AI Agent Hub
Back to skills
Requirement Clarifier icon

Requirement Clarifier

Development Updated 2026.08.30

Paste the following prompt into your AI chat to install this skill:

Please follow https://skillhub.cn/install/skillhub.md and install @user_84fd5459/requirement--clarifier

About this skill

Problem

Requirement docs often hide three kinds of risk: vague business goals, implicit technical assumptions, and missing test boundaries. A single review pass can miss cross-role contradictions or mix blockers with optional polish.

How It Works

requirement-clarifier runs multi-round reviews in order: product manager, tech lead, senior test engineer. Each round asks one critical question at a time, prioritizing red critical defects, yellow important gaps, and green optimization suggestions; user answers are recorded, then the same role re-reviews the updated context. PM checks business goals, user stories, acceptance criteria, scope, and market context; tech lead checks feasibility, data models/APIs, performance/security, architecture impact, and solution trade-offs; test engineer checks testability, edge cases, verifiable acceptance, data consistency, and logic consistency.

It then outputs {requirement}_review_report_{YYYY-MM-DD}.md and {requirement}_requirement_doc_refined_{YYYY-MM-DD}.md, preserving original wording, marking TBD items and pending decisions instead of rewriting or inventing requirements.

Notes

Best for PRDs, requirement notes, and feature specs; .docx input should be converted to Markdown first. If a role exceeds three rounds with unresolved critical issues, the workflow suggests pausing for more information. Green suggestions are listed rather than interrogated, making them suitable for selective adoption.

Use Cases

  • Before development, check the requirement doc for missing business goals, acceptance criteria, or interface boundaries.
  • Convert a .docx requirement note to Markdown, then clarify it through PM, tech-lead, and test passes.
  • Before test design, review edge cases, error scenarios, data consistency, and logic coherence.
  • Generate a review report and refined requirement doc, listing pending items and optimization ideas as TBD.

Best For

  • Product manager drafting a PRD who wants to catch contradictions, gaps, and untestable acceptance criteria.
  • Tech lead planning a release who wants to validate APIs, data models, and performance constraints.
  • Test engineer designing cases who wants to review edge cases, error paths, and consistency risks.
  • Project owner taking over legacy specs who wants missing sections and pending decisions listed clearly.