What problem does it solve? Starting to code with vague or ambiguous requirements leads to rework, misaligned expectations, and architectural mistakes. This Skill forces clarity by writing a reviewed specification before any implementation begins. ## Core Features & Use Cases - Gated Four-Phase Workflow: Progresses through Specify, Plan, Tasks, and Implement phases, each requiring human review before advancing. - Structured Spec Template: Covers objective, commands, project structure, code style, testing strategy, and three-tier boundaries (Always/Ask First/Never). - Assumption Surfacing: Lists implicit assumptions upfront and reframes vague requirements into concrete, testable success criteria. - Use Case: When asked to "make the dashboard faster," the Skill reframes this into measurable targets like LCP under 2.5s and CLS under 0.1, then produces a spec, plan, and task list before writing code. ## Quick Start Ask the AI to write a specification for your new feature before implementing it, covering the objective, tech stack, commands, and success criteria.