AwesomeThe control-plane story
Six chapters. The problem, the solution, the layers, the pivot, how to adopt it, and where it goes next.
The problem
Out of the box, every AI assistant starts from scratch on every session. It doesn't know your naming conventions, your accessibility requirements, or your branching strategy. Multiply that across a team and you get a different implicit rulebook for every developer — and code review becomes a game of catching AI-generated patterns that don't match your standards.
The insight
A single .github repository that all your other repos inherit from is the fix. Define standards once in machine-readable formats and let every AI tool consume them. GitHub calls it a "special repository" — we call it the control plane.
The layers
The control plane has two distinct layers: a hooks layer that describes semantic intent ("label new issues"), and a workflow layer that compiles those intents into GitHub Actions. Instructions, agents, skills, and prompts sit above both — they shape reasoning before code is written.
The pivot
Beyond governance files, the control plane is becoming a plugin distribution system. A manifest describes what a plugin pack contains; consuming repos install it with one command. Standards become packages. The whole team moves with a single version bump.
Adopt it
Start with the catalogue — browse by type, copy what you need, install into VS Code with one click, or grab the raw URL. Every resource carries installation instructions and links directly to the source file on the develop branch.
The horizon
The next stage: plugin packs you install like npm packages, AI-powered code review that knows your exact standards, and auto-syncing governance that stays current without manual upkeep. The foundations are in this repository right now.
Read the references →