What problem does it solve? Jumping into implementation with unclear requirements leads to rework, scope creep, and misaligned outcomes. This Skill structures the pre-implementation phase by interviewing the user one question at a time, researching technical options, and converging on a documented MVP scope. ## Core Features & Use Cases - Task-First Capture: Immediately creates a task directory and seeds a PRD so ideas are recorded before any Q&A begins. - Question Gating: Filters out low-value questions by deriving answers from repo inspection, docs, and research, asking only blocking or preference questions. - Research-First Decisions: Delegates technical-choice research to a trellis-research sub-agent that persists findings to research files, then presents 2-3 concrete approaches with trade-offs. - Diverge-Converge Workflow: Expands thinking across future evolution, related scenarios, and edge cases, then locks an explicit MVP scope with out-of-scope items. - Use Case: When a user describes a vague new feature like "add group chat delegation", the Skill creates the task, inspects the codebase, researches orchestration patterns, asks targeted preference questions, and produces a PRD with acceptance criteria and an implementation plan. ## Quick Start Ask the AI to brainstorm requirements for a new feature you are considering before writing any code.
