From game-design
Use when the user wants help finding the fun in a prototype, pivoting a game's direction based on playtesting, deciding what to keep or cut after prototyping, or using emergent design methodology, including rapid prototyping, following the fun, and recognizing when bugs or unintended behaviors should become features
How this skill is triggered — by the user, by Claude, or both
Slash command
/game-design:design-discoveryThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Help the user discover what their prototype is trying to become. Read `references/framework.md` for the full design discovery model, and check `references/gotchas.md` for common mistakes that derail the process.
Help the user discover what their prototype is trying to become. Read references/framework.md for the full design discovery model, and check references/gotchas.md for common mistakes that derail the process.
The best games are often discovered, not designed. This is bottom-up design: prototype fast, play it, recognize what's fun -- even if unintended -- then redesign around it. Juggle mechanics born from combat bugs, defining aerial moves born from physics accidents, horror games born from what was supposed to be exploration. The methodology is prototype, play, recognize emergent fun, redesign around it.
Bad starting ideas are fine. Many successful games began with loose, fuzzy, or derivative premises. The fun came from iteration, not from the quality of the seed idea. Don't wait for a brilliant starting concept -- start building and listen.
This applies most powerfully in early development -- the "white-hot malleable magma" phase -- when the game can still become anything. But the same methodology (play, observe, redirect) is valid across the full development arc: puzzle generation in a mature engine, content design late in production, tonal refinement near ship.
Creative constraints (thematic anchors) narrow the search space and speed discovery. Anchoring one element -- a theme, a core mechanic, a visual style -- makes iteration faster, not slower, because you're exploring deeply within a defined space rather than broadly across an undefined one.
Walk the user through what their prototype is telling them. Identify emergent fun, evaluate whether it deserves to become the core, and help them commit or cut.
Provides a checklist for code reviews covering functionality, security, performance, maintainability, tests, and quality. Use for pull requests, audits, team standards, and developer training.
npx claudepluginhub tielur/claude-game-studiokit