What problem does it solve?
Software projects accumulate complexity from small convenience promises—fallback parsers, compatibility aliases, duplicate fast paths, pixel-perfect reproductions—that each force an entire second system to be maintained. This Skill provides a disciplined decision framework for finding the one small promise whose refusal collapses a disproportionately large code family while keeping the product's core value intact.
Core Features & Use Cases
- Refusal Point Analysis: Systematically lists candidate refusal points (legacy aliases, rare modes, provider exceptions, exact reproduction) and maps each to the code family it owns.
- Decision Templates: Provides structured templates for specs and design notes covering the product sentence, deletion prize, user loss, and final decision.
- UI Refactor Guidance: Includes an artifact ladder (shared primitive, screenshot, sketch, partial tree, full HTML) so teams use the smallest artifact that preserves the design decision.
- Use Case: When refactoring authentication, evaluate whether refusing a Google GIS fast path lets you delete per-provider helpers, blocked-browser UI, and duplicate test paths while keeping one OAuth 2.1 PKCE invariant.
Quick Start
Ask the AI to apply the asymmetric wins framework to evaluate whether a specific feature or compatibility promise in my codebase is worth the complexity it forces.