This LoRA remains relatively uncompleted. The training asset contains only 41 images for more than 5 categories of monsters, so I'm not that sure whether each of them can run smoothly. To ensure stable outcomes, efforts of manually img2img sketch would be somehow needed.

Prompt: aic_monster

aic_worm_wall, worms, worms on body, worm sex, black worms,

aic_monster surrounding,

aic_mole_snake, tentacles, black tentacles, restrained,

aic_puppy, tentacles, black tentacles, (optional: slime_(creature))

aic_giant, monster,

aic_lord_of_the_forest, tentacles, restrained, vines, black vines, black tentacles, (optional: crotch_drilling, tentacle sex)

aic_slime, black slime, slime_(creature),

aic_cow, cow,

(optional: aic_monster rune, rhombus rune for the rhombus shape on monsters)

(Some of the prompts are almost ineffective, donnot count too much on them...)