MOSS Layout YDR

MOSS Layout YDR

A visual layout editor for MOSS-Image in ComfyUI. Draw regions on the image, describe what belongs in each one, and send the finished layout to MOSS.

<p align="center"><img src="examples/assets/moss-layout-example.png" alt="Example MOSS-Image result created with a layout prompt" width="100%"></p>

Works with both ComfyUI's classic (Legacy) nodes and Node 2.0. The image above is only a demonstration; the example workflow does not need an image file.

What it does

• Draw and edit boxes for characters, objects, and text.
• Add descriptions, optional visible text, and color palettes.
• Choose the canvas size in the node or connect width, height, and megapixels from other nodes.
• Import a MOSS layout from JSON or text.
• Save the whole editor state with the workflow, including boxes and fields.
• Use the latest generated image as a preview, or connect an optional background image.

Install

Download this repository and copy ComfyUI-MOSS-Layout-YDR into ComfyUI/custom_nodes/, then restart ComfyUI. Find MOSS Layout YDR under MOSS/Layout.

Connect it

For a regular MOSS-Image workflow, connect prompt to the positive prompt input and width and height to the image size inputs.

You can also connect the outputs to ComfyUI's built-in MOSS-Image Layout Prompt node:

• moss_elements → element
• summary_caption → prompt
• width and height → the matching size inputs
• character_boxes → character_boxes

Example workflow

• moss-layout-ydr-moss-workflow.json — the workflow supplied for this project, with its saved MOSS Layout YDR state.

The workflow does not include model checkpoints. Install the models required by your ComfyUI MOSS-Image setup.

Credit and license

The visual editor builds on patterns from Deno2026's ComfyUI custom nodes. Thank you, Deno2026, for sharing this work as open source. This project is licensed under GPL-3.0-only; see LICENSE and THIRD_PARTY_NOTICES.md.