Save 4 hours! We auto-setup your workflow! Free!

Drop your workflow.json — we handle every dependency, custom node, and model. Just open the link and run.

Auto-Setup Workflow Json (Free) Now!
ComfyUI > Nodes > ComfyUI-MossPanelDataset > Choose a composition (preview images)

ComfyUI Node: Choose a composition (preview images)

Class Name

MossCompositionTemplate

Category
Manga panel dataset
Author
sepiablue-ai (Account age: 467days)
Extension
ComfyUI-MossPanelDataset
Latest Updated
2026-11-05
Github Stars
0.04K

How to Install ComfyUI-MossPanelDataset

Install this extension via the ComfyUI Manager by searching for ComfyUI-MossPanelDataset
  • 1. Click the Manager button in the main menu
  • 2. Select Custom Nodes Manager button
  • 3. Enter ComfyUI-MossPanelDataset in the search bar
After installation, click the Restart button to restart ComfyUI. Then, manually refresh your browser to clear the cache and access the updated list of nodes.

Visit ComfyUI Online for ready-to-use ComfyUI environment

  • Free trial available
  • 16GB VRAM to 80GB VRAM GPU machines
  • 400+ preloaded models/nodes
  • Freedom to upload custom models/nodes
  • 200+ ready-to-run workflows
  • 100% private workspace with up to 200GB storage
  • Dedicated Support

Run ComfyUI Online

Choose a composition (preview images) Description

Loads saved character composition templates to guide consistent staging in manga panels.

Choose a composition (preview images):

MossCompositionTemplate lets you reuse prepared character compositions when building manga panels in the ComfyUI-MossPanelDataset workflow. Instead of describing and arranging every character placement from scratch, you select an existing composition whose stored arrangement and English description can guide subsequent panel creation. The node loads a saved template rather than generating a new composition or rendering an image, making it useful for maintaining consistent staging across panels and exploring different arrangements with less manual setup. The supplied dataset is tested to contain 120 compositions, divided evenly into 40 one-character, 40 two-character, and 40 three-character arrangements. Each template has associated reference-image and generation information, although this node itself does not display or return a reference image.

Choose a composition (preview images) Input Parameters:

構図テンプレート

Required selection field of type COMBO[STRING] that determines which saved composition the node loads. Its options come from the composition catalog: all entries’ name values are listed first, followed by all entries’ name_en values, allowing you to select templates using their catalog names or English names. The exact available names depend on the installed catalog and are not hard-coded in the node. Choosing a different entry changes both the loaded composition data and the English composition prompt returned with it. No explicit default is declared in the source; the first catalog option is the initial available choice in the usual dropdown behavior. Minimum and maximum values do not apply because this is a selection rather than a numeric control. If the catalog contains no entries, the dropdown instead shows 生成済みの構図がありません / No compositions; this is an empty-catalog notice, not a usable composition template.

Choose a composition (preview images) Output Parameters:

Composition for panel

An output of type MOSS_COMPOSITION containing the complete composition record loaded for your selection. This is structured template data rather than an image or a plain-text description. The supplied context shows that composition records include an English prompt and character-placement boxes, and that the dataset also stores identifying, reference-image, and generation-related information. The placement boxes describe the character regions that compatible downstream panel-building nodes can use to arrange the panel’s characters. Connect this output to a compatible composition input when you want the saved arrangement to participate in panel construction, rather than using only its descriptive wording. The node returns the loaded record directly without editing its contents.

Composition English prompt

An output of type STRING containing the selected composition record’s prompt value. This is the saved English description of the composition, returned directly without translation, rewriting, or additional text generation. It lets you inspect or reuse the template’s descriptive wording independently of the structured composition data. The distinction matters because a text connection carries the description only; it does not carry the character-placement boxes or the rest of the template record. Use this output wherever your workflow needs the English composition wording, and use the MOSS_COMPOSITION output when a compatible panel node also needs the saved placement information.

Choose a composition (preview images) Usage Tips:

  • Choose a template with the same number of character placements as the panel you are building. The supplied dataset contains one-, two-, and three-character compositions, and downstream tests demonstrate that a mismatch between a panel’s characters and a template’s boxes can raise a validation error.
  • Use the English entries in the dropdown if they are easier for you to identify. The node lists both the catalog’s name entries and its name_en entries; these are alternative ways to select saved templates, not separate language-generation settings.
  • Connect Composition for panel to a compatible panel-building composition input when you want to reuse the saved character arrangement. Connecting only Composition English prompt provides wording but does not transfer the structured placement boxes.
  • Inspect Composition English prompt before using the template in a larger workflow. This helps you check whether the stored composition description suits the scene you intend to create without assuming that the node will adapt or rewrite it automatically.
  • Use the dataset’s reference images, where accessible, to compare compositions visually. The supplied dataset tests verify 1024 × 1024 WebP references, but this node has no image output, so viewing those references requires a separate image viewer or an appropriate part of the extension interface.
  • If you update the composition catalog, refresh the node’s available selections before choosing newly added templates. The node tracks the catalog file’s modification time for change detection, but its dropdown options are obtained when the input definition is read.

Choose a composition (preview images) Common Errors and Solutions:

生成済みの構図がありません / No compositions

  • Explanation: This exact text is the dropdown’s fallback when the catalog returns no compositions. It is a status notice rather than an exception message, and selecting it does not provide a real template to load.
  • Solution: Check that the extension’s composition catalog and dataset files are installed and available. Restore missing dataset files if necessary, then refresh or restart ComfyUI so the dropdown can be populated again.

Selected composition cannot be loaded

  • Explanation: This is a descriptive troubleshooting label, not an exact exception string shown in the supplied source. The node passes your selected name directly to the composition loader, so a stale selection or an unavailable catalog entry may prevent loading. The loader’s exact error wording is not included in the provided context.
  • Solution: Refresh the available options and select an entry currently present in the dropdown. If the problem persists, check that the catalog and composition data belong to the same complete extension installation.

Character count does not match the composition

  • Explanation: This is a downstream validation issue rather than an explicitly defined error message in MossCompositionTemplate. The supplied tests confirm that a panel can raise a ValueError when its characters do not match the template’s placement boxes; for example, a panel expecting two characters cannot use a template containing only one character box.
  • Solution: Select a composition with the same number of character boxes as the characters assigned to your panel, or adjust the panel’s character configuration to match the chosen template.

Invalid composition boxes

  • Explanation: The supplied tests cover invalid-box validation in downstream panel processing, but do not provide the exact exception text. Damaged or manually edited placement data may load as a template yet fail when a later node tries to use its character regions.
  • Solution: Try an unmodified template from the supplied dataset. If that succeeds, restore the affected template from the original dataset rather than guessing replacement coordinates or box structures.

KeyError: 'prompt'

  • Explanation: The node directly reads the loaded record’s prompt field. If a damaged or incompatible record lacks this field, Python can raise this error because the English description required by the node is missing.
  • Solution: Restore the affected composition record or reinstall the matching dataset. If you maintain custom composition data, ensure each loaded record includes the expected prompt field.

FileNotFoundError or PermissionError during catalog access

  • Explanation: These are possible file-access exceptions, not custom messages defined by this node. The node checks the catalog’s modification time, and loading a selection also depends on accessible composition data. Missing files or insufficient access permissions can interrupt these operations; the full exception text depends on the failing operation.
  • Solution: Check the file path reported in the ComfyUI error log, restore missing files, and make sure the account running ComfyUI can read the extension’s catalog and dataset. If files were moved or replaced while ComfyUI was running, restart it after correcting the installation.

Choose a composition (preview images) Related Nodes

Go back to the extension to check out more related nodes.
ComfyUI-MossPanelDataset
RunComfy
Copyright 2025 RunComfy. All Rights Reserved.

RunComfy is the premier ComfyUI platform, offering ComfyUI online environment and services, along with ComfyUI workflows featuring stunning visuals. RunComfy also provides AI Models, enabling artists to harness the latest AI tools to create incredible art.

Choose a composition (preview images)