Directing the agent,
not just trying it
Building software by directing an AI coding agent is a real capability — or a risk, depending on how it's done. This check scores your team on the workflow (is the agent set up and used well?), the guardrails (is AI-generated code safe in your delivery process?), and the fit (is it a team capability, or one person's trick?).
Workflow
The agent set up with real context, people who know when to reach for it, and the skill to steer it back when it drifts.
Guardrails
Every AI-generated change reviewed, tested, and owned by a human — with scope control and a security check.
Fit & spread
A capability more than one person has, in the real delivery process, with a shared view of what it's good for.
// Answers are scored in your browser. Nothing is sent to a server and nothing is stored after you close the tab.
Work the fix
Your check points at the area to start with. The Vibe Coding guide walks through the fix for all three — plus a walkthrough deck, and a course on the way.
Your scores go to Jon, with your address as the reply-to. Nothing is stored on the site.
Tried it and still stuck? Bring Jon in on that one problem.
A 30-minute call to walk through your score and where an engagement would start.
Book a conversation →