bmad-prd

Create, update, and validate product requirement documents through guided discovery and reviewer gates.

Updated Mar 14, 2026
One-click install
npx skills add https://github.com/ArchaonHW/MingGoRTS --skill bmad-prd-archaonhw
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: bmad-prd
Source: https://github.com/ArchaonHW/MingGoRTS/tree/main/.agents/skills/bmad-prd
Command: npx skills add https://github.com/ArchaonHW/MingGoRTS --skill bmad-prd-archaonhw

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) and assets (resource) components.

What problem does it solve? Writing a PRD that downstream teams can actually build from is hard: requirements end up vague, assumptions go unrecorded, and reviews happen too late. This Skill guides the full PRD lifecycle—creation, updating, and validation—with structured discovery, a persistent decision log, and multi-reviewer quality gates. ## Core Features & Use Cases - Three intent modes: Create a new PRD from a brain dump, Update an existing PRD against a change signal, or Validate a PRD with a seven-dimension quality rubric without modifying it. - Guided discovery with working modes: Choose a Fast path (batched questions, tagged assumptions) or Coaching path (Vision + Features or Journey-led entry points) calibrated to project stakes. - Auditable workflow: Every decision, change, and override is appended to a memlog; reviewer subagents write findings to per-reviewer files synthesized into HTML and Markdown validation reports. - Use Case: A product manager needs a launch-grade PRD for a new consumer app. The Skill elicits user journeys with named protagonists, drafts grouped features with globally numbered FRs, runs rubric and adversarial reviewers, and finalizes a polished prd.md plus addendum. ## Quick Start Ask the assistant to help you create a PRD for your product idea and answer the discovery questions it asks.

Frequently Asked Questions about bmad-prd

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I create a PRD with an AI assistant?▼

Invoke the skill with your product idea and it runs a discovery phase: brain dump, stakes calibration, and a choice between Fast path (drafted PRD with tagged assumptions) or Coaching path (section-by-section collaboration). The result is a prd.md with grouped features and numbered functional requirements.

How do I validate an existing PRD for quality?▼

Use the Validate intent with the path to your prd.md. A rubric-walker subagent scores seven dimensions such as decision-readiness and done-ness clarity, and the results are synthesized into HTML and Markdown validation reports with severity-ranked findings.

What is the difference between Fast path and Coaching path?▼

Fast path batches remaining questions into one or two rounds, then drafts the full PRD with [ASSUMPTION] tags for inferred content. Coaching path walks PM-thinking sections together, starting from either a Vision + Features or Journey-led entry point.

Can the PRD workflow run headless without user interaction?▼

Yes. When invoked headless, the skill infers intent from provided inputs, records unconfirmed values as assumptions or open questions, and returns a JSON status payload with artifact paths instead of prompting the user.

How are PRD decisions and changes tracked across sessions?▼

Every decision, change, and override is appended as a one-line entry to a .memlog.md file in the run workspace via a shared script. The memlog is audited at finalize and used to resume in-progress runs or reconcile updates.

When should I use a product brief instead of a full PRD?▼

For one-pager scope or very small efforts, the skill itself suggests alternatives during its misroute scan, such as a product brief or direct build workflow. Full PRDs suit launch-grade or chain-top work feeding UX, architecture, and epics.