Module 03 · Delivery stages
runnableStage Gate
Make the path from appointment to build visible.
Runnable v0.1 slice available on GitHub — architecture and scope below.
Problem
AI-assisted projects can jump from an appointment straight into generated code. The missing middle—prerequisites, requirements, decisions, and clear advance criteria—is usually spread across documents and understood by only a few people.
Where it sits
Sits over an existing delivery harness and makes its stages visible. It is meant to guide work from pre-requirements through requirements, architecture, build, and specification without forcing one development framework.
Think of it as the difference between booking the appointment and doing the work. A project first checks pre-requirements, then moves through requirements and architecture before build or specification work is allowed to advance—on top of a harness you already have, for example VS Code plus a side agent.
The proposed flow connects high-level requirements to functional and non-functional requirements, ADRs, and implementation specs. SpecKit or OpenSpec can plug in where useful, but neither is required. A non-technical owner should be able to follow the same flow and answer “when do I advance?” without living in a terminal.
How it fits
Stage Gate and Shadow Versioning share the same project ID: the gate explains where the work is, while versioning keeps the related artifacts tied together. Context Engine can supply indexed project material without turning the gate into another repository browser.
Capabilities
- Intended visual stages over an existing delivery harness
- Pre-requirement checks before requirements and build work
- Advance criteria from HLRs through FRs, NFRs, ADRs, and specs
- Optional SpecKit and OpenSpec integration—not locked to one method
- Side-agent guidance inside VS Code
- Readable stage status for non-technical project owners
Other modules
- Smart RouterModel access gets scattered across provider keys, endpoints, deployment rules, and separate UIs. Even a VS Code assistant and a home chatbot can end up repeating the same integration work.
- Evals & English GovernanceEach model platform has its own eval screens, scoring terms, and policy controls. Teams end up comparing results by hand, while governance rules live in documents that do not map cleanly to provider settings.
- Shadow VersioningOne project can have its brief in Confluence, approvals in SharePoint, code in Git, and evidence in a testing system. Humans see one piece of work, but the tools expose unrelated histories and technical version language.
- Context EngineAn agent working across 15 repositories can spend more tokens rediscovering the codebase than answering the question. Every tool builds its own partial picture, then throws much of that work away.
- Data LineageOnce a prompt crosses a gateway, it can be hard to explain which provider, region, model, or internal service handled it. Existing lineage tools cover data estates well, but AI request paths often need a smaller, trace-linked view.
- AI-DORAAI coding agents inflate commit counts and PR volume, hiding code churn, review bottlenecks, and rework. Traditional DORA dashboards track deployment frequency without telling you if half the AI-written lines were reverted two weeks later.