---
name: character-simulation
description: Simulate character decisions under goals, beliefs, fears, knowledge limits, relationships, incentives, constraints, and current pressure, in order to test agency and behavioural consistency. Use as a plausibility test for any significant character choice.
---
# Character Simulation v2.1

Simulate the decision the character would actually face, using only the information and pressures they actually have. This is a plausibility test, not a canon source and not a way to outsource authorship.

## Inputs

At a decision point, inspect the character's current knowledge, goals, fears, beliefs, values, constraints, relationships, recent events, physical and material state, and the options genuinely available to them.

The knowledge limit is the discipline here. Simulation that quietly grants the character authorial knowledge produces conclusions about a person who does not exist.

## Method

1. Establish what the character wants right now, and what they believe about the situation.
2. Enumerate the options their beliefs and circumstances actually permit.
3. Identify the dominant motivation for each viable option, and the strongest reason to reject it.
4. Note which options are blocked by constraint rather than by preference.
5. Name the most likely choice, and the strongest alternative they might plausibly have made.
6. State the consequences of each.

Weight motivations qualitatively and state your reasoning. Do not produce false precision in the form of numeric probabilities that imply a rigour the evidence does not support.

## Flags

- Unexplained personality shifts.
- Impossible knowledge.
- Coercion not established, used as a shortcut for compliance.
- Behaviour that contradicts an established moral boundary without anything breaking that boundary.
- Author convenience — a choice made because the plot needs it and nothing in the character would produce it.
- Second-choice blindness — the character takes an obviously inferior option for no visible reason.

## Rules

Never force simulation results into canon. Where simulation contradicts the plan, report the contradiction and offer the smallest adjustment to either the pressure or the character that would resolve it. Changing the plan to suit the simulation is legitimate; changing the character to suit the plan is a retcon and routes through `canon-manager`.

## Coordination

Use `character-development` for the underlying psychology, `character-arc-engine` for where the decision sits in the arc, `mystery-engine` for knowledge states, and `dialogue` for how the decision is expressed.