菜单
Build Product Requirements Documents from Product Briefs. Use when transforming product vision into actionable requirements for engineers and designers. Produces high-level design principles, use case compendiums, and milestone definitions. A PRD focuses on what, not how.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
| name | writing-prds |
| description | Build Product Requirements Documents from Product Briefs. Use when transforming product vision into actionable requirements for engineers and designers. Produces high-level design principles, use case compendiums, and milestone definitions. A PRD focuses on what, not how. |
| allowed-tools | ["AskUserQuestion","Read","Write","Task"] |
Transform Product Briefs into actionable PRDs that unblock engineers and UI designers. A PRD captures the "what" while leaving the "how" to implementation teams.
Point me at a Product Brief and I'll extract personas, scenarios, and metrics, then walk through five phases (Extraction → High-Level Requirements → Use Case Compendium → Milestone Definition → Review) to produce a PRD. If no path is provided, paste the Brief inline.
When invoked:
I'll help you build a Product Requirements Document from a Product Brief.
Please provide:
- A path to your Product Brief, OR
- Paste the Product Brief content directly
A PRD includes:
1. High-level requirements and design principles
2. Use case compendium (fine-grained requirements)
3. Milestone definitions
I'll guide you through extracting requirements from your north star scenarios.
When AskUserQuestion is unavailable (e.g., running as a sub-agent), do not stall waiting for user confirmation. Proceed using best-judgment defaults at each phase boundary and surface assumptions explicitly in the output (e.g., an "Assumptions" note above the section, or inline italics on chosen priorities). The caller will review the full document and request changes.
A complete PRD has three sections:
The workflow has five phases. The PRD document has three sections (above) — extraction and review are workflow phases, not document sections.
┌───────────┐ ┌───────────┐ ┌───────────┐ ┌───────────┐ ┌───────────┐
│Extraction │──▶│High-Level │──▶│ Use Case │──▶│ Milestone │──▶│ Review │
│ │ │ Reqs │ │ Compendium│ │ Definition│ │ │
│ Personas, │ │Principles │ │ Atomic │ │ MVP / MLP │ │ Coverage │
│ scenarios │ │& trade- │ │ reqs in │ │ scope for │ │ + polish │
│ & metrics │ │offs │ │ one table │ │ release │ │ │
└───────────┘ └───────────┘ └───────────┘ └───────────┘ └───────────┘
For full phase details, see workflow-phases.md. For section-level authoring guidance (formats, examples, prioritization strategy), see authoring-guide.md.
Phase 1 — Extraction: Read the Product Brief; identify personas, metrics, north star scenarios, and risks.
Phase 2 — High-Level Requirements: State design principles and priority trade-offs.
Phase 3 — Use Case Compendium: Break each north star scenario into atomic requirements; combine into one unified table.
/thinking atomic-thought to decompose scenarios systematically.Phase 4 — Milestone Definition: Choose one north star to define milestone 1; mark requirements 1.0 / 1.5 / blank.
/thinking tree-of-thoughts for north star selection and prioritization.Phase 5 — Review: Verify coverage, check quality, polish prose for human readers, save or hand off.
/thinking self-consistency to validate before finalizing.writing-for-humans skill. The Use Case Compendium table is preserved verbatim.Use these values in the Use Case Compendium:
Target distribution: 15–25% at 1.0, 10–15% at 1.5, 60–75% blank. If more than 40% have milestone values, you're listing, not prioritizing.
Bad: "Helpy uses a WebSocket connection to maintain chat state" Bad: "A React modal component displays the chat interface" Good: "Lisa can continue her support conversation across page navigation"
Bad: "Helpy is discoverable" Bad: "Users can get support" Good: "Lisa can easily find Helpy on the home page without first logging in"
Bad: "Helpy is on the home page" Good: "Lisa can easily find Helpy on the home page without first logging in when she's looking for tech support."
Bad: Requirements that don't trace to any scenario Good: Every requirement references its source scenario via the North Star column
PRD prioritization is consequential — wrong decisions waste engineering effort. Visible reasoning surfaces assumptions, documents trade-offs, and enables course-correction.
| Phase | Pattern | Why |
|---|---|---|
| Phase 3 — extracting requirements | /thinking atomic-thought | Decompose each scenario into independent requirement atoms |
| Phase 4 — choosing the north star | /thinking tree-of-thoughts | Compare trade-offs between candidate scenarios |
Phase 4 — assigning 1.0 vs 1.5 | /thinking tree-of-thoughts | Evaluate bang-for-buck across requirements |
| Phase 5 — final review | /thinking self-consistency | Validate coverage and prioritization decisions |
From: writing-product-briefs → writing-prds. Thesis informs high-level requirements; personas and scenarios carry over directly.
Uses: writing-prds invokes writing-for-humans in Phase 5 as a post-processing pass over prose sections. PRDs are read by engineers, designers, and stakeholders — readability matters. The Use Case Compendium table is excluded from the rewrite (structured content, not prose).
To: writing-prds → slicing-elephant-carpaccio (slice features into thin vertical increments), writing-agile-stories (each use case becomes a story candidate), or planning-tdd (use cases define what to build; milestones define scope).
Quick scaffold below. For per-section formatting rules and examples, see authoring-guide.md.
# Product Requirements Document: [Product Name]
## High-Level Requirements
- [Primary goal and constraints from Product Brief]
- [Persona/scenario priority for first milestone]
- [Key trade-off decisions]
- [Safety/guardrail stance]
## Use Case Compendium
| Requirement | Milestone | Persona | North Star |
|-------------|-----------|---------|------------|
| [North star requirement 1] | 1.0 | [Persona] | [North star scenario] |
| [North star requirement 2] | 1.0 | [Persona] | [North star scenario] |
| [North star stretch goal] | 1.5 | [Persona] | [North star scenario] |
| [Future requirement] | | [Persona] | [North star scenario] |
| [Other scenario requirement] | | [Persona] | [Other scenario] |
| ... | | ... | ... |
*Note: 60–75% of requirements should have blank milestone values.*
## Milestone Definitions
### Milestone 1: [Theme]
**Goal**: [What this proves]
**Persona focus**: [Primary persona]
**North star scenario**: [The scenario that defines this milestone]
**Scope**: [Capabilities included]
**Excluded**: [What's deferred and why]
CRITICAL: Only define Milestone 1. Do NOT create placeholder sections for future milestones (Milestone 2, 3, etc.) with "To be defined" or "Anticipated scope" content. Requirements for future work are already captured in the Use Case Compendium with blank milestone values — that's sufficient. Future milestone definitions happen when planning actually begins for those milestones.
Creates TDD implementation plans where tests ARE the plan. Specifies test specs and structural context per phase — implementation emerges during execution, not planning. Verification is fully automated; no manual verification step. Use when planning features test-first, creating TDD plans from design artifacts, or when the user asks for a test-driven implementation plan. NOT for executing a plan test-first (the Red-Green-Refactor cycle during implementation) — that is practicing-tdd.
Breaks an epic into an ordered backlog of vertical, sprint-sized user stories. Use when splitting an epic into stories, planning a feature's story backlog, or when a story feels too large but isn't ready for task breakdown. Produces 4-8 demoable stories, each independently shippable and grounded in user-visible behavior. NOT for sub-story implementation slicing (see slicing-elephant-carpaccio) or for writing one story's full BDD spec (see writing-agile-stories).
Slices a single feature or story into ultra-thin vertical increments (minutes-to-hours each) using Alistair Cockburn's Elephant Carpaccio methodology. Use during implementation planning when one story or feature is too large to build in a session and you need demoable increments across layers. Produces an ordered backlog of 10-20 thin slices, each independently working, testable, and demoable. NOT for splitting an epic into sprint-sized stories (see decomposing-epics).
Write well-scoped, verifiable development tasks for non-user-facing work — refactors, test work, and dependency/tooling changes. Use to "write a dev task", create a "refactor task", a "tech-debt ticket", a "test task", or a "dependency upgrade task"; to capture "non-user-facing work" or a "technical task"; or to write a "definition of done for a refactor". Produces an implementation-focused task with a verifiable Definition of Done. For user-facing behavior, use writing-agile-stories instead.
Expert guidance for creating, writing, and refining Claude Code Skills. Use when working with SKILL.md files, authoring new skills, improving existing skills, or understanding skill structure and best practices.
Examines a git-machete stack of dependent branches and drafts preemptive PR comments warning reviewers that code in an earlier PR is already refactored, renamed, moved, deleted, or fixed later in the stack. Output is DRAFT ONLY — the user pastes it manually; the skill never posts to GitHub. Use for stacked PRs, a git-machete stack, when reviewers comment on code that is "already fixed/refactored/handled later in the stack", to preempt reviewer comments on superseded code, to draft PR comments for a branch stack, or to find code in an earlier branch that a later branch supersedes.