---
name: game-design
description: Discuss and plan game experiences, mechanics, content scope, and focused playable experiments. Use for concept exploration, design tradeoffs, balancing plans, or deciding what to prototype next.
---

# Game Design

Read existing project decisions and the user's current request. Preserve explicit choices; do not inherit a genre, camera, grid, party size, resource model, or art specification from another project. Use the decision states in [shared contracts](../../shared/contracts.md) when discussion needs continuity.

Start from a concrete play situation: what the player observes, does, chooses between, and receives as feedback. Identify the largest unresolved design question. Offer a few meaningfully different approaches with experience and production tradeoffs, then ask only what changes the next step. Do not require a fixed number of questions or discussion rounds.

When enough is known, summarize the intended experience, repeated play loop, adopted decisions, provisional assumptions, and smallest playable experiment. Describe what the experiment can establish and what would lead to revision. A prototype can test usability or rule behavior; claims about enjoyment still need play feedback.

For numerical systems, define units, turn or time semantics, resource sources and sinks, bounds, and interactions before tuning values. Use calculations or simulations for specific balance questions and distinguish model outcomes from actual player behavior.

For content planning, connect encounters, spaces, narrative events, and progression to the intended player experience. Estimate scope using concrete content and implementation needs, not an invented duration guarantee. Keep deferred ideas separate from committed work.

If implementation is requested, hand off the playable target and its acceptance observations to the relevant production skill. If the request is planning only, return a reviewable plan without modifying the game.