---

name: custom-graphs description: "Guides Krita AI Diffusion Graph workspace custom ComfyUI workflows, ETN Krita placeholders, parameter metadata, outputs, live mode, and animation tasks." disable-model-invocation: true metadata: disco-role: operating license: GPL 3.0

custom-graphs

Use this sub-skill when the task is about importing, validating, parameterizing, or running custom ComfyUI workflows through the Krita AI Diffusion Graph workspace.

Trigger examples:

• A user has a ComfyUI workflow.json and wants to know which Krita UI controls will be generated.
• Debugging ETN placeholder nodes such as ETN_KritaCanvas, ETN_KritaOutput, ETN_Parameter, ETN_KritaImageLayer, ETN_KritaMaskLayer, ETN_KritaStyle, or ETN_KritaStyleAndPrompt.
• Explaining workflow_parameters, CustomParam, ParamKind, grouped order, custom workflow save/import/overwrite behavior, or document-embedded graphs.
• Diagnosing custom Graph validation warnings, live custom generation, or animation mode.

Safe entry point

python sub-skills/custom-graphs/scripts/inspect_custom_workflow.py workflow.json

The helper parses a workflow JSON file statically and does not connect to ComfyUI, load models, launch Krita, or run generation.

References

• references/custom-graph-reference.md: placeholder node and parameter metadata contract.
• references/workflow-expansion.md: how the Graph workspace turns placeholders/params into inputs, live output, and animation work.
• references/troubleshooting.md: custom graph validation and import/output failures.
• scripts/inspect_custom_workflow.py: static custom workflow inspector.

Boundaries

• For the final WorkflowInput(kind=custom) payload and image request fields, route to inference-workflows.
• For server-side availability of ETN custom nodes and model resources, route to server-resources.
• For document/layer/mask/image semantics used by placeholders, route to document-image-state.