Give this LoRA a try when looking to squeeze a slightly better image out of your prompts. What does it do? In general, it tends to add more natural posing and detail to a shot. It works well with my FULLY_REAL_XL checkpoint. (Note that it is partially baked in to V5 of my checkpoint so use a weaker weight when pairing the two.) Images posted show with and without the LoRA applied.
No trigger word needed. Weight of 0.4 to 1.0 depending on checkpoint model.