What problem does it solve? Domain terminology drifts and architectural decisions get forgotten, leaving future readers guessing why the code looks the way it does. This Skill actively sharpens a project's domain model by challenging fuzzy terms, stress-testing concepts with concrete scenarios, and recording decisions the moment they crystallize. ## Core Features & Use Cases - Ubiquitous Language Management: Maintains an opinionated CONTEXT.md glossary with canonical terms and explicit _Avoid_ lists, supporting both single-context repos and multi-context setups via CONTEXT-MAP.md. - Architectural Decision Records: Creates numbered ADRs in docs/adr/ only when a decision is hard to reverse, surprising without context, and the result of a real trade-off. - Live Model Sharpening: Challenges conflicting terminology, invents edge-case scenarios to probe domain boundaries, and cross-references user statements against the actual code. - Use Case: While designing a billing feature, the user says "account" — the Skill asks whether they mean Customer or User, resolves the term, updates CONTEXT.md inline, and offers an ADR when the team picks event-sourced writes over a CRUD model. ## Quick Start Ask the assistant to help pin down the domain terminology for your project and record any significant architectural decisions as ADRs.