Generate Professional AI Art and Images
A skill for generating and editing images via the Stability AI API - text-to-image, inpainting, upscale, and 15 art styles.
Why it matters
Leverage Stability AI's advanced models (SD3.5, Ultra, Core) to generate high-quality artistic and photorealistic images from text prompts, with capabilities for image-to-image transformations, inpainting, upscaling, and background removal.
Outcomes
What it gets done
Generate images from text prompts using various Stability AI models.
Perform image editing tasks like inpainting, upscaling, and background removal.
Transform existing images into new artistic styles using image-to-image generation.
Create concept art, illustrations, and photorealistic visuals for diverse applications.
Install
Add it to your toolbox
Run in your project directory:
curl -fsSL https://spark.entire.vc/get/ag-stability-ai | bash Overview
Stability AI - Professional Image Generator
A skill for generating and editing images via the Stability AI API - text-to-image, img2img, inpainting, upscale, background removal, and search-and-replace - with 15 art style presets. Use it for digital art, illustration, or photorealistic image generation and editing; prefer a Gemini-based image skill instead for humanized social-media or casual phone-camera photos.
What it does
This skill generates and edits artistic and photorealistic images via the Stability AI API (SD3.5, Ultra, Core models), free under the Community License with no limit for personal/small-business use. It supports ten operation modes via CLI flags: generate (text-to-image on SD3.5), ultra (premium text-to-image), core (fast text-to-image), img2img (image plus text into a new image), upscale (conservative resolution increase) and upscale-creative (detail-adding upscale), remove-bg (transparent-background removal), inpaint (mask-based partial editing), search-replace (swap an object by description), and erase (remove a masked element). It defines eight aspect-ratio presets with aliases (square/1:1 for Instagram feed, portrait/2:3, landscape/3:2, photo/4:5 for optimized IG feed, wide/16:9 for cinema/YouTube, ultrawide/21:9, stories/9:16 for TikTok/IG Stories, phone/9:21 for wallpapers) and 15 style presets that auto-append qualifiers to the prompt (photorealistic, anime, digital-art, oil-painting, watercolor, pixel-art, 3d-render, concept-art, comic, minimalist, fantasy, sci-fi, sketch, pop-art, noir). It provides explicit guidance on when to use this skill versus a complementary Gemini-based image skill: use Stability AI for digital art/illustration/concept art, cinematic 8K photorealism, posters/wallpapers/book covers/game assets, and any inpainting/upscale/background-removal/search-replace/erase editing operation; use the Gemini-based alternative for humanized social-media photos, casual phone-camera realism, and educational visuals. Setup requires a free platform.stability.ai account, an API key placed in .env as STABILITY_API_KEY, and installing script requirements. Output images are saved to data/outputs/ with a {mode}_{style}_{timestamp}_{index}.png naming convention, alongside a .meta.json recording the original and final prompt, model, aspect ratio, seed, generation time, and file size. It integrates with other skills - generating art to publish via an Instagram skill, or sending generated images through a Telegram bot. Rate limits and safety measures include the Community License's 150 requests/10 seconds, a configurable daily image cap (default enforced via SAFETY_MAX_IMAGES_PER_DAY), automatic exponential-backoff retry on HTTP 429, and fallback to backup API keys if the primary fails.
When to use - and when NOT to
Use it when generating digital art, illustrations, concept art, cinematic photorealistic renders, posters/wallpapers/covers/game assets, or performing image editing operations (inpainting, upscaling, background removal, search-and-replace, erasing). Prefer the complementary Gemini-based image skill instead for humanized social-media photos, casual phone-camera style, or educational-material visuals - this skill is not the right choice for those use cases.
Inputs and outputs
Inputs: a text prompt, operation mode, and (for editing modes) a source image and optional mask; aspect ratio and style preset are configurable.
Outputs: generated or edited PNG images saved with a structured filename, plus a metadata JSON recording the prompt, model, and generation parameters.
python scripts/generate.py --prompt "cinematic portrait, dramatic lighting" --mode ultra --aspect-ratio 16:9
python scripts/generate.py --mode inpaint --image jardim.jpg --mask mascara.png --prompt "red roses"
Who it's for
Creators and developers who need art, illustration, or photorealistic image generation and editing (upscale, background removal, inpainting) via the Stability AI API, as a complement to more casual/humanized photo-generation tools.
FAQ
Common questions
Discussion
Questions & comments · 0
Sign In Sign in to leave a comment.