---

name: ui-workspaces description: "Guides Krita AI Diffusion Qt model, workspace, settings, job queue, connection state, live preview, upscaling, animation, and UI routing tasks." disable-model-invocation: true metadata: disco-role: operating license: GPL 3.0

ui-workspaces

Use this sub-skill when the task is about plugin UI/model state rather than the backend request dataclasses alone.

Trigger examples:

• Understanding DocumentModel, Workspace, QueueMode, ProgressKind, ErrorKind, JobKind, or JobState.
• Debugging why Generate, Live, Upscale, Animation, or Graph workspace enqueues a job or reports a validation/error state.
• Working with observable properties, settings persistence, connection state, or job queue behavior.
• Running headless Qt model tests or static enum inspection.
• Mapping UI controls to WorkflowInput fields.

Safe entry points

python sub-skills/ui-workspaces/scripts/inspect_workspace_enums.py --static-only
QT_QPA_PLATFORM=offscreen python sub-skills/ui-workspaces/scripts/inspect_workspace_enums.py --import-live

Static mode is safest and performs no imports. Live import mode requires a package/test environment where ai_diffusion imports successfully.

References

• references/workspace-model-reference.md: workspace/model/job state and observable property map.
• references/ui-routing-and-settings.md: how UI settings route generation, server mode, apply behavior, persistence, and test setup.
• references/troubleshooting.md: Qt/headless, job queue, connection, and validation recovery.
• scripts/inspect_workspace_enums.py: static/live enum and model-property inspector.

Boundaries

• For WorkflowInput payload internals after a workspace enqueues work, route to inference-workflows.
• For server mode/backend/URL and connection resources, route to server-resources.
• For custom Graph workspace node metadata, route to custom-graphs.
• For document, layer, image, mask, prompt, style, and persistence details, route to document-image-state.