| name | concept-game-builder |
| description | Use when Codex should turn a concept, topic, tutorial, lesson, algorithm, system, workflow, or abstract idea into a playable browser game that teaches through interaction. Use for planning or implementing educational web games with the Game Studio plugin, including concept-to-mechanic translation, gameplay loop design, learning progression, and browser-game QA. |
| metadata | {"dependencies":[]} |
Concept Game Builder
Create a browser game that teaches by making the player act on the concept.
This skill owns the learning design and concept-to-game translation; Game Studio
owns engine choice, runtime implementation, assets, UI, and playtesting.
Core Workflow
- State the learning goal in one sentence: what should the player understand,
predict, or do after playing?
- Decide whether a game is the right medium. Use a game when the topic has
decisions, constraints, feedback loops, failure states, strategy, incentives,
sequencing, systems, or tradeoffs. If the request needs an explorable article
more than a game, route to
visual-learner instead.
- Start concrete. Put a playable challenge in the first viewport before
definitions, lectures, or long explanations.
- Convert the concept into a playable model:
- player verbs
- state variables
- rules and constraints
- visible consequences
- success and failure conditions
- inspectable explanations after action
- Use Game Studio for production planning and implementation:
- Load
game-studio:game-studio or $game-studio after the learning model is
clear.
- Default to a 2D Phaser route unless the user asks for 3D, Three.js, React
Three Fiber, or a specific engine.
- Route HUD and overlays through
game-studio:game-ui-frontend when present.
- Route browser QA through
game-studio:game-playtest before calling the game
done.
- Build or plan the smallest complete game that teaches the topic. Prefer one
polished core loop over many disconnected minigames.
Game Design Contract
Define these before writing code:
- Player mission: a 10-30 second challenge the player can attempt with no
prior vocabulary.
- Learning model: the concept's variables, rules, and relationships.
- Misconception: the naive idea the game should test or overturn.
- Player verbs: actions that map directly to the concept, such as allocate,
route, tune, trade, predict, combine, debug, classify, or negotiate.
- Feedback: how the game shows consequences through motion, score, state,
resources, failure, or changed world behavior.
- Progression: 3-5 levels, rounds, or scenarios that add one concept at a
time.
- Explanation timing: explain after the player acts; keep text synchronized
with the current game state.
- Inspection path: hover, focus, click, pause, replay, or detail panels that
reveal the rule, formula, trace, or evidence behind the result.
Game Studio Handoff
When handing the plan to Game Studio, include this compact brief:
Learning goal:
Core loop:
Player verbs:
Concept model:
Progression:
UI/HUD needs:
Asset style:
Target runtime:
Playtest checks:
For implementation tasks, continue through Game Studio rather than inventing a
separate stack. Use its specialist routing for Phaser, Three.js, React Three
Fiber, UI, assets, and playtesting. If Game Studio is unavailable in the current
session, follow its defaults manually: 2D Phaser first, DOM HUD for menus and
learning panels, responsive controls, and browser playtest verification.
Quality Rules
- Do not make a quiz with decorative motion and call it a game. The answer
choices must change game state, strategy, or consequences.
- Do not hide the concept behind lore. Theme should make the model easier to
remember, not replace the model.
- Do not front-load exposition. Let the player do something first, then name the
concept.
- Keep the first build small enough to finish and playtest in one session.
- Make win, loss, replay, reset, and pause states explicit.
- Keep touch, keyboard, and responsive layouts viable from the start.
- Expose toy-model assumptions when the numbers are illustrative.
- Verify with an actual browser when possible; check that the game renders,
responds to input, avoids overlapping text, and teaches the intended takeaway.
Reference
Read teaching-game-patterns.md when
choosing mechanics for a topic or when the first concept-to-game mapping feels
too literal.