> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.loreos.app/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.loreos.app/_mcp/server.

# Visual system

LoreOS supports characters that can appear visually, not only in text. The visual system
does not treat images as one-off generations: it keeps image generation grounded in
identity, reusable assets, readiness checks, and probes so a character can appear
consistently over time.

## Core concepts

### Identity image

An image that anchors the character's face and visual identity.

### Visual assets

Reference assets such as outfits, accessories, hairstyles, inventory items, and scene
elements. Assets can be uploaded and managed per character.

### Visual inventory

A structured list of assets and image-relevant details the runtime can use when planning
or generating images.

### Image readiness

Readiness checks tell you whether a character has enough visual grounding to generate
consistent images.

### Image probe

An image probe is a controlled test run. Use it to see whether the visual setup behaves
as expected before exposing image generation to users.

## Recommended setup

1. Create the character.
2. Upload a strong identity image.
3. Add representative wardrobe and accessory assets.
4. Check visual readiness.
5. Run an image probe.
6. Use the generated result as launch evidence.

Start with the **Visual assets** endpoints in the API Reference.