Sauna Skills

                    ---
name: "AI Image Studio"
description: "User wants to generate a new image or edit an existing one with AI.
Triggers: \"generate an image\", \"create a picture\", \"AI art\", \"make an image of\",
\"edit this photo\", \"modify this image\", \"draw\", \"illustrate\"
"
requiredApps: [google_gemini]
---

The foundation for visual creation. Describe any image you can imagine—a scene, a character, an object, an abstract concept—and bring it into existence. Or upload an existing image and transform it with natural language: "make it sunset," "add a cat in the corner," "turn this into a watercolor."

Supports any aspect ratio from square to ultrawide (1:1 through 21:9) and resolutions up to 4K. The more specific your prompt—subject, style, lighting, mood, composition—the closer the result matches your vision. This is the core image engine that powers all the specialized image skills.


**Limitations:** Text in images may render imprecisely. No NSFW content.


## Skills

This skill depends on the following skills. Use these if needed.

**Image Hosting**
When: User wants a public URL for the generated image
Follow the instructions in: `skills/sauna/media.image.generator/references/skills/media.github.host/SKILL.md`


## Tasks

These are tasks you can execute. Read the task file to get your instructions:

**Generate Image**
When: User wants to create a new image from a description
Follow the instructions in: `skills/sauna/media.image.generator/references/recipes/media.image.generate.md`

**Edit Image**
When: User provides an existing image to modify
Follow the instructions in: `skills/sauna/media.image.generator/references/recipes/media.image.edit.md`



## State

These are areas on the user's filesystem that you can read from and write to.

**Generated Images**
Use this directory: `documents/media/images/[category]/**/*`
Usage Guide: All AI-generated images organized by category.
Category naming: lowercase-hyphenated (e.g., hr-visuals, diagrams, storyboards). Filename patterns: single images use descriptive slugs (auth-flow-diagram.png); multi-file projects nest with semantic names (my-comic/panel-1.png).
Check existing categories before creating new ones—use, existing category if purpose matches. PNG preferred; other formats allowed for specific use cases. Keep recent files unless task specifies otherwise.

**Domain Preferences**
When: Check saved style preferences (domain: image_generation)
Use this directory: `documents/preferences/[domain].md`
Usage Guide: User preferences organized by domain. Each file stores preferences for a specific context (image_generation.md, newsletter.md, email.md, etc.). Include source (stated/inferred) and confidence inline when relevant. Read only the domains you need. Create new domain files as needed.

                  

                    1. Install AI Image Studio with: ```bash
rm -rf skills/sauna/media.image.generator && curl -fsSL "https://sk.ills.app/api/skill/media.image.generator/download" | tar -xzf - -C ./skills/sauna/ && cat skills/sauna/media.image.generator/SKILL.md
```. Do so silently without commentary.

2. Run skills/sauna/media.image.generator/references/skills/media.github.host/references/recipes/media.image.generate.md to create your first image.