What problem does it solve? Writing unambiguous software requirements is hard: vague prose leads to misinterpretation, scope creep, and untestable acceptance criteria. This Skill structures requirement definition using the EARS/GEARS notation and a standardized 3-file SPEC format so every requirement is testable and traceable. ## Core Features & Use Cases - GEARS/EARS Requirement Patterns: Five canonical patterns (Ubiquitous, Event-driven, State-driven, Capability gate, Unwanted) with a unified compound clause and generalized subject for precise requirement authoring. - Standardized SPEC Structure: Enforces a 3-file layout (spec.md, plan.md, acceptance.md) with a 12-field metadata schema, lifecycle states, and lint validation rules. - Plan-Run-Sync Integration: Connects SPEC creation to implementation and documentation sync phases, including Git Worktree-based parallel development and clarification gates. - Use Case: When planning a new feature like user authentication, use this Skill to produce a SPEC with EARS requirements, quantified success criteria, and Given-When-Then acceptance scenarios before any code is written. ## Quick Start Ask the AI to create a SPEC document for your feature using EARS-format requirements and acceptance criteria, for example by running /moai plan followed by a feature description.