Flywheel PM Marketplace

A Claude Code plugin marketplace for product managers.

Available Plugins

Plugin	Description	Version
flywheel-pm	Small pushes, unstoppable momentum — 32 commands, 13 agents, 13 skills. Your PM operating system that gets sharper with every product cycle	2.2.0

What's New in v2.0

v1.0 introduced artifact compounding — specs, decisions, and learnings get better with every cycle. v2.0 adds a second layer: judgment compounding — simulation accuracy, bias awareness, and segmentation instincts get sharper the more you use them.

Two layers of compounding:

1. Artifact compounding (v1.0): Your specs, decisions, and learnings improve because the system remembers what worked and what didn't.
2. Judgment compounding (v2.0): Your simulation accuracy, bias awareness, and segmentation instincts sharpen because the system trains your product intuition.

v2.0 adds 15 new commands, 5 new skills, and 4 new agents across skill development, strategy, and growth.

Installation

# Add the marketplace
claude plugin marketplace add https://github.com/abhitsian/flywheel-pm-marketplace

# Install the plugin
claude plugin install flywheel-pm

Restart Claude Code after installation.

Quick Start

/pm:welcome     # Guided first-run experience
/pm:onboard     # Create your PM profile
/pm:context     # Load your work environment

Then run the core loop on something real:

/pm:opportunity  # Frame the problem
/pm:solution     # Design solutions
/pm:spec         # Write the spec
/pm:review       # 5 agents review in parallel
/pm:measure      # Define predictions
/pm:compound     # Capture learnings (never skip this)

Build judgment with the v2.0 skill-development commands:

/pm:teardown     # Tear down a product to train simulation muscle
/pm:simulate     # Simulate user reactions before shipping
/pm:bias-check   # Check your decisions for cognitive biases

How It Works

Flywheel PM encodes your PM operating system into three knowledge layers:

• pm-profile.yaml — How you think. Your frameworks, vocabulary, and past patterns. Auto-enriched by /pm:compound after each cycle.
• team-profile.yaml — What the team knows collectively. Built via /pm:share.
• work-context.yaml — What you're working within. Company OKRs, team capacity, customer signals, competitive landscape.

Every command loads all three layers to produce personalized, context-aware output.

Observe → Opportunity → Solution → Spec → Review → Measure → Compound
   ^                                                              |
   └──────────────── learnings feed back ──────────────────────────┘

Teardown → Simulate → Bias-Check → Compound
   ^                                    |
   └──────── judgment sharpens ─────────┘

Cycle 1: Generic frameworks. Predictions are guesses. Cycle 10: System knows your estimation bias, flags similar past failures, pre-fills metrics by project type. Cycle 20: Your second brain knows more about your PM operating system than you consciously remember.

All Commands (31)

Setup

Command	Purpose
/pm:welcome	Guided first-run experience
/pm:onboard	Create your PM profile from docs or interactive Q&A
/pm:context	Set up your operating environment

Research & Discovery

Command	Purpose
/pm:research	Ingest signals from support, NPS, feature requests, sales, usage data
/pm:interview	Prepare guides, capture notes, synthesize patterns

Core Product Loop

Command	Purpose
/pm:opportunity	Signal -> problem -> business context -> prioritized opportunity
/pm:solution	20 creativity techniques -> risk/impact matrix -> classification
/pm:spec	Vision narrative -> journey map -> metrics -> launch criteria
/pm:review	5 specialized agents review your spec in parallel
/pm:measure	Predictions framework with activation funnel and continue/pause/pivot criteria
/pm:compound	Compare predictions vs actuals, update profile, share with team

Decision-Making & Communication

Command	Purpose
/pm:decide	Frame decisions — options, criteria, pre-mortem, reversibility
/pm:buyin	Stakeholder-specific narratives using 6 buy-in techniques
/pm:product-review	Weekly reviews for team, leadership, or all-hands
/pm:share	Promote a learning to the team knowledge base

Skill Development (v2.0)

Command	Purpose
/pm:teardown	Guided product teardown that trains simulation muscle
/pm:simulate	Simulate user reactions before shipping
/pm:bias-check	Check decisions for cognitive biases
/pm:friction-log	Structured friction logging
/pm:wayrtd	"What Are You Really Trying To Do?" — peel back feature requests to find the real need
/pm:motivation-map	Map the adoption equation (motivation, ability, triggers)
/pm:ux-psychology	Apply behavioral design principles to user flows

Strategy (v2.0)

Command	Purpose
/pm:segment	Needs-based user segmentation
/pm:differentiate	Competitive differentiation using the 7 Powers framework
/pm:strategy	Full strategy document

Growth & Metrics (v2.0)

Command	Purpose
/pm:growth-model	Growth loop design
/pm:retention	Retention analysis
/pm:north-star	North Star Metric design
/pm:goals	Team goal setting derived from NSM
/pm:tradeoff	Structured tradeoff analysis

Shortcut

Command	Purpose
/pm:lfg	Full pipeline: opportunity -> solution -> spec -> review -> measure -> buyin

Agents (13)

Review Agents (run in parallel during /pm:review)

Agent	Focus
customer-advocate	Does this solve the real customer problem?
metrics-reviewer	Are metrics measurable? Are thresholds specific?
business-case-reviewer	Is revenue quantified with methodology?
scope-creep-detector	One problem well or three problems poorly?
past-pattern-checker	Have we seen this before? What did we learn?

Research Agents

Agent	Focus
opportunity-researcher	Search past work for similar opportunities
pm-profile-analyzer	Extract frameworks from PM docs during onboarding
research-synthesizer	Find patterns across multiple interviews
competitive-landscape	Research competitive positioning and market context

Judgment Agents (v2.0)

Agent	Focus
simulation-coach	Reviews teardowns for blind spots and simulation gaps
bias-detector	Detects cognitive biases in documents and decisions
strategy-reviewer	Reviews strategy docs for coherence and completeness
growth-analyst	Analyzes growth models for loop integrity and retention risks

Skills (11)

Core Skills

Skill	Expertise
pm-operating-system	Loads PM profile and adapts every command
metrics-design	Metrics hierarchies, activation funnels, engagement dimensions
opportunity-evaluation	5-question framework, business context, prioritization
solution-creativity	20 systematic techniques for solution generation
vision-narrative	3-part aspirational narratives
stakeholder-buyin	6 techniques: framing, social proof, goal seek, inception, citation, narration

v2.0 Skills

Skill	Expertise
simulation	Cognitive empathy engine — predict user reactions with increasing accuracy
behavioral-design	Motivation theory and UX psychology for product flows
strategy-craft	Needs-based segmentation and competitive differentiation
growth-systems	Growth loop design, retention analysis, and flywheel mechanics
analytical-thinking	North Star Metrics, goal decomposition, and structured tradeoffs

Documentation

• Plugin README — Full plugin documentation
• Onboarding Guide — 5-day guided setup
• Cheat Sheet — Quick reference
• Sample Profile — Example PM profile

License

MIT