What problem does it solve?
The question skill prevents wasted research and misaligned implementation by forcing early clarification of intent, scope, edge cases, defaults, and user preferences using evidence from existing docs and code.
Core Features & Use Cases
- Doc- and code-grounded interviews: Searches and reads relevant vault artifacts and inspects the code for existing behavior and vocabulary, asking humans only for missing intent, tradeoffs, priorities, or domain facts.
- Branch-by-branch design grilling: Walks the design tree through user-visible behavior, scope, ownership/persistence, compatibility/migration, failure cases, non-goals, and acceptance boundaries, updating the decision tree after each answer.
- Reusable Question Summary handoff: Produces a structured Question Summary containing project vocabulary, evidence map, resolved answers with rationale, covered decision tree, blocking open questions, recommended defaults, and explicit non-goals.
Quick Start
Ask an AI to run the question phase for your feature request and return a Question Summary suitable for carrying into later QRDSPI phases.