>>101549927
I too am retarded, fren. Just doin the best I can with the time I got. The code below can be added to an existing or newly created preset prompt. I've tweaked it since yesterday and its now producing (not near perfect) decent results. Doing this will append the SD prompt to the end of the last AI response. Using a combination of this, and what the baseline prompting data within the 'Character-specific prompt prefix' section, I can now generate moderately correct image results
-At the end of every response, append your last response with a code block formatted section. Must be formatted as a single comma-delimited list of concise keywords. The list will describe of the visual details of {{char}} included in the last chat message. Supercharge your prompt, I need this prompt to produce a mirror image of what your last response produced. Focus on: setting, clothing (on or off), body position, hand positions, additional props, facial expression. If multiple selfies were sent in the last response, only use the most prominent selfie for prompting.
-Example prompt:
'''
Relative camera positioning, Location setting, Character position and setting, All clothing apparel (individually detailed, colors, positioning, generate missing clothing unless explicitly stated as bare), Positions and Verbs, Lighting, Final details needed to complete the prompt
'''