at 768x768

AGENT
  1. 1Set DRAWTHINGS_URL to the local DrawThings API address
  2. 2Call generate.py with the prompt, --batch-size 5, --width 768, --height 768
  3. 3Apply DPM++ 2M Karras sampler with 20 steps and cfg-scale 7.5
  4. 4Receive base64-encoded PNG responses from the API
  5. 5Save each image with embedded metadata to the output directory
OUTPUT

5 PNG files named drawthings_output_YYYYMMDD_HHMMSS.png, each containing the prompt, seed, and generation parameters in metadata

Requirements

Accounts, API keys, or tools you or your AI assistant may need to set up while using this skill.

DrawThings app installed on Mac with API server enabled in preferencesDRAWTHINGS_URL environment variable (defaults to http://127.0.0.1:7860)