The Golden Thread

A machine-readable thread that ties executive intent, build, and proof into a living ledger.

HomesFlow is a real app. But the app was never the point.

It's a test of one question: can one person keep AI-assisted development honest without slowing it to a crawl? Not "prompt-driven development," where governance is whatever you typed into the chat window last. Real governance, at AI speed.

For a CTO, the value isn't the methodology. It's control.

The base workflow is standard. I use Spec-Driven Development with GitHub's Spec Kit to work iteratively through a PRD, feature spec, plan, tasks, and code. But I write the intent down before anyone, human or agent, builds it, then move up and down the stack as the work demands.

Standard wasn't enough. But a spec can exist and still drift. Tasks get marked done without proving which intent they served. Tests pass without making clear which acceptance criterion they answer for. AI is especially good at producing code that looks right without being justified by anything.

So I added one more layer: a machine-checkable thread. Every requirement (a statement of intent) in HomesFlow.prd.md gets a stable ID. That ID has to survive, visibly, through the spec, the tasks, the Swift code (@covers annotations), and the test names themselves (test_AC_SYNC_05...). A script, my Gate 2, checks the whole chain on every push. Reference an ID that doesn't exist, skip a Carries: field, or ship an acceptance criterion with no test and no tracked gap, and the build fails.

The snapshot as of this note's last update:

82 registry IDs in the PRD
67 proven, 10 riding as tracked debt, 5 waiting in backlog
0 silent gaps; Gate 2 passing in CI

Ten IDs are not yet proven. That is not a hole in the gate, it's the gate working: each one rides on an open task where anyone can see it, and five more wait in backlog, minted and visible. Passing doesn't mean zero gaps. It means zero hidden ones.

Those numbers are the point. They're the difference between "we have documentation" and "the documentation governs the work."

AI lowers the cost of writing code, and just as much, the cost of writing code that merely looks right. The Golden Thread forces the same four questions on every change: What intent does this serve? Where's it built? What proves it? If proof is missing, is that gap visible or hidden?

It is not perfect. Renaming the project from HomeFlow to HomesFlow touched the repo, Xcode targets, the bundle ID, Supabase, SonarCloud, CI paths, and docs. Gate 2 held the core traceability intact. Everything else still needed a manual pass.

It is good. Scope creep becomes visible. Untested acceptance criteria become visible. Drift between what was promised and what shipped gets much harder to hide. It even makes it possible to quickly and meaningfully answer questions like "What is this program costing us?" – in real time, before the bills even arrive.

None of that required a new methodology: GitHub's Spec Kit gives the workflow, and the gate turns the chain into something CI enforces instead of something a slide deck asserts.

The receipts are in the repo: github.com/rdryfoos/HomesFlow. Or if you'd rather talk it through first: book some time.