For LoRAs that I make I always add examples where I don't use any other LoRA/TI/Emb. You should be able to get good result just following the prompting scheme: [trigger words], [pose], [location], [quality tags]. If You have problem with the eyes try to use ADetailer or different checkpoint - maybe the way you are writing the prompt have some problems with model that You are using.

@TactileMist First of all, try to change checkpoint to some of mine. Here I can see that You are using base Pony checkpoint, which was never good for me either. Also, You don't have to use my nickname in the prompt, it is more like signature for LoRA :). Also work on Your negatives, use more stuff like "bad quality, score_1, score_2, score_3, score_6, score_5, score_4, low quality, censored, deformed, bad hands, old, greyscale, jpeg artifacts, blurry, messy drawing, amateur drawing, bad anatomy" etc. and other negatives embedding, only one pack might be not enough. Same goes for positives, if it is Your full prompt - that is for sure too basic. Also You can join my Discord to talk to the people and have discussions about stuff :)