I created this simple workflow to have auto-prompt along with normal Lora loader as I have been facing severe slowness with xlabs lora loader version. This works great.

Update 3.0: Had to make this little bit messy to make it run faster. KSampler was using negative prompts and slowing down generations. Metadata was not getting stored properly. Now the issues are fixed!
Update 3.2: Combined the independent nodes and organized the workflow to make it look simple without spaghetti lines.
The Lora I used will be uploaded in few days under Indian Girl Faces - Beta 1 | Flux LoRA | Civitai, please follow me to get updated when the Lora drops.
Sorry to do this - I need buzz to train my next Lora, so requesting a buzz donation to support my work!

Heartfelt thanks for the ones who have tipped! :) You made my day!