stability-ai

What This Skill Does

The Stability AI skill is a powerful generative tool for OpenClaw that allows users to create high-quality, professional-grade imagery directly within their workflow. By leveraging advanced models such as SDXL, SD3, and Stable Image Core, this skill acts as a digital artist assistant. Whether you need photorealistic landscapes, stylized digital art, or complex artistic compositions, the skill handles the heavy lifting of API communication, model selection, and image processing. It produces both the resulting image file and a comprehensive JSON metadata manifest, ensuring that every generation is reproducible and well-documented.

Installation

Getting started with the Stability AI skill is straightforward. First, add the skill to your environment using the command: clawhub install openclaw/skills/skills/1999azzar/stability-ai. Once installed, navigate to the skill directory, copy the .env.example file to .env, and populate your STABILITY_API_KEY. If you are using a self-hosted instance or a custom proxy, configure the API_HOST variable. The system handles all dependency management during the initial execution, allowing for a seamless "plug-and-play" experience.

Use Cases

This skill is perfect for creators and developers who require visual assets without leaving the terminal. Use it to generate concept art for game design, create placeholder images for web and application development, or produce unique visuals for marketing materials and social media posts. The ability to use seed values makes it an ideal tool for iterative design, allowing you to refine a prompt until the visual output perfectly matches your creative vision.

Example Prompts

1. "Draw a cyberpunk city at night with neon signs and flying vehicles in 16:9 ratio."
2. "Generate a photographic portrait of an elderly traveler wearing a leather hat, use the cinematic style."
3. "Create an image of an abstract landscape with floating islands, use seed 999 for reproducibility."

Tips & Limitations

To maximize the quality of your outputs, utilize the negative prompt feature to steer the model away from unwanted artifacts like blurriness or low-quality textures. Be mindful of the aspect ratio settings; while 1:1 is the default, choosing a specific ratio like 21:9 can drastically change the composition. Note that this skill consumes API credits per generation depending on the model chosen; refer to the documentation for specific cost breakdowns. Finally, the built-in auto-cleanup feature maintains a limit of 20 images; if you need to keep your work long-term, ensure you move your favorite outputs to a permanent storage directory, as older files will be removed by the script.