Product Management Skill for Claude Code

繁體中文

AI-era Product Manager expert skill for Claude Code, integrating classic methodologies with AI collaboration workflows.

Features

• Classic Methodologies: Marty Cagan (Problem-first), Teresa Torres (Continuous Discovery), Clayton Christensen (JTBD)
• AI Collaboration Workflow: Define boundaries → AI generates draft → PM reviews → Perspective switching
• Practical Prompt Library: Ready-to-use prompts for Edge Cases, Sequence Diagrams, Risk Assessment
• Perspective Switching: Auto-generate Engineer version / Executive version of PRDs

Installation

# Copy to Claude Code skills directory
git clone https://github.com/panda/product-management-skill.git
cp -r product-management-skill ~/.claude/skills/product-management

Or use symlink (easier to update):

git clone https://github.com/panda/product-management-skill.git
ln -s $(pwd)/product-management-skill ~/.claude/skills/product-management

Usage

Method	Example
Slash command	/product-management
Keyword trigger	"Help me write a PRD", "Analyze this feature request"
Natural language	"How should I evaluate this feature?", "Plan next quarter's roadmap"

Use Cases

1. Write a PRD with AI Collaboration

You: I need a PRD for a user notification system

Claude: [Loads skill, asks for context/boundaries]
        → Generates draft with edge cases
        → You review and refine
        → Output: Full PRD with JSON Schema + Sequence Diagram

2. Analyze Feature Request

You: Should we add dark mode? Here's the user feedback...

Claude: [Applies RICE scoring]
        → Impact vs Effort analysis
        → Risk assessment
        → Recommendation with rationale

3. Perspective Switching

You: Convert this PRD for my engineering team

Claude: → Strips business context
        → Adds Mermaid sequence diagram
        → Outputs JSON Schema for API
        → Includes sample request/response

You: I need an executive summary for my CEO

Claude: → Risk Assessment Matrix
        → Business impact analysis
        → Decision points highlighted
        → Cost/benefit summary

4. Edge Case Discovery

You: What edge cases should I consider for this payment flow?

Claude: → Lists 5+ edge cases
        → Bank API timeout handling
        → Partial payment scenarios
        → Currency conversion errors
        → Refund edge cases

5. Consistency Check

You: Review this PRD for consistency issues

Claude: → Scans for terminology conflicts
        → "Line 12 says 'ETH', but line 45 uses 'Ethereum' - standardize?"
        → "Amount unit inconsistent: wei (line 23) vs ETH (line 67)"
        → "Role naming: 'user' vs 'customer' vs '用戶' - pick one"
        → Outputs checklist of fixes needed

6. Decision Risk Assessment

You: What risks should I flag for stakeholders?

Claude: → Cost scaling risk
        "Will costs grow non-linearly with usage?
         Current: $0.01/tx → At 100x volume: $2.50/tx (not $1.00)"

        → Fallback impact analysis
        "If payment API fails: UX degradation or revenue loss?
         Recommendation: This affects revenue - needs P0 SLA"

        → Responsibility attribution
        "If AI agent makes wrong decision, who is accountable?
         Flag for legal review before launch"

How It Works

┌─────────────────────────────────────────────────────┐
│  1. You provide: Context, Goals, Boundaries         │
└─────────────────────┬───────────────────────────────┘
                      ▼
┌─────────────────────────────────────────────────────┐
│  2. AI generates: Draft PRD with edge cases         │
└─────────────────────┬───────────────────────────────┘
                      ▼
┌─────────────────────────────────────────────────────┐
│  3. You review: Logic, scenarios, consistency       │
└─────────────────────┬───────────────────────────────┘
                      ▼
┌─────────────────────────────────────────────────────┐
│  4. Iterate: Refine until complete                  │
└─────────────────────┬───────────────────────────────┘
                      ▼
┌─────────────────────────────────────────────────────┐
│  5. Transform: Engineer view / Executive view       │
└─────────────────────────────────────────────────────┘

Structure

product-management-skill/
├── SKILL.md                     # Main skill definition
└── references/
    └── ai-prd-workflow.md       # AI PRD workflow + Prompt library

Core Capabilities

Capability	Description
PRD Writing	Quick PRD, Full PRD, Discovery Checklist
AI Collaboration	Define boundaries → Generate → Review → Transform
Feature Analysis	Impact vs Effort Matrix, RICE Scoring
Prompt Library	Edge Cases, Sequence Diagrams, Risk Assessment

Methodology Sources

Expert	Framework
Marty Cagan	Problem space vs Solution space
Teresa Torres	Continuous Discovery, Opportunity Trees
Clayton Christensen	Jobs to be Done

Credits

AI PRD Workflow: Inspired by Darren's article on AI-augmented product management - the core concepts of "define boundaries → AI generates → PM reviews → perspective switching" and decision risk exposure.

Classic Methodologies: Supplemented with established PM frameworks:

• Marty Cagan (SVPG): Problem space vs Solution space, Discovery vs Delivery
• Teresa Torres: Continuous Discovery, Opportunity Solution Trees
• Clayton Christensen: Jobs to be Done

License

MIT