pm-pmm

# PM/PMM Skill

> **Version:** 1.0.0
> **Framework:** Jerry PM/PMM Decision Framework
> **Constitutional Compliance:** Jerry Constitution v1.0
> **SSOT Reference:** `.context/rules/quality-enforcement.md`
> **Architecture Reference:** PROJ-018 Issue #123

## Document Audience (Triple-Lens)

| Level | Audience | Sections to Focus On |
|-------|----------|---------------------|
| **L0 (ELI5)** | New users, stakeholders | [Purpose](#purpose), [When to Use](#when-to-use-this-skill), [Quick Reference](#quick-reference) |
| **L1 (Engineer)** | Developers invoking agents | [Available Agents](#available-agents), [Invoking an Agent](#invoking-an-agent), [Discovery vs Delivery Mode](#discovery-vs-delivery-mode), [Framework Catalog](#framework-catalog) |
| **L2 (Architect)** | Workflow designers | [P-003 Compliance](#p-003-compliance), [Cross-Agent Data Flow](#cross-agent-data-flow), [Integration Points](#integration-points-with-other-skills), [Artifact Ownership Matrix](#artifact-ownership-matrix) |

---

## Purpose

The PM/PMM skill provides a **decision-focused product management and product marketing capability** for the Jerry Framework. It helps PMs and PMMs make better product decisions by ensuring every decision is grounded in evidence, structured by proven frameworks, and documented as a living artifact.

### What It Does

- **Product Strategy:** Creates PRDs, product vision documents, roadmaps, and use case specifications grounded in JTBD, RICE, Kano, Playing to Win, and Opportunity Solution Tree frameworks
- **Customer Insight:** Builds JTBD-oriented personas, customer journey maps with Moments of Truth, and VOC research reports synthesized from interview data
- **Business Analysis:** Produces business cases, TAM/SAM/SOM market sizing, and pricing strategy analysis using Van Westendorp, Lean Canvas, and SaaS financial metrics
- **Competitive Intelligence:** Delivers competitive analysis with Porter's Five Forces, Blue Ocean value curves, battle cards with talk tracks, and win/loss pattern analysis
- **Go-to-Market:** Creates GTM plans with Dunford positioning, MRDs, and buying-committee buyer personas

### Core Design Principles

1. **Cagan's Two Risks:** Organized around Value Risk ("Will they buy/use it?") and Business Viability Risk ("Does it work for the business?"). Usability Risk and Feasibility Risk are out of scope.
2. **Discovery Before Delivery:** Every agent defaults to discovery mode. Validates assumptions cheaply before committing to polished artifacts.
3. **18 Validated Frameworks:** Not name-drops -- each framework is operationalized with methodology steps that produce canonical output structures.
4. **Evidence, Not Intuition:** All claims require evidence citations or explicit hypothesis marking with confidence levels.

---

## When to Use This Skill

**Activate when:**

- Building a new product or feature and need to define what to build and why
- Creating or updating product requirements (PRDs) with prioritization frameworks
- Understanding customers through personas, journey maps, or VOC research
- Assessing business viability through business cases, market sizing, or pricing analysis
- Analyzing the competitive landscape or creating battle cards
- Planning go-to-market strategy, positioning, messaging, or launch execution
- Needing structured product/market frameworks rather than ad-hoc analysis

**Do NOT use when:**

- Writing code, designing architecture, or making technical implementation decisions (use `/architecture` or `/eng-team`)
- Running adversarial quality reviews on deliverables (use `/adversary`)
- Performing root cause analysis on bugs or failures (use `/problem-solving`)
- Creating UX wireframes or design specifications (out of scope -- Usability Risk)
- Managing work items or tracking project status (use `/worktracker`)
- Parsing transcripts or meeting notes (use `/transcript`)
- Deploying, testing, or managing CI/CD (use `/eng-team`)

### Negative Keywords (Prevent False Routing)

These keywords in a request should suppress `/pm-pmm` routing:

| Keyword | Routes To Instead |
|---------|-------------------|
| code review | `/problem-solving` or `/eng-team` |
| architecture, ADR | `/architecture` |
| engineering, implementation | `/eng-team` |
| deployment, CI/CD | `/eng-team` |
| testing, test coverage | `/eng-team` |
| infrastructure pricing, cloud pricing | `/eng-team` (not product pricing) |
| adversarial, tournament | `/adversary` |
| transcript, VTT, SRT | `/transcript` |
| penetration test, exploit | `/red-team` |

---

## Available Agents

| Agent | Decision Question | Risk Domain | Model | Status |
|-------|------------------|-------------|-------|--------|
| `pm-product-strategist` | "What should we build, and why?" | Value + Viability (strategy) | opus | **Tier 1 -- Active** |
| `pm-customer-insight` | "Who are our customers, and what do they need?" | Value Risk | opus | **Tier 1 -- Active** |
| `pm-market-strategist` | "How do we bring this to market?" | Viability Risk (GTM) | opus | **Tier 1 -- Active** |
| `pm-business-analyst` | "Is this worth investing in?" | Viability Risk (financial) | sonnet | **Tier 2 -- Active** |
| `pm-competitive-analyst` | "Who are we up against?" | Viability Risk (market) | sonnet | **Tier 2 -- Active** |

### Agent-to-Artifact Ownership

| Agent | Primary Artifacts | Count |
|-------|-------------------|-------|
| pm-product-strategist | PRD, Product Vision, Roadmap, Use Cases | 4 |
| pm-customer-insight | User Personas, Customer Journey Maps, VOC Research Reports | 3 |
| pm-market-strategist | GTM Plan, MRD, Buyer Personas | 3 |
| pm-business-analyst | Business Case, Market Sizing (TAM/SAM/SOM) | 2 |
| pm-competitive-analyst | Competitive Analysis, Battle Cards, Win/Loss Analysis | 3 |
| **Total** | | **15** |

### Agent Selection Hints

| User Says | Route To | Rationale |
|-----------|----------|-----------|
| "Write a PRD" / "product requirements" / "roadmap" / "prioritize features" | pm-product-strategist | Owns all product requirements and prioritization artifacts |
| "Create personas" / "journey map" / "customer interviews" / "VOC" / "pain points" | pm-customer-insight | Owns user personas, journey maps, VOC research |
| "GTM plan" / "positioning" / "messaging" / "buyer persona" / "launch plan" | pm-market-strategist | Owns GTM, MRD, buyer personas |
| "Business case" / "TAM" / "market sizing" / "pricing model" / "unit economics" | pm-business-analyst | Owns business cases, market sizing, pricing analysis |
| "Competitive analysis" / "battle card" / "win/loss" / "Porter's" / "SWOT" | pm-competitive-analyst | Owns competitive analysis, battle cards, win/loss reports |

---

## P-003 Compliance

All PM/PMM agents are **workers**, NOT orchestrators. The MAIN CONTEXT (Claude session) orchestrates all workflows.

```
P-003 AGENT HIERARCHY:
======================

  +-------------------+
  | MAIN CONTEXT      |  <-- Orchestrator (Claude session)
  | (orchestrator)    |
  +-------------------+
     |        |        |
     v        v        v
  +------+ +------+ +------+
  | pm-  | | pm-  | | pm-  |   <-- Tier 1 Workers
  |prod- | |cust- | |mkt-  |
  |strat | |insght| |strat |
  +------+ +------+ +------+

  +------+ +------+
  | pm-  | | pm-  |   <-- Tier 2 Workers (Active)
  |biz-  | |comp- |
  |anlst | |anlst |
  +------+ +------+

  - Agents CANNOT invoke other agents.
  - Agents CANNOT spawn subagents.
  - Agents do NOT have the Task tool.
  - Only MAIN CONTEXT orchestrates the sequence.
  - All cross-agent data flows through file artifacts.
```

---

## Invoking an Agent

### Option 1: Natural Language Request

Simply describe what you need:

```
"Create a PRD for the self-service onboarding feature"
"Build personas for our platform engineering customers"
"Create a GTM plan with positioning for our developer platform"
"Help me prioritize these features using RICE"
"Map the customer journey for enterprise onboarding"
"Write an MRD for the APAC market expansion"
```

The orchestrator selects the appropriate agent based on keywords and context.

### Option 2: Explicit Agent Request

Request a specific agent:

```
"Use pm-product-strategist to create a product vision using Playing to Win"
"Have pm-customer-insight build a persona for DevOps engineers"
"I need pm-market-strategist to create Dunford positioning for our launch"
```

### Option 3: Multi-Agent Workflow

Request work that spans multiple agents (orchestrated by main context):

```
"Research our customers, then create a PRD grounded in the persona data"
  -> pm-customer-insight (personas) -> pm-product-strategist (PRD with persona cross-refs)

"Position our product and then create a GTM launch plan"
  -> pm-market-strategist (positioning hypothesis) -> pm-market-strategist (GTM plan)

"Analyze the competitive landscape and build a business case for our platform"
  -> pm-competitive-analyst (competitive analysis) -> pm-business-analyst (business case with competitive context)

"Size the market, analyze competitors, and create pricing recommendations"
  -> pm-business-analyst (TAM/SAM/SOM) -> pm-competitive-analyst (competitive pricing) -> pm-business-analyst (pricing strategy)
```

---

## Discovery vs Delivery Mode

Every agent operates in two modes. **Discovery is always the default.**

| Dimension | Discovery Mode | Delivery Mode |
|-----------|---------------|---------------|
| **Purpose** | Explore, hypothesize, validate assumptions | Produce stakeholder-ready artifacts |
| **Output length** | 1-2 pages | Full framework depth (5-20 pages) |
| **Framework depth** | Lightweight application, key dimensions only | Complete framework execution with all sections |
| **Evidence standard** | Hypotheses with stated confidence | Data-validated claims with citations |
| **Audience** | PM/PMM internal working document | Cross-functional stakeholders, executives |
| **Status** | `discovery` | `delivery` or `final` |
| **Default?** | Yes -- always start here | Explicit user request required |

### Mode Selection

```
IF user explicitly says "delivery" or "full" or "stakeholder-ready":
    mode = delivery
ELIF prior discovery artifact exists for this topic:
    mode = delivery (suggest upgrade, confirm with user per P-020)
ELSE:
    mode = discovery (default)
```

### Why Discovery First?

Discovery before delivery prevents the most expensive PM failure mode: building a polished artifact for a product nobody wants. Discovery validates assumptions cheaply (1-2 pages, hypothesis-driven). Delivery commits resources to documentation (5-20 pages, evidence-validated).

---

## Framework Catalog

18 primary frameworks mapped to agents. Each framework is operationalized with methodology steps in the agent's `<methodology>` section.

### pm-product-strategist Frameworks (6)

| # | Framework | Operationalization |
|---|-----------|-------------------|
| 1 | Opportunity Solution Trees (Torres) | Tree structure: outcome root, opportunity branches, solution leaves with evidence links and confidence |
| 2 | JTBD (Christensen/Ulwick) -- secondary | Decomposes needs into functional/emotional/social jobs; Opportunity Scoring |
| 3 | RICE Prioritization (Intercom) | Scores: Reach, Impact, Confidence, Effort; dimension-level breakdown table |
| 4 | Kano Model | Feature classification: Must-Have, Performance, Delighter; questionnaire pairs |
| 5 | Product Kata (Perri) | Cycle: Direction -> Current State -> Next Target Condition -> First Step |
| 6 | Playing to Win (Lafley & Martin) | Strategy cascade: Aspiration, Where to Play, How to Win, Capabilities, Systems |

**Supporting methods:** ICE/MoSCoW/WSJF (alternative prioritization), North Star Metric (vision anchoring), Story Mapping (use case organization)

### pm-customer-insight Frameworks (4)

| # | Framework | Operationalization |
|---|-----------|-------------------|
| 7 | JTBD (Christensen/Ulwick) -- primary | Functional/emotional/social jobs with importance/satisfaction scoring |
| 8 | Customer Development (Blank) | Four phases: Discovery, Validation, Creation, Building with evidence |
| 9 | Moments of Truth (P&G/Google) | Maps ZMOT, FMOT, SMOT, UMOT at each journey stage |
| 10 | Service Blueprint (Shostack) | Five-lane blueprint: physical evidence, customer/frontstage/backstage/support |

**Supporting methods:** NPS/CSAT/CES measurement, Opportunity Scoring (Ulwick/ODI)

### pm-market-strategist Frameworks (3)

| # | Framework | Operationalization |
|---|-----------|-------------------|
| 11 | Positioning Framework (Dunford) | 5-step: competitive alternatives, attributes, value, segment, category |
| 12 | PMF Survey (Ellis) | "Very disappointed" test design, 40% threshold, segment analysis |
| 13 | Lauchengco PMM Model | 4-role mapping: Ambassador, Strategist, Storyteller, Evangelist |

**Supporting methods:** Crossing the Chasm (adoption lifecycle, bowling alley), StoryBrand (messaging narrative)

### pm-business-analyst Frameworks (3)

| # | Framework | Operationalization |
|---|-----------|-------------------|
| 14 | Van Westendorp PSM | Four price-point questions, price sensitivity curves |
| 15 | Lean Canvas / BMC | 9-block canvas: Problem, Solution, Metrics, UVP, Channels, etc. |
| 16 | SaaS Financial Metrics | Rule of 40, LTV:CAC, NRR, Magic Number with BVP benchmarks |

**Supporting methods:** Good-Better-Best pricing, Conjoint analysis, NPV/IRR/break-even

### pm-competitive-analyst Frameworks (3)

| # | Framework | Operationalization |
|---|-----------|-------------------|
| 17 | Porter's Five Forces | All 5 forces assessed with high/medium/low rating and evidence |
| 18 | Blue Ocean / Value Curve | Value curve with eliminate-reduce-raise-create actions |
| -- | Crossing the Chasm (Moore) -- shared | Technology Adoption Lifecycle position, bowling alley targeting |

**Supporting methods:** SWOT Analysis, Gartner MQ / Forrester Wave, Category Design

---

## Artifact Ownership Matrix

Every artifact has exactly one primary owner. Contributing agents provide inputs but do not produce the artifact.

| # | Artifact | Primary Agent | Contributors | Sensitivity Default |
|---|----------|--------------|-------------|-------------------|
| 1 | PRD | pm-product-strategist | pm-customer-insight | internal |
| 2 | Product Vision | pm-product-strategist | pm-business-analyst | internal |
| 3 | Roadmap | pm-product-strategist | pm-business-analyst, pm-customer-insight | internal |
| 4 | Use Cases | pm-product-strategist | pm-customer-insight | internal |
| 5 | User Personas | pm-customer-insight | -- | confidential |
| 6 | Journey Maps | pm-customer-insight | -- | confidential |
| 7 | VOC Reports | pm-customer-insight | -- | confidential |
| 8 | Business Case | pm-business-analyst | pm-product-strategist, pm-competitive-analyst | restricted |
| 9 | Market Sizing | pm-business-analyst | pm-competitive-analyst | restricted |
| 10 | Competitive Analysis | pm-competitive-analyst | pm-market-strategist | restricted |
| 11 | Battle Cards | pm-competitive-analyst | pm-market-strategist | restricted |
| 12 | Win/Loss Analysis | pm-competitive-analyst | pm-market-strategist | restricted |
| 13 | GTM Plan | pm-market-strategist | pm-competitive-analyst, pm-customer-insight | internal |
| 14 | MRD | pm-market-strategist | pm-competitive-analyst, pm-product-strategist | internal |
| 15 | Buyer Personas | pm-market-strategist | pm-competitive-analyst | internal |

---

## Cross-Agent Data Flow

All agent interactions are mediated by the Jerry main context (orchestrator). No agent invokes another agent directly. Data flows through file artifacts.

| From | To | Data | Mechanism |
|------|-----|------|-----------|
| pm-customer-insight | pm-product-strategist | Persona file paths, VOC themes, JTBD statements | File paths in handoff artifacts array |
| pm-customer-insight | pm-market-strategist | User persona references for buyer-user alignment | cross_refs frontmatter |
| pm-competitive-analyst | pm-market-strategist | Competitive positioning, battle card references | File paths in handoff |
| pm-competitive-analyst | pm-business-analyst | Competitive pricing data, market share estimates | Orchestrator passes file paths in handoff |
| pm-business-analyst | pm-product-strategist | Market sizing, feasibility verdict | cross_refs frontmatter |
| pm-business-analyst | pm-market-strategist | Pricing model, packaging recommendations | Orchestrator passes file paths in handoff |
| pm-product-strategist | pm-market-strategist | Product strategy, feature differentiation | cross_refs frontmatter |
| pm-product-strategist | pm-business-analyst | Product scope, investment estimation inputs | Orchestrator passes file paths in handoff |

### Conflict Resolution

When agents produce conflicting recommendations:
1. Orchestrator surfaces both outputs with explicit conflict statement
2. Presents the evidence from each agent
3. Asks the user to decide (P-020: User Decides)
4. Does NOT silently pick a winner

---

## Integration Points with Other Skills

| # | Skill | Integration | Details |
|---|-------|-------------|---------|
| 1 | `/worktracker` | Bidirectional | Artifact IDs (PM-PS-001) link to worktracker entities (FEAT-042) via cross_refs. Roadmap items map to Epics/Features/Stories. |
| 2 | `/adversary` | Unidirectional | PM/PMM artifacts at C2+ submitted for quality review using 6-dimension weighted composite. Minimum 3-iteration creator-critic-revision cycle per H-14. |
| 3 | `/problem-solving` | Unidirectional | ps-researcher feeds pm-product-strategist and pm-competitive-analyst with external research. ps-analyst supports pm-business-analyst with root cause analysis. |
| 4 | `/architecture` | Bidirectional | ADR decisions inform PRD technical constraints. Product requirements referenced in ADR context. |
| 5 | `/nasa-se` | Unidirectional | PRD requirements flow to SE verification. Requirements IDs appear in V&V traceability matrix. |
| 6 | `/use-case` | Unidirectional | pm-product-strategist use cases feed to `/use-case` for slicing and implementation planning. |

---

## Quick Reference

### Common Workflows

| Need | Agent | Example Prompt |
|------|-------|---------------|
| Write a PRD | pm-product-strategist | "Create a PRD for the self-service onboarding feature using RICE prioritization" |
| Create product vision | pm-product-strategist | "Build a product vision using Playing to Win for our developer platform" |
| Build personas | pm-customer-insight | "Create JTBD-oriented personas for our platform engineering customers" |
| Map customer journey | pm-customer-insight | "Map the enterprise onboarding journey with Moments of Truth" |
| Synthesize VOC | pm-customer-insight | "Analyze these 12 interview transcripts and extract VOC themes" |
| Position product | pm-market-strategist | "Create Dunford positioning for our developer platform launch" |
| Plan GTM | pm-market-strategist | "Design a go-to-market plan for APAC expansion" |
| Create buyer personas | pm-market-strategist | "Build buyer personas for the enterprise buying committee" |
| Prioritize backlog | pm-product-strategist | "Prioritize these 8 features using RICE scoring" |
| Assess PMF | pm-market-strategist | "Design a PMF survey using the Ellis 40% test" |
| Build business case | pm-business-analyst | "Create a business case for the self-service platform with NPV analysis" |
| Size market | pm-business-analyst | "Calculate TAM/SAM/SOM for the developer platform market" |
| Analyze pricing | pm-business-analyst | "Run Van Westendorp pricing analysis for our enterprise tier" |
| Analyze competitors | pm-competitive-analyst | "Analyze the competitive landscape using Porter's Five Forces" |
| Create battle cards | pm-competitive-analyst | "Build battle cards for our top 3 competitors with talk tracks" |
| Analyze win/loss | pm-competitive-analyst | "Analyze Q1 win/loss patterns from our sales data" |

### Routing Keyword Quick-Map

| Keywords | Agent |
|----------|-------|
| PRD, product requirements, roadmap, prioritize, RICE, Kano, product vision, product strategy, "what to build", opportunity, north star, feature prioritization | pm-product-strategist |
| persona, customer interview, journey map, VOC, voice of customer, churn, NPS, CSAT, CES, customer discovery, pain points, user needs | pm-customer-insight |
| GTM, go-to-market, positioning, messaging, MRD, launch plan, sales enablement, buyer persona, product marketing, PLG, category | pm-market-strategist |
| business case, financial model, market sizing, TAM, SAM, SOM, pricing, unit economics, LTV, CAC, NRR, NPV, break-even, feasibility, revenue model, Van Westendorp, Lean Canvas, Rule of 40, Magic Number, payback period | pm-business-analyst |
| competitive analysis, battle card, win/loss, competitor, Porter's, SWOT, competitive landscape, differentiation, market intelligence, competitive threat, Blue Ocean, value curve, Crossing the Chasm | pm-competitive-analyst |

> **Persona routing disambiguation:** Standalone "persona" routes to pm-customer-insight (user personas). "buyer persona" routes to pm-market-strategist (buying committee personas). This distinction aligns with the user persona vs. buyer persona ownership boundary.

---

## Trigger Map Entry

For registration in `mandatory-skill-usage.md`:

| Detected Keywords | Negative Keywords | Priority | Compound Triggers | Skill |
|---|---|---|---|---|
| product strategy, PRD, product requirements, roadmap, prioritize, RICE, customer insight, persona, journey map, VOC, voice of customer, business case, market sizing, TAM, pricing, unit economics, competitive analysis, battle card, win/loss, GTM, go-to-market, positioning, messaging, MRD, launch plan, product marketing, buyer persona, JTBD, north star, Kano, product vision | code review, architecture, engineering, implementation, deployment, testing, CI/CD, infrastructure pricing, cloud pricing, adversarial, transcript, penetration test, exploit, strategy (standalone) | 9 | "product strategy" OR "product requirements" OR "market sizing" OR "go-to-market" OR "competitive analysis" OR "business case" OR "buyer persona" (phrase match) | `/pm-pmm` |

> **Priority 9 rationale:** Below /ast (8) and /adversary (7), ensuring /pm-pmm does not capture general analysis or AST requests. PM/PMM domain terms provide sufficient specificity via compound triggers. Standalone "strategy" replaced with compound "product strategy" to avoid collision with /problem-solving.

---

## Dependencies / Prerequisites

### Templates (Phase 4)

15 artifact templates in `skills/pm-pmm/templates/` will be created during Phase 4 (naming: `{NN}-{artifact-slug}.template.md`). Templates cover all 15 artifact types from the Artifact Ownership Matrix. Each template includes both discovery and delivery mode sections.

### SSOT Files

- `.context/rules/quality-enforcement.md` -- Quality gate thresholds, criticality levels, adversarial strategies
- `.context/rules/agent-development-standards.md` -- H-34 dual-file architecture, tool tiers
- `docs/schemas/agent-governance-v1.schema.json` -- Governance YAML validation schema

---

## Constitutional Compliance

All agents adhere to the **Jerry Constitution v1.0**:

| Principle | Requirement |
|-----------|-------------|
| P-001: Truth and Accuracy | Framework application produces canonical structures. Claims based on evidence. |
| P-002: File Persistence | All outputs persisted to project-relative paths per ADR-output-path-resolution-001. |
| P-003: No Recursive Subagents | All agents are workers. No agent has the Task tool. Main context orchestrates. |
| P-011: Evidence-Based | All claims tied to data, citations, or stated as hypotheses with confidence levels. |
| P-020: User Authority | Conflicting recommendations surface both sides. User decides. Never override. |
| P-022: No Deception | Discovery artifacts labeled as hypotheses. Confidence levels reported honestly. |

---

## References

| Source | Content |
|--------|---------|
| GitHub Issue #123 | Full PM/PMM skill specification |
| `eng/phase-1-research/architecture.md` | 5-agent architecture, artifact ownership, data flows |
| `eng/phase-1-research/frontmatter-schema.md` | Frontmatter fields, governance YAML structures |
| `.context/rules/quality-enforcement.md` | Quality gate SSOT |
| `.context/rules/agent-development-standards.md` | H-34 dual-file architecture |
| `docs/schemas/agent-governance-v1.schema.json` | Governance YAML schema |
| `docs/governance/JERRY_CONSTITUTION.md` | Constitutional principles |

---

## Context Budget Note

> This SKILL.md exceeds the typical ~500-token Tier 1 budget due to the skill's scope (5 agents, 18 frameworks, 15 artifacts). The triple-lens navigation table enables selective section loading. Framework catalog and cross-agent data flow sections are reference material that agents load selectively via Tier 2/3.

---

*Skill Version: 1.0.0*
*Constitutional Compliance: Jerry Constitution v1.0*
*SSOT: `.context/rules/quality-enforcement.md`*
*Architecture: PROJ-018 PM/PMM Skill, Issue #123*
*Created: 2026-03-01*