civitai-ai-art

What This Skill Does

The civitai-ai-art skill provides a direct interface for OpenClaw agents to interact with CivitAI's powerful image generation ecosystem. By leveraging the official CivitAI JavaScript SDK, this skill allows users to generate high-quality AI artwork, anime illustrations, and custom character portraits directly from their terminal or automated workflows. It supports advanced configuration including model selection (URNs), LoRA network injection for style control, and fine-tuning parameters such as CFG scale, sampling steps, and clip-skip layers.

Installation

To begin, ensure you have a Node.js 18+ environment ready. Install the skill using the OpenClaw command: clawhub install openclaw/skills/skills/shinnpuru/civitai-api-art. Once installed, ensure your environment variable CIVITAI_API_TOKEN is set with your valid CivitAI API credentials. You will also need to have the civitai npm package installed in your project directory via npm install civitai to enable the underlying SDK functionality.

Use Cases

This skill is perfect for artists, content creators, and developers. Use it to rapidly prototype visual assets for game development, generate consistent character art for storytelling, or produce high-fidelity illustrations by swapping different checkpoints and LoRAs. It is particularly effective for users who require high-resolution, specific stylistic outputs that default web-based generators might struggle to replicate consistently.

Example Prompts

1. "Generate a high-quality portrait of a magical girl with silver hair and purple eyes using the default SDXL checkpoint with a resolution of 1024x1536."
2. "Create an anime-style illustration of a maid with red hair and blue eyes, and apply the LoRA from URN urn:air:sdxl:lora:civitai:162141@182559 with a strength of 0.8."
3. "Generate a fantasy landscape with floating islands and waterfalls, using the checkpoint URN urn:air:sdxl:checkpoint:civitai:101055@128641."

Tips & Limitations

To achieve the best results, always define a strong negative prompt to avoid common AI artifacts like 'bad hands' or 'extra digits'. Note that image generation is resource-intensive; ensure your system handles the output files appropriately. When using multiple LoRAs, be mindful of their total combined weight to prevent image corruption. Finally, always verify that your provided Model URNs are accurate, as incorrect identifiers will cause the generation process to fail.