Authors a web-game concept — pitch, core loop, MDA, mechanics, progression/economy, content scope, visual/audio direction, technical feasibility (engine, frame/memory/load budgets on mobile web, networking), accessibility, metrics, prototype plan. Produces docs/specs/game-concept.md.
Scanned 9/22/2026
Install to Claude Code
npx -y skills add gonimar/claude-web-studio --skill game-concept --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Game Concept?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/gonimar-game-concept)More formats (shields.io, HTML) on the badges page.
---
name: game-concept
description: "Authors a web-game concept — pitch, core loop, MDA, mechanics, progression/economy, content scope, visual/audio direction, technical feasibility (engine, frame/memory/load budgets on mobile web, networking), accessibility, metrics, prototype plan. Produces docs/specs/game-concept.md."
argument-hint: "[game title] | gate"
user-invocable: true
allowed-tools: Read, Glob, Grep, Write, Edit, AskUserQuestion, Task
model: sonnet
agent: game-lead
---
# Game Concept
Reply in the project conversation language (CLAUDE.md → Language); code, identifiers, paths and commit messages stay in English.
Template `.claude/docs/templates/game-concept.md`. Reference `threejs-webgames.md`.
## Phase 1: Pitch and loop
Questions: genre, the player's "verb", session (minutes), platform (mobile web first?), single/multiplayer, references. Three core-loop variants with "why it is fun to repeat".
## Phase 2: Mechanics, progression, content
Rules with parameters (mark: in data); a difficulty-curve table; MVP volume vs full version.
## Phase 3: Feasibility
Engine with justification (3D → three.js r185 WebGPU/WebGL2; 2D → PixiJS 8/Phaser); budget: 16.6 ms frame, draw calls, memory, first load on 4G; networking (multiplayer → server-authoritative on Go, tick rate); saves; **spikes** to test risks (e.g. "1000 instances on a mid-range phone").
`threejs-engineer`/`web-game-engineer` via Task — risk assessment for their part (in parallel).
## Phase 4: Accessibility, metrics, prototype
Accessibility settings; retention metrics; prototype plan: what playability validates first, timeline, the "fun" criterion — measurable (sessions, minutes, a question to N testers), so `gate` can record a result, not an opinion.
## Phase 4b: `gate` — the prototype go/no-go
`/game-concept gate` (after the first playable slice, before backend/multiplayer investment): read §11 of `docs/specs/game-concept.md`, ask for the measured result against the "fun" criterion (playtest notes, numbers), and record `GO | NO-GO | PIVOT (what changes)` with the evidence in `production/releases/gate-prototype.md` and as a "Result" block under §11 — after "May I write?". `NO-GO`/`PIVOT` → the next step is a concept revision, not the next feature. The catalog step `prototype-gate` and `/help` look for this artefact.
## Phase 5: Write
"May I write `docs/specs/game-concept.md`?" — one `AskUserQuestion`: write (Recommended) · show the draft/diff first · not now. Propose an engine ADR (`/architecture-decision`). After the "write" answer: `touch .claude/.write-consent` (rule 7 — the consent-guard hook checks the marker).
Verdict: `APPROVED` | `NEEDS REVISION` | `GO` | `NO-GO` | `PIVOT`. Next step — one `AskUserQuestion`: `/product-spec` (light) (Recommended) · `/architecture-decision` (engine) · prototype via `/dev-story`.
Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.
No comments yet. Be the first to comment!