I've set up an image generator on perchance, you can check it out here: https://perchance.org/wam-image-generator-1
Attached are some images I've gotten out of it. It generates nudes too but I can't post them here. Feel free to share any you get that you like
I would like some suggestions on what the prompts should be so I can get pictures where the mess is falling down the subject and in her hair. Right now the pictures that come out well are more wet and less messy. Does anyone have any tips?

This is excellent and I look forward to seeing what it comes up with.
I'm an n00b and was wondering how I could edit this to add some of my own prompts? Specifically I want to just alter the clothing.

Sure, there's an edit button in the top-right corner. Then look for the section called dress in the panel on the left. There's a generic perchance tutorial on the top of every page on desktop if you need help.
It does not always listen to my requests for clothing so if you crack it please let me know

Perchance uses Stable Diffusion, the stuff your making is really good from the base version!

If you want to make even better images you want to run it locally you can then use checkpoints and lora from Civitai.com to get images in the style of things that interest you, or failing that you can possibly train your own styles depending on the spec of your pc. Also no worries from filters.

I can't run it on my machine so I have to keep at it with the online generators.

I read that as 'if you crack please let me know'. As it will ignore my terms and randomly erase them that could happen at some point...

Wow nice. You got some really nice results. I learned a few things from your prompts.

I do have some tips for increasing the mess and less wet. I cannot get any results where mess is falling down on the subject though.

I took one random prompt:

"A professional photo of a mid-20s woman dumping goopy orange slime over her body. She is fully clothed in casual attire. The orange slime runs down her face, soaking into her messy hair and staining her clothes."

The first image shows the result I got for this. I like that she is nude, but the prompt should have made her wearing casual clothing. I don't like the water splash and background.

I found a few ways to increase the amount of mess. Sometimes you can get better results if you crease the weights. When I run stable diffusion locally, it supports changing weights. I don't know if online tools support this. You can change weights like "(goopy orange slime over her body:1.4)". You have to tweak the weight 1.4. Sometimes that will increase the mess, and you need to increase or lower 1.4 to get more or less. Adding more will often distort the image when it is higher like 1.5 or 1.6. I tried this on the prompt though, and I could not really get a good result.

I tried to improve this prompt to get a better result. I write prompts for stable diffusion with comma separated terms, so I changed it to be comma separated. I removed the word "soaked" and "dumped" because I think that was causing water splashes. The words "soaked" and "dumped" do seem to cause some good results though. I added a kitchen background. I added a bunch more, and repeated the messy substance a lot. Here is the prompt:

"mid-20s woman, goopy orange slime over her body, fully clothed in casual attire, goopy orange slime down her orange painted face, goopy orange slime on her head oozing down her orange hair, goopy orange slime staining her clothes, kitchen background, soft smile"

The 2nd image is the result. There is much more in her hair, but her face is still clean. The main reason for this is that "mid-20s" has a strong influence on the face being clean and it is first in the prompt. All I did was move it last:

"woman, goopy orange slime over her body, fully clothed in casual attire, goopy orange slime down her orange painted face, goopy orange slime on her head oozing down her orange hair, goopy orange slime staining her clothes, kitchen background, soft smile, mid-20s"

It is a little better. There are ways to continuously change this to make it better. It is very hard for me to get a good messy face result though. I generally avoid messy faces in the images I make just because it is easier, and I find it attractive.

I did not use any specially trained model or lora for this. It is just a generic realism one. The thing is that you might copy this prompt and not get good results due to small differences in the model. Hopefully this helps though.

Thank you for the advice! I like the images you were able to get, but it doesn't seem to be working as well for me. I need to learn more about the generator it's using behind the scenes. If I can set weights that would be huge.
Here is the generator if anyone wants to mess with it: https://perchance.org/wam-image-generator-2