A few suggestions

First of all I want to express my deep gratitude for this plugin. **Thank you very much**
Here are some feature suggestions based on my experience with this and other plugins.

I am a CG artist and have been using SD extensively since last year.
For a long time I worked in Krita with the auto-sd-paint-ext plugin, since there was no decent alternative for PS before.
Now the Krita plugin is not being developed, it does not have controlnet support, and so on.
But some of the solutions in auto-sd-paint-ext are quite convenient and allow you to work faster in some situations.

_**Suggestion 1:**_
**Add the ability to generate an image in a layer without additional steps and attachments to folders.**
That is, when you click on the generate button, the folder is not created and confirmation is not required.
The generated image immediately appears as a separate layer.
This mode is necessary to speed up the work. Since in most cases, creating detailed art requires a lot of img2img generations to refine individual elements.

**An example of a desired pipeline:**
-Select the desired part of the image
-Generate img2img
-Erasing a part of the inappropriate image (for example, along the edges, for better blending)
-Overpaint in the same layer (as needed)
-Repeat

**Current pipeline:**
-Select the desired part of the image
-Generate img2img
-Confirm by clicking on the green checkmark
-Create a mask for the folder with the image to erase the excess
-Erasing a part of the inappropriate image
-Create a new layer above for overpaint (or switch from mask to layer)
-Overpaint
-Repeat

In the current pipeline, extra steps slow things down a lot.
The image refinement stage uses low CFG, so there is no need to choose from different options.
The priority is fast generation with manual refinement, and so on in a circle.
**I propose to add an option to the settings that disables the creation of a folder and the need for confirmation.**

_**Suggestion 2:**_
**Adding an additional generation size mode: base_size - max_size**
Similar to the system as in the auto-sd-paint-ext plugin in Krita.
Below is a description of how it works:
_"The selection will be resized such that the shorter dimension is base_size. However, if the aforementioned resize causes the longer dimension to exceed max_size, the shorter dimension will be resized to less than base_size. Setting base_size and max_size higher can be used to generate higher resolution images."_
To generate the original image, it's best to stick to a size around 512 (to avoid tiling),
but when reworking with img2img with low CFG, it's better to generate images around 1024, even if a smaller area is selected.
the base_size - max_size mode allows you to set the size once, and generate images not lower than the specified size for any selection, while not being afraid that the requested image will be too large to get "out of memory"
**I propose to add an additional mode base_size - max_size, or replace the ignore mode with it**

_An example of generation sizes for different selection options (approximately)._

_**Suggestion 3:**_
Add **Clip skip** option. (important for working with anime and styling)

_**Suggestion 4:**_
Add **SD ​​VAE** selection option. (important for working with anime and other)

_**Suggestion 5:**_
Make a sampler selection in the pop-up list (the current option takes up a lot of space when expanded; when collapsed, you can forget which sampler is used).

_**Suggestion 6:**_
The ability to increase the field prompt input (a large prompt may not fit)

Please let me know if you need clarification or an example.
Thanks in advance for your reply.