DEATH or TREAT

--- ## 👻 Recommended Personal

※ In this main story, the character is treated as selfish and self-centered

Name:
Gender:
Age: 2 years after graduating high school (i.e., 20 years old)
Personality (at the start):

• Emotional and a bit self-centered
• Acts before thinking
• But people don't leave him because of the 'loneliness' hidden behind his clumsiness

Relationships (from other characters' perspectives):

• Reina: "I can't just leave him alone"
• Kana: Thinks of him as "the one who took everything from me," but can't hate him completely
• Haruto: "I got involved out of curiosity," but was unexpectedly influenced
• Riku: "A presence that I'm always curious about, for no reason"

👻 How to Play

🩸 On Halloween night, 7:00 PM to 11:59 PM
You will be killed many times during that time.

🩸 Every time you die, you are asked, "DEATH or TREAT?"

🩸 If you choose TREAT, you'll return to the beginning of the party.
The fate of death will follow you, changing its form.

🩸 Difficulty: High (Recommended: Those who like to analyze and deduce)

🩸 However, fate looks at your actions and words more than game-like strategies. Depending on you, you might be able to change the fate of death.
☝️ If you're unsure, just choosing an option is fine. It won't break anything, no matter what you input.
Sometimes, choosing DEATH might not be so bad.

🩸 If you write down the events of each loop in user notes or long-term memory, the AI might use them in the performance (but they will pretend to forget each time the night returns)

🪞 About SATSUI

🩸 In this mansion, the invisible "heat of emotion" is shown as a numerical value, "SATSUI."

🩸 But that number is just a guideline.
Numbers aren't everything. What's important is the atmosphere, the timing, and the weight of words.

🩸 What you should believe is not SATSUI, but your own "feelings."
Even if it's +100 or -100, that single moment of gaze might be everything.

🧠 About the AI Model

▶︎ High-performance model recommended
→ Superior in notation stability, depth of acting, and event understanding

▶︎ When using the standard model (not recommended)
→ Starting with the high-performance model for the first few turns makes it easier to stabilize even after switching to the standard model (Among the standard models, Dino is a bit straightforward, but it's generally easy to use. Donald and Deepseek are also acceptable)

▶︎ It's effective to switch models regularly (Stability UP even with exchanges between standard models. Even better with a high-performance model in between)

▶︎ Even if the notation is slightly off, the conversation can continue without problems
→ If it's off, manually adjusting it may cause it to relearn.
→ Even with high-performance models, the format becomes sloppy when the acting gets intense

🥀 Request

▶︎ If there's any clearly strange behavior, please let me know in the comments.

🪽 Referral code for 100 feathers → nQwUnKrEHI