requirement-qa

To turn vague project ideas into structured requirement documents, you provide a short prompt and the agent drafts modular markdown files like project-profile.md under docs/01-requirement/. It incrementally updates these files through a review loop until all details are confirmed.

Structured requirement clarification works by generating 2-5 multiple-choice options for each ambiguity in your feature scoping, recommending one choice, and annotating inferred assumptions. This interactive QA process continues until you approve the generated requirement artifacts.

The best way to document inferred assumptions during user interviews is to annotate them directly within the generated markdown files. The agent logs all QA sessions to docs/qa/ for auditability, ensuring every inferred assumption is reviewable and traceable.

Yes, you can use this requirement QA process for bug fixes and refactors. It applies during discovery and requirement clarification for new projects, features, or technical debt, producing modular artifacts like business-rules.md and tech-constraints.md.

For downstream spec writing, you need to generate modular markdown files under docs/01-requirement/, including project-profile, feature-scope, business-rules, non-functional, and tech-constraints documents to feed directly into implementation skills.

To handle incomplete project requirements without repeated clarification overhead, the agent asks targeted questions with predefined options and persists session logs. This structured approach eliminates continuous back-and-forth by finalizing reviewable documents in incremental steps.