Sorry for the slow reply. My work at the office and my college activities take up a lot of my time.
Here's the example prompt for Isabella:
Positive
((best quality)), ((masterpiece)), ((anime)), (detailed), portrait, close up, young female, blue eyes, (wearing a serafuku),(serafuku), ((pink hair)), long hair, highly detailed skin, looking at viewer, cyberpunk art, stunning gradient colors, no watermark signature, detailed background, town, alley, dark alley, closed mouth, insanely detailed, , ((masterpiece)), absurdres, HDR, <lora:Muscle2:1.3>, (muscular:1.3), muscle, big breasts,
negative
wires, ear rings, dirty face (deformed iris, deformed pupils, bad eyes, semi-realistic:1.4) (bad-image-v2-39000, bad_prompt_version2, bad-hands-5:1.5, EasyNegative, ng_deepnegative_v1_75t, bad-artist-anime:0.7),(worst quality, low quality:1.3), (depth of field, blurry:1.2), (greyscale, monochrome:1.1), nose, cropped, lowres, text, jpeg artifacts, signature, watermark, username, blurry, artist name, trademark, watermark, title, (tan, loli, petite, child, infant, toddlers, chibi, sd character:1.1), multiple view, Reference sheet, long neck, nsfw
Note: The results of the art generated in my game do not purely come from stable diffusion, I always use controlnet to control the pose of the character in Stable Diffusion. I also fixed the error of the stable diffusion image generation using Photoshop. Then to produce satisfactory results I also use LORA and Models (checkpoint), which I'm very sorry, I can't reveal, because it's part of my work secrets. So using the positive or negative prompts that I have provided above may not necessarily produce the satisfactory results that you want. Thank you for asking.