Playbook 1: Greenfield
/pureinn [idea]▊Build a new product from scratch - from first idea to launch. Full discovery, validation, and JIT delivery.
Use when: new product, zero users, need to validate PMF before building.
How to read this page
This is a route map - which phase comes next, when a whole phase can be skipped, and what unlocks the next one. Each phase links to its own runbook page with the per-skill detail (when to run/skip each skill, what you get, what it doesn't).
Phase overview
Phase 1 Phase 2 Phase 3a Phase 3b Phase 4 Phase 5 Phase 6+7
Foundation → Discovery → Validation → Commercial → Domain → Feature → JIT Delivery
┃GO gate┃ Definition + Registers Planning (per feature)
┗━hard━━┛
Express path (skip discovery/validation)
- When to use: you already have a validated idea, know exactly what to build, and don't need discovery or market/customer validation.
- Skips: Phase 1, 2, 3a, 3b entirely.
- Never skips: domain registers and a feature list with FEAT-IDs -
pm-feature-designcannot run without them.
/pureinn [idea] # workspace setup, 3 intake questions
/pm-entity-registry # entity list + key states only (lean mode)
/pm-business-rules-library # core rules in Draft
/pm-features-list # feature inventory + FEAT-IDs + KANO + V×C
/pm-mvp-scope # MVP scope + Delivery Stripe assignment
/pm-feature-design [FEAT-ID] # → JIT delivery engine
Done when: the first feature enters pm-feature-design. From there you're in the same JIT cycle as the full path.
Full flow
| Phase | Enter when | What happens | Detail |
|---|---|---|---|
| 1 - Foundation | Start of every Greenfield project | Scope, team, stakeholders defined - scales down to nothing for solo builders | Phase 1 |
| 2 - Discovery | Phase 1 artifacts exist (or /pureinn discover) | 4 parallel tracks (Tech, Domain, Market, Customer) converge into a Problem Validation Summary | Phase 2 |
| 3a - Validation | Problem Validation Summary exists (can start in parallel with late Phase 2) | Real-world experiments → hard GO/PIVOT/STOP gate | Phase 3a |
| 3b - Commercial Definition | Phase 3a GO verdict (no workaround) | Business model, KPIs, business case, PRD - PRD frozen at exit | Phase 3b |
| Pre-Phase 6 - Technical Foundation | PRD frozen | /common-ground (fullstack-dev-skills) - tech stack decision, repo structure, COMMON-GROUND.md | - |
| 4 - Domain Modeling | Technical Foundation done | Live Registers (entities, rules, decision tables) initialized in Draft | Phase 4 |
| 5 - Feature Planning | Registers initialized | Feature inventory, prioritization, MVP scope, Delivery Stripes | Phase 5 |
| 6+7 - JIT Delivery | Features assigned to Stripes | Per feature, just before build: spec → design inspection → build → review → ship | Phase 6+7 |
Key rules
- Phase 3a is a hard gate. Only GO advances. PIVOT loops back to targeted re-validation. STOP ends the project. No FORCE bypass.
- Phase 3b requires Phase 3a GO. Cannot be entered without a GO verdict - there is no workaround.
- PRD is frozen after Phase 3b. No scope changes without a formal change request tracked in the PRD changelog.
- Spec gate before build (Phase 6, hard gate). Feature Card Sections 1-3 complete + Design Inspection passed before any feature enters code.