From rlm-gaming
The Warden + The Playtester run a synthetic playtest, balance Monte-Carlo, and telemetry tuning, then route tuning recommendations to the design heads with an accessibility-floor check.
How this command is triggered — by the user, by Claude, or both
Slash command
/rlm-gaming:game-balance-pass <system/encounter to balance> [--telemetry] [--iterations <n>]sonnetcommands/The summary Claude sees in its command listing — used to decide when to auto-load this command
# /game-balance-pass **The Warden** (QA/balance) and **The Playtester** (synthetic players) run a balance sweep: persona-driven synthetic playtests, a balance Monte-Carlo simulation, and telemetry-informed tuning, producing concrete tuning recommendations that route back to the owning design heads — and a check that no change drops the game below its accessibility floor. Uses the `game-qa-and-balance` skill. This command tunes existing design; it does not create new features. ## Steps 1. **Frame** — The Warden defines the balance question, the fun-proxy metrics, and the win/clear-rate ta...
The Warden (QA/balance) and The Playtester (synthetic players) run a balance sweep: persona-driven synthetic playtests, a balance Monte-Carlo simulation, and telemetry-informed tuning, producing concrete tuning recommendations that route back to the owning design heads — and a check that no change drops the game below its accessibility floor. Uses the game-qa-and-balance skill. This command tunes existing design; it does not create new features.
--iterations to find dominant strategies, dead options, and difficulty cliffs.--telemetry, fold in live clear-rate/drop-off/build-popularity data to ground the sim against reality.game-accessibility-guidelines (owner: The Warden) to confirm no tuning change raises the skill floor past the accessibility baseline (e.g. assist-mode targets still clearable).DECISION_RECORD with the recommendation table and verdicts; eights.memory.add(domain="gaming")./game-balance-pass "ranged-vs-melee weapon parity" --telemetry --iterations 5000
The Playtester finds melee unviable above tier 3, the Monte-Carlo confirms a ranged-dominant meta, telemetry agrees, and The Warden routes per-weapon tuning to The Systemsmith plus a difficulty-floor note to The Duelist — all while passing the accessibility-floor check.
This is largely an in-crown evaluation; tuning recommendations are design artifacts handed to design heads. If a recommendation requires a code/tuning-table change, the owning head emits a DEV_TASK to engineering (game-bug-fix-team or game-feature-team). No assets are commissioned.
npx claudepluginhub lebobo88/rlm-gaming --plugin rlm-gaming