This page explains the intended lifecycle. Factory aims to turn business objectives and customer problems into lasting customer value across products. See values for the decision rules and build review for current implementation evidence.
Lifecycle
Planning scopes a slice, the smallest useful change, stating its intended outcome, how to measure it and what would disprove it. Delivery builds a candidate: the exact proposed change. Assurance tests it independently and issues a technical verdict. Operations releases it safely. After Assurance checks the measurement, Planning records an assessment comparing observed results with the intended outcome: Supported, Rejected or Inconclusive. Only a Supported assessment counts as value, not finished work or a release.
Decision domains
| Domain | Decides |
|---|---|
| Planning | Value, priority, slice scope, required capabilities and retirement. |
| Delivery | How a slice becomes a candidate; repair. |
| Assurance | Independent scenarios, checks, verdicts and measurement verification. |
| Operations | Integration, release, health, recovery and vendor access. |
| Growth | Docs, websites, launches and acquisition from confirmed release facts. |
The adoption layers are Module → Workflow → Factory. Modules are bounded building blocks; Workflows perform complete jobs such as research, implementation, verification and marketing; Factory combines them across the lifecycle. Each lower layer must work in another system with its declared dependencies. Architecture and adoption evidence distinguish this goal from what works today.
The proposed Factory shares three foundations across domains: Execution runs durable, isolated work; the Evidence store keeps evidence and provenance; the Action gateway checks authority and policy, and reconciles external effects.
Protecting customers
Assurance sets acceptance scenarios before building; builders cannot weaken them. Potentially unsafe changes reach alpha before stable; health, security and privacy checks decide promotion or rollback. Retirement is deliberate; migration, recovery and deletion obligations stay binding.
Autonomy and evolution
Within owner-granted authority and budget, the factory acts on its own and may revise its operating policy; only owners widen authority. It asks people only for a decision, access or resource it cannot obtain. Independent recovery restores the factory and its products; unattended self-update waits until recovery drills pass.
The proposed system would recover from failures, research improvements nightly and simplify weekly, restoring accidental loss. These changes follow the same lifecycle; only measured gains are adopted. Interfaces are renewed when evidence shows friction or accessibility gaps.
Pilot, spending/capacity limits, thresholds, recovery targets, exact schedules and stack await commissioning. No implementation is claimed.
