What problem does it solve? Designing puzzle and engine-builder mechanics that feel fair and earned is hard: additive scoring feels flat, unpredictable outcomes read as slot machines, and untested balance changes break quota curves. This Skill encodes the design principles distilled from Balatro, Slay the Spire, Wordle, and NYT Connections research so every new part, quota curve, or boss twist is judged against proven criteria. ## Core Features & Use Cases - Mechanic evaluation checklist: Judge any proposed mechanic against five ordered questions covering predictability, interaction depth, positional relevance, bot-verifiability, and mobile legibility. - Balance tuning method: Use lazy-bot and competent-bot harnesses to verify quota curves, treating bot results as the pessimistic bound. - Boss twist and daily puzzle rules: Design Jam rounds that attack the build rather than luck, keep daily seeds comparable, and share performance without spoiling boards. - Use Case: When adding a new part to the Payload engine builder, run it through the checklist to confirm players can predict its consequence, it interacts with at least two existing parts, and the placement preview covers it. ## Quick Start Ask the AI to evaluate a proposed new game part or quota curve change using the puzzle-design principles before implementing it.
