General tags:
cskrstg: general tags, included in every images1skst: this is for a skinsuit that is clean preventing in imagewearing skst: for person who is wearing the skinsuitexposed skst back: when you can see the gap of skinsuit after worn (untested)skst pull: anchor for skinsuit pull related tags unworn head: only when the skinsuit's head is on the wearer's headdetched head: when the skinsuit's head is not connect to skinsuit and not on head, is holding on hand or just somewhere else; (this will not be very good if you have too many figures in the image)holding skst： holding a skinsuit (very unstable, and might not work; don't treat it as a valid tag)Modifiers:
(for wearing skst):
unworn upper body, covered breasts: cover-level is around and cover the breats and upper elbow (This is very broken, will improve in the next iteration)unworn upper body: just like half covered, around the waist levelunworn lower body: only cover the part below the waist (missing data, so it is weak)(for exposed skst back):
open zipper: there is a zipper around the hollow and is openedclosed zipper: there is a zipper around the hollow and is closedno zipper: no zipper on back(for exposed skst back):
face pull: pull the face part of skinsuitneck pull: pull the neck part of skinsuitOR: <body part> pull, this seems to work with <cheek> and <mouth>, so help yourself
There are some tags are used in trained but untested because of lack of data or I forgot to actually include such tags in caption files xd, check meta if you want to know.
Before Anima version, now I am planning a illustrious iteration to improve style leakages and balancing concepts.
Description
Blah blah blah