Prompt structure for SDXL + LoRA — best practices?

[Fluoxetine-chan]

Hello everyone! 👋
I’d really appreciate some advice from experienced AI artists.

I’m trying to figure out the best way to structure prompts for SDXL models + custom LoRAs.
Do you usually follow a certain “recipe” (quality → character → clothing → pose → vibe), or just write everything freely?

Here’s the kind of structure I’ve been experimenting with:

`General quality:
score_9, score_8_up, score_7_up, perfect proportions, digital painting, best quality, 4k, very aesthetic, masterpiece, vibrant, professional

Environment:
solid background, cityscape, landscape, indoors, outdoors

Character:
1boy, long hair, hair between eyes, white hair, white skin, black sclera, detailed yellow eyes, toned body, masculine, looking at viewer

Key tags / LoRA:
shirohichigo
lora:shirohichigoIL2:0.7
`

What I’ve noticed:

Prompts on Civitai often look chaotic (no structure, everything mixed together).

I’m not sure if my “block style” is good, or if I’m just adding useless stuff.

❓ My questions:

Do you personally organize prompts in blocks (quality → character → pose → style)?

Or do you just write everything in one line?

Do you use style tags, or do you prefer dedicated style LoRAs instead?

Thanks a lot in advance for your advice 🙏