Product Management Skill

AI-native product management for startups. Transform Claude into an expert PM that processes signals, not just feature lists.

Core Philosophy

WINNING = Pain × Timing × Execution Capability

Filter aggressively from 50 gaps to 3-5 high-conviction priorities. Expert PMs track signals with confidence scores, timestamps, and velocity.

Commands Quick Reference

Command	Purpose
/pm:analyze	Scan codebase + interview for product inventory
/pm:landscape	Research competitor landscape
/pm:gaps	Run gap analysis with WINNING filter
/pm:file	Batch create GitHub Issues for approved gaps
/pm:prd	Generate PRD and create GitHub Issue
/pm:sync	Sync local cache with GitHub Issues

Agents

This plugin provides specialized agents for autonomous tasks:

Agent	Triggers On	Purpose
research-agent	"research [competitor]", "scout [name]"	Deep autonomous web research
gap-analyst	"find gaps", "what should we build"	Systematic gap identification with scoring
prd-generator	"create PRD for [feature]"	Generate PRD + create GitHub Issue

Data Storage

All data stored in .pm/ folder at project root:

.pm/
├── config.md                 # Positioning, scoring weights
├── product/                  # Product inventory, architecture
├── competitors/              # Competitor profiles
├── gaps/                     # Gap analyses with scores
├── requests/                 # Synced GitHub Issues (for dedup)
├── prds/                     # Generated PRDs
└── cache/last-updated.json   # Staleness tracking

See references/data-structure.md for complete file templates.

WINNING Filter Scoring

Hybrid scoring approach - Claude suggests researchable criteria, user scores domain-specific:

Criterion	Scorer	Source
Pain Intensity (1-10)	Claude	Review sentiment, support data
Market Timing (1-10)	Claude	Search trends, competitor velocity
Execution Capability (1-10)	User	Architecture fit, team skills
Strategic Fit (1-10)	User	Positioning alignment
Revenue Potential (1-10)	User	Conversion/retention impact
Competitive Moat (1-10)	User	Defensibility once built

Total: X/60 → Recommendation:

• 40+ → FILE (high conviction)
• 25-39 → WAIT (monitor)
• <25 → SKIP (not worth it)

See references/winning-filter.md for detailed scoring criteria.

Deduplication & Sync

Prevent duplicate feature requests by syncing with GitHub Issues:

On Session Start

1. Check .pm/cache/last-updated.json for staleness
2. If >24 hours since last sync, prompt for /pm:sync

/pm:sync Process

1. Fetch all GitHub Issues with pm:* labels via gh issue list --json
2. Update .pm/requests/[issue-number].md for each issue
3. Update last-updated.json timestamp

Deduplication During Gap Analysis

1. Load existing issues from .pm/requests/
2. For each new gap, fuzzy match against existing:
   • Title similarity (Levenshtein): 40% weight
   • Keyword overlap: 30% weight
   • Label match: 20% weight
   • Description similarity: 10% weight
3. Mark gaps as:
   • EXISTING (>80% match) → Show linked issue
   • SIMILAR (50-80%) → Warn, ask user
   • NEW (<50%) → Proceed normally

Output Format

| Gap | WINNING | Status | Match |
|-----|---------|--------|-------|
| OAuth support | 47/60 | EXISTING | #42 (95%) |
| Dark mode | 38/60 | NEW | - |

Flow Summaries

Product Analysis (/pm:analyze)

1. Scan codebase for features (routes, components, APIs, models)
2. Interview user for business context
3. Generate inventory with technical moats and debt flags
4. Save to .pm/product/

Competitive Intelligence (/pm:landscape, /pm:scout [name])

1. Research competitors via WebFetch/WebSearch
2. Categorize features: Tablestakes, Differentiators, Emerging, Deprecated
3. Save profiles to .pm/competitors/

Gap Analysis (/pm:gaps)

1. Load product inventory + competitor profiles
2. Check staleness (>30 days → prompt refresh)
3. Sync with GitHub Issues for deduplication
4. Identify all gaps, score with WINNING filter
5. Mark as NEW/EXISTING/SIMILAR
6. Save to .pm/gaps/[date]-analysis.md

Feature Filing (/pm:review, /pm:file)

1. /pm:review - Walk through gaps, decide FILE/WAIT/SKIP
2. /pm:file - Create GitHub Issues for approved gaps (skips duplicates)
3. Apply labels: pm:feature-request, winning:*, priority:*

PRD Generation (/pm:prd [feature])

1. Load feature context from gap analysis or GitHub Issue
2. Generate PRD: Problem, User Stories, Acceptance Criteria, etc.
3. Save to .pm/prds/[slug].md
4. Create GitHub Issue with PRD content as feature request

Backlog & Roadmap (/pm:backlog, /pm:roadmap)

1. Fetch open issues with pm: labels
2. Sort by WINNING score or RICE
3. Organize into Now/Next/Later priorities

GitHub Integration

Labels (Auto-Created)

pm:feature-request    pm:gap-identified    pm:competitor-intel
priority:now          priority:next        priority:later
winning:high (40+)    winning:medium (25-39)    winning:low (<25)

Prerequisites

• GitHub CLI (gh) installed and authenticated
• Run gh auth status to verify

See references/github-labels.md for label definitions. See references/issue-template.md for issue format.

Integration with spec-kit

This plugin handles WHAT to build and WHY (product discovery). For HOW to build it, use spec-kit:

PM Plugin → GitHub Issue → spec-kit
/pm:file     Creates issue   /speckit.specify
/pm:prd      Creates issue   /speckit.plan → /speckit.implement

The GitHub Issue IS the handoff—no separate command needed.

Staleness Handling

PM data ages. Handle proactively:

• Competitor data >30 days: Prompt refresh before gap analysis
• Gap analysis >14 days: Warn when viewing backlog
• GitHub sync >24 hours: Suggest /pm:sync on session start

⚠️ Competitor data is 45 days old. Run `/pm:landscape` to refresh.

Additional Resources

Reference Files

• references/winning-filter.md - Detailed WINNING scoring criteria
• references/github-labels.md - Label definitions and colors
• references/issue-template.md - GitHub Issue template
• references/data-structure.md - Complete .pm/ folder structure

Example Files

• examples/gap-analysis.md - Sample gap analysis output
• examples/competitor-profile.md - Sample competitor profile