SuperAPI AI Directory

unity-game-balance

Agents & models

View repository

Updated 10-16 08:49

About this skill

Unity game numeric design

Convert experience goals into explainable, adjustable, and verifiable models; do not substitute undocumented subjective tuning for evidence.

Inputs and Decisions

Read the Work Item, related gameplay, configuration, tests, GDD, and docs/balance.md. Hand off to when there are substantive trade-offs in target pacing, economy, monetization impact, allowed deviation, acceptance thresholds, or high-risk curves; within the remaining scope, reversible configuration may proceed after being recorded in the implementation package. $unity-game-grilling; reversible configurations within the remaining scope shall be advanced after being recorded in the implementation package.

Execution and Handoff

  1. Define player pacing, success rates, resource sources/sinks, failure costs, pity, upper/lower bounds, and abnormal behavior.
  2. In balance.md record parameters, units, default values, sources, formulas, sensitivity, impact, and validation methods; at runtime, read through centralized, versionable configuration.
  3. For key curves, save reproducible calculations, simulation seeds, sample sizes, and playtest records; do not treat spreadsheet results as in-game validation.
  4. Provide configuration, boundaries, and acceptance examples to gameplay, and provide statistical definitions, tolerances, and high-risk scenarios to QA.
  5. When changing parameters, mark affected scenarios, saves, tests, and performance conclusions, and trigger necessary revalidation.