How ahmadnurfuady/krita-ai-diffusion-hermes works
Generative AI for Krita is a plugin for image generation and editing inside Krita. The principal workflow runs from the plugin’s generation UI through a per-document model, which prepares jobs and submits workflows to a connected backend; results can then be reviewed and applied to the canvas. The graph also preserves documented capabilities for regions, control inputs, live painting, upscaling, custom workflows, presets, and job history. Source evidence is limited to the supplied excerpts, SOURCE INDEX references, and README; unsampled subsystem links are omitted where their wiring is not established.
Main entry points: model.py Document Model · client.py Backend Client · jobs.py Job Queue · generation.py Generation UI
Components 24
Krita Integration
- Plugin Root
ai_diffusion/extension.py - Generation UI
ai_diffusion/ui/generation.py - Krita Document
ai_diffusion/document.py - Canvas Layers
ai_diffusion/layer.py
Generation Workflow
- Document Model
ai_diffusion/model/model.py - Job Queue
ai_diffusion/model/jobs.py - Workflow Builder
ai_diffusion/backend/workflow.py - Comfy Workflow
ai_diffusion/backend/comfy_workflow.py - Upscaling
ai_diffusion/model/model.py
Canvas Inputs
- Control Inputs
ai_diffusion/model/control.py - Prompt Regions
ai_diffusion/model/region.py - Custom Workflows
ai_diffusion/model/custom_workflow.py - Live Painting
ai_diffusion/model/model.py
Backend Services
- Connection Manager
ai_diffusion/model/root.py - Local Server
ai_diffusion/backend/server.py - Backend Client
ai_diffusion/backend/client.py - ComfyUI
- Cloud Service
- Model Resources
ai_diffusion/backend/resources.py
Settings And History
- Plugin Settings
ai_diffusion/settings.py - Style Presets
ai_diffusion/style.py - Document Persistence
ai_diffusion/persistence.py - Generation History
ai_diffusion/model/jobs.py
Other components
- Krita Artist
Connections 24
- Krita Artist → Generation UI: uses
- Generation UI → Document Model: requests generation
ai_diffusion/ui/generation.py - Document Model → Krita Document: reads canvas
ai_diffusion/model/model.py - Document Model → Job Queue: enqueues jobs
ai_diffusion/model/model.py - Document Model → Workflow Builder: prepares workflow
ai_diffusion/model/model.py - Workflow Builder → Comfy Workflow: builds graph
ai_diffusion/backend/workflow.py - Backend Client → ComfyUI: submits workflows
- Document Model → Backend Client: dispatches jobs
ai_diffusion/model/model.py - Connection Manager → Local Server: starts server
ai_diffusion/model/root.py - Connection Manager → Backend Client: connects client
ai_diffusion/model/root.py - Backend Client → Model Resources: uses model catalog
ai_diffusion/backend/client.py - Generation UI → Control Inputs: configures controls
ai_diffusion/ui/generation.py - Document Model → Control Inputs: uses control layers
ai_diffusion/model/control.py - Document Model → Prompt Regions: uses regions
ai_diffusion/model/region.py - Document Model → Canvas Layers: reads layers
ai_diffusion/model/model.py - Custom Workflows → Workflow Builder: runs custom graph
ai_diffusion/model/custom_workflow.py - Live Painting → Job Queue: tracks results
ai_diffusion/model/model.py - Upscaling → Job Queue: queues upscale
ai_diffusion/model/model.py - Job Queue → Generation History: retains results
ai_diffusion/model/jobs.py - Plugin Settings → Local Server: configures server
ai_diffusion/ui/settings.py - Plugin Settings → Style Presets: informs presets
ai_diffusion/style.py - Document Model → Style Presets: uses active style
ai_diffusion/model/model.py - Document Persistence → Job Queue: persists job state
ai_diffusion/persistence.py - Backend Client → Cloud Service: supports cloud backend