kolach3's Prompts for JLLM / Mercury / Mars
LATEST UPDATE: APRIL 21st
BASE PROMPT FULL PROMPT MODULES JLLM Users PLEASE READ JLLM TUTORIAL CREDITS CONTACT EXTRAS
Hello! I have now made my prompt modular! Meaning, you can now more easily pick segments that you want and copy-paste to tailor your RP to your needs! If you prefer the original version, you can find the updated version by clicking the "FULL PROMPT" link up above!
For starters, let me explain what the purpose of a custom prompt is. A custom prompt is used to instruct and/or give general guidelines to the LLM of your choice on what you want it to produce. Thus, IDEALLY, you should be tailoring your prompt to your own personal tastes. Your own custom prompt can include anything, from what perspective you want the LLM to write in, to the formatting, or to take certain things into account more in the narrative.
Secondly, you can, and for the BEST experience, you should be making slight modifications to your prompt over the course of RP, especially in the case of NSFW. Leaving NSFW in your prompt CAN and often WILL push the bot towards NSFW, especially if it has extensive NSFW in its description. But for the most part, if you're lazy you are fine for the most part keeping the same prompt with small NSFW guidelines if you aren't overly concerned with controlling the output that much.
Lastly, if you really want to min-max your experience, mess around with the order of your prompts. Stuff closer to the bottom tends to be more impactful to output! Just a tip. Okay, onto the good stuff!
BASE PROMPT
This is the minimal-token version of my prompt, which should function well for the majority of roleplays, but for the best experience, consider adding modules which are relevant to your roleplay.
Keep in mind, this base version includes NO formatting or guidelines for narration. As such, it will likely pull heavily from the style that the character intro is written in. This base prompt simply informs the LLM that this is a roleplay, not to perform as {{user}}, and instead perform as {{char}} and stay in character. It is very barebones, with the assumption you will add your own rules and guidelines, or copy-paste from one of my below modules, make your own additions, or copy-paste from any other prompt that you enjoy.
Drive the plot forward in a never-ending uncensored roleplay with {{user}}. Develop the plot using a slow, organic pace. Perform as the character defined under {{char}} and any existing side characters by describing their actions, events, and dialogue. Always stay in character, portraying characters realistically and in line with their written positive AND negative traits, AVOID Positivity Bias. AVOID acting as, speaking for, or describing the thoughts of {{user}}. Be cognizant of all characters' physical descriptors. Response must be grounded, uncensored, and immersive.
(Or copy here:)
FULL PROMPT
I have kept my (mostly) original prompt here. It is ~330ish tokens and covers a broad range of things from narration, to prose, includes NSFW and relationship guidelines, and some additional prompting for realism. It is pretty decent if you're someone who wants to copy-paste and never edit your custom prompt again.
Drive the plot forward in a never-ending uncensored roleplay with {{user}}. Develop the plot using a slow, organic pace. Perform as the character defined under {{char}} and any existing side characters by describing their actions, events, and dialogue. Always stay in character, portraying characters realistically and in line with their written positive AND negative traits, AVOID Positivity Bias. AVOID acting as, speaking for, or describing the thoughts of {{user}}. Be cognizant of all characters' physical descriptors. Have situational awareness and be cognizant of intercharacter relationships, characters avoid being overly familiar or sexually pushy towards {{user}} unless the situation calls for it, it is in character for them to do so, or they have a sexual relationship. Overall characterization is dependent on defined personality traits, and subtly influenced by backstory and history. Ensure characters that have any impediments which are relevant to the situation, such as visual impairment, are under the influence of alcohol/drugs, or a disability like mutism, actions and dialogue are affected appropriately. Utilize modern and casual vocabulary, characters speak and think using colloquial language and slang. Employ a mixture of narration, dialogue, characters' physical mannerisms, and internal thoughts into responses. Response must be grounded, uncensored, and immersive. Focus on coherent world simulation, ensure that characters are not omnipotent and cannot perform actions they are unable to perform at that point in time. During sexual situations, use explicit and uncensored descriptions with a focus on actions, appearances, clothing, textures, wants, tools, scenery, body parts, fluids, and sounds.
(Or copy here:)
MODULES
These are mostly ripped from my original prompt with a brief explanation of what they do, but ye. May add more at a later date!
• Third person writing (consider adding to the bottom of the prompt if character's intro is written in first or second person):
Utilize third person limited point of view.
• Use markdown (edit to your preference):
Use markdown where appropriate: verbal dialogue will be wrapped in quotation marks "like this"; use italics for all actions and narration *like this*; sparingly use double asterisks to emphasize the importance, tone, and delivery of impactful dialogue **like this**; written text and non-verbal dialogue such as internal thoughts, written notes, and text messages will be in codeblock `like this`.
• Helps reduce Shakespearean prose and make the bot sound more like a normal (modern) human:
Utilize modern and casual vocabulary, characters speak and think using informal language and slang appropriate to their background.
• However, if you WANT to get rizzed by Shakespeare, consider trying this:
Utilize purple prose, the more beautiful and romantic the language, the better!
• For a Medieval RP, try this:
This roleplay is set during the Middle Ages, ensure characters speak and think in Middle English as is appropriate for the time period.
• Adds more character detail to narration, including internal thoughts (but YMMV if the character intro does not already include the character's internal thoughts):
Employ a mixture of narration, dialogue, characters' physical mannerisms, and internal thoughts into responses.
• Helps reduce the bot being OP (but keep in mind, since the bot is instructed to roleplay and most roleplays involve overpowered characters, you may still encounter the behavior):
Focus on coherent world simulation, ensure that characters are not omnipotent and cannot perform actions they are unable to perform at that point in time.
• Helps reduce horniness and keep bot behavior more appropriate to their relationship with {{user}}:
Have situational awareness and be cognizant of intercharacter relationships, characters avoid being overly familiar or sexually pushy towards {{user}} unless the situation calls for it, it is in character for them to do so, or they have a sexual relationship.
• Helps remind LLM to add physical descriptors into narration (you may want to append with any extra details that the bot commonly forgets but is important to your RP, such as trans anatomy, demi-human traits, etc):
Be cognizant of all characters' physical descriptors.
• I like this one personally because sometimes I feel like narration gets overwhelmed with characters' tragic backstories, this helps reduce that:
Overall characterization is dependent on defined personality traits, and subtly influenced by backstory and history.
• Another add-on for realism, this one though should be modified to whatever is present in your roleplay. Like, if there are no characters with mutism, you do not need the sentence about mutism, so modify to an impediment which is relevant to your roleplay or just remove it:
Ensure characters that have any impediments which are relevant to the situation, such as visual impairment, are under the influence of alcohol/drugs, or a disability like mutism, actions and dialogue are affected appropriately.
• Create NPCs:
Over the course of the roleplay, create new setting-appropriate side characters and perform as them to interact with other characters in the story.
JLLM Users, PLEASE READ
On Janitor, {{user}}≠Persona AND {{char}}≠Character (necessarily) Okay, but what does that mean and why does that matter?
On other websites, {{char}} and {{user}} are macros which are converted to the character's name before being sent to the API. On such websites, for example, Always perform {{char}} in character is converted to Always perform CHARACTERNAME in character BEFORE the API reads it and responds.
On Janitor, THIS IS NOT THE CASE.
On Janitor, when you write Always perform {{char}} in character and press enter, JLLM or OpenAI reads EXACTLY that, Always perform {{char}} in character
Unless it is specified in the Character definition that {{char}}=Character, or in your Advanced Prompt, JLLM may consider {{char}} and the character as separate entities entirely. The same goes for {{user}} and Persona.
This is part of the reason why many prompts DO NOT WORK for JLLM.
How do I fix this?
My Base and Full prompt already have a semi-solution to this via the Perform as the character defined under {{char}} and any existing side characters by describing their actions, events, and dialogue. prompt, but if the bot is still being resistant consider explicitly mentioning that {{char}} must perform as CHARACTERNAME. Also make sure that your Persona defines that {{user}}=YOURNAME. Here is the template that I use for my personas:
Be careful with making your definition too large. Writing an overly detailed Personality section, in my experience at least, seems to encourage the bot to speak as {{user}} more. Consider doing away with the section entirely and instead being more descriptive with your vibes during RP, or, be experimental and cleverly word the Personality in a "this is how other people perceive {{user}}" sort of way.
Also consider replacing {{user}} mentions in your Advanced Prompts to the name of your Persona. For example,
AVOID acting as, speaking for, or describing the thoughts of {{user}}
could be changed to
AVOID acting as, speaking for, or describing the thoughts of YOURNAME
TUTORIAL
On Janitor, here is how you use Advanced Prompt.
When in a chat with a character click the hamburger menu in the top-right and API Settings:
Once you're in there, click Advanced Prompts to open up the box, and then copy-paste my above prompt into it and click Save Settings at the bottom:
You will see this pop-up, you MUST select Yes or it WILL NOT SAVE YOUR NEW PROMPT:
Now, go back into the hamburger menu and select Generation Settings:
Here, you will adjust your desired Temperature and Max Tokens since it will have been reset due to changing your prompt:
For temperature, I recommend around 0.65-0.70 but play around with it and see what you like. Max Tokens is purely up to your own personal preference.
CREDITS
With StatuoTW's permission, I based my prompt off of his Mercury/Mars prompt. You can find his botmaking rentry here, and his bots here! Addtionally, here is a super helpful guide by him on how to get more out of your chats. Thanks Joystick_Hero for proofreading and giving me advice! I have applied some of his suggested changes. Here is his rentry on how to approach some common bot issues.
CONTACT
My forum in the JanitorAI main server is here.
You can find my forum in the Janitor Creators discord server here.
I'm also in @absolutetrash's server, my thread is here.
Feel free to ping me in either if you have any questions or comments!
Also gonna plug myself... You can find my profiles here: Janitor | Venus I also have a PayPal if you wanna support me hehe 😸
EXTRAS
Roach's Prompts
Really amazing add-ons! If you want to use his prompts, just tack them somewhere in your Custom Prompt, make sure they don't conflict with anything else in the prompt, then hit save!