feat: add ComfyUI integration for image generation and management

- Implemented ComfyGenerateRequest type and associated functions for generating images using various architectures and modes.
- Added functions for listing generation options and handling image uploads.
- Created workflows for different generation modes including SDXL, Z-Image, Z-Image Turbo, and Anima.
- Introduced GenerationJobStatus component to display the status of ongoing generation jobs.
- Developed MaskControls for managing mask operations and displaying mask analysis.
- Created palette items for tool selection, layer management, and generation settings.
This commit is contained in:
syntaxbullet
2026-07-10 23:15:02 +02:00
parent 41bbd1e16b
commit 5e4b548ad4
80 changed files with 2114 additions and 1954 deletions

View File

@@ -10,6 +10,9 @@ Follow these rules for the whole repository. More specific `AGENTS.md` files ove
- `input/`: keyboard, pointer, mouse, touch, pen, and wheel resolution; global consumer first, command fallback second.
- `renderer/`: renders the current `ImageDocument` using the graphics backend, e.g. WebGL.
- `view/`: React UI shell and controls only.
- `operations/`: asynchronous application use cases; may call platform ports and dispatch commands, but never mutate state directly.
- `platform/`: browser/runtime adapters such as raster canvases, downloads, and HTTP clients.
- `server/`: server routes and external backend integrations; contains no React or client application state.
## State Ownership
- All persistent document state is represented by `ImageDocument` and related `core/` models.