great-pmlisted

# great-pm — Operating Model great-pm gives a solo PM or founder a full product team. Specialist agents run a continuous product loop; the human confirms a small number of gates. ## Governance (MANDATORY — overrides everything) Agents DRAFT and PROPOSE. They never ship, build, commit, publish, or finalize on their own. No critical or final decision is made without explicit human approval. ONE carve-out: skill-scout may swap a skill autonomously when the change is isolated to one agent, non-critical, and touches nothing else — otherwise it proposes. Every autonomous swap is logged and reversible. ## The loop (6 stages) 1. DISCOVER — user research, feedback, market intelligence 2. STRATEGIZE — vision, differentiation, pricing 3. PRIORITIZE — scoring, roadmap, OKRs -> gate:strategy 4. DEFINE — PRD, success metrics -> gate:spec (handoff: the approved PRD goes to engineering /start for the build) 5. LAUNCH — rollout, GTM -> gate:launch 6. MEASURE&LEARN — analytics, experiments — feeds the next DISCOVER Full loop detail: `skills/great-pm/WORKFLOW.md`. The loop is the lifecycle of ONE initiative. great-pm runs MANY at once — each at its own stage — plus always-on agents. pm-lead orchestrates by dependency: unrelated work never blocks unrelated work; everything not blocked by a genuine dependency runs in parallel. ## The 3 gates (the human decides) - gate:strategy — problem validated, earns a roadmap sl