What problem does it solve?
This Skill helps product teams turn a structured PMContext into two aligned PRD outputs: one optimized for AI execution and one optimized for human review. It prevents teams from drafting requirement documents directly from vague chat context, which often leads to lost assumptions, untraceable decisions, and inconsistent delivery artifacts.
Core Features & Use Cases
- Dual-form PRD generation: Produces both ai-prd.md and human-prd.md from the same PMContext source so execution rules and review narrative stay aligned.
- Strict traceability: Ensures every rule, acceptance criterion, and boundary maps back to facts, assumptions, conflicts, or decisions already captured in PMContext.
- Workflow orchestration: Coordinates downstream PRD generation steps, supports auto mode, and handles missing PMContext by routing through the upstream requirement flow when arguments are provided.
- Risk-aware output: Preserves [待确认] and [冲突] signals instead of upgrading them into facts, and limits output when uncertainty is too high.
- Use case: A product manager has already captured user needs, constraints, and open questions in PMContext and now needs both an AI-executable product spec for implementation support and a human-readable PRD for stakeholder review.
Quick Start
Ask the agent to run pm-prd on your existing PMContext to generate both the AI PRD and the human PRD, or provide a requirement description with auto mode to create the context and PRDs in one flow.