PM Agent Team Skill

SkillAI & models

PM Agent Team - Automated product discovery, strategy, and PRD generation. Runs PM analysis workflow to produce comprehensive PRD before PDCA Plan phase.

Available today. Use it from your connected AI after setup.

Connect ahel once, and every AI you use reads what you have installed.

Then ask your AI: use the PM Agent Team Skill skill

What this skill tells your AI

The instructions your AI receives, as published by ww-w-ai/bkit-gemini in skills/pm-discovery/SKILL.md and read by ahel’s review.

Automated product discovery, strategy analysis, and PRD generation. Produces a comprehensive PRD before the PDCA Plan phase begins.

Overview

The PM Agent Team applies 8 professional product management frameworks to systematically analyze a feature before implementation begins. This ensures that development is grounded in validated strategy rather than assumptions.

PM Frameworks

The following frameworks are applied across the 4-phase workflow:

#FrameworkPhasePurpose
1Opportunity Solution Tree (OST)DiscoveryMap problem space systematically
2Teresa Torres' Continuous DiscoveryDiscoveryEvidence-based opportunity identification
3JTBD 6-Part Value PropositionStrategyDefine who, why, and what progress
4Lean CanvasStrategyValidate business model viability
5User Personas (JTBD-based)ResearchHumanize target segments
6Competitor Analysis MatrixResearchIdentify differentiation opportunities
7TAM/SAM/SOM Dual ValidationResearchSize the market with confidence
8Beachhead Segment + GTM (Moore)PRDDefine go-to-market strategy

4-Phase Workflow

Phase 1: Discovery (OST)
    Desired Outcome → Opportunities → Solutions → Experiments
        │
Phase 2: Strategy (VP + Lean Canvas)
    JTBD 6-Part → Lean Canvas → Key Assumptions
        │
Phase 3: Research (Personas + Competitors + Market)
    3 Personas → 5 Competitors → TAM/SAM/SOM
        │
Phase 4: PRD Synthesis
    Beachhead → GTM → 8-Section PRD
        │
    Output: docs/00-pm/{feature}.prd.md

Usage

Via PDCA Command

/pdca pm {feature-name}

This runs the full PM Team workflow for the specified feature and outputs the PRD to docs/00-pm/{feature-name}.prd.md.

Direct Agent Invocation

Individual phases can be triggered by mentioning relevant keywords:

  • Discovery: "opportunity analysis", "OST", "customer pain points"
  • Strategy: "value proposition", "lean canvas", "JTBD"
  • Research: "persona", "competitor analysis", "market sizing"
  • PRD: "product requirements", "beachhead", "GTM strategy"

PDCA Integration

The PM Team workflow is a pre-phase that feeds into the PDCA cycle:

PM Team → PRD
              ↓
         PDCA Plan  → references PRD for requirements
              ↓
         PDCA Design → translates PRD into technical design
              ↓
         PDCA Do     → implements the design
              ↓
         PDCA Check  → verifies implementation against design
              ↓
         PDCA Act    → iterates until quality threshold met

Handoff

After the PRD is generated, the PM lead agent guides the user to:

  1. Review the PRD for completeness and accuracy
  2. Run /pdca plan {feature} to begin the implementation cycle
  3. The Plan phase automatically references the PRD as its source document

Output Structure

docs/
  00-pm/
    {feature}.prd.md          # Complete PRD with all 4 phases

The PRD includes all analysis outputs in a single consolidated document:

  • Discovery analysis (OST map, opportunity ranking)
  • Strategy analysis (JTBD table, Lean Canvas, assumptions)
  • Research analysis (personas, competitor matrix, market sizing)
  • Beachhead selection and GTM strategy
  • Full 8-section PRD with user stories and acceptance criteria

Agents

AgentRoleModelTemp
pm-leadOrchestrates 4-phase workflowgemini-3.1-pro0.3
pm-discoveryOST analysisgemini-3.1-pro0.5
pm-strategyJTBD + Lean Canvasgemini-3.1-pro0.4
pm-researchPersonas + Competitors + Marketgemini-3.1-pro0.4
pm-prdPRD synthesis + Beachhead + GTMgemini-3.1-pro0.3

Signals

GitHub stars
65
Forks
16
Last commit
May 2026
Advanced
Catalog kind
skill
Gateway key
pm-discovery-ww-w-ai
Source
github.com/ww-w-ai/bkit-gemini