THE CLOSE

What you now hold

Nine diseases, each with a receipt trail from a real city. And at the end of each chapter, questions — the same questions a surveyor would put to your builder, rephrased so you can ask them yourself, tonight, with no technical background and no access to anything but your own product and your own AI.

Ask them honestly and you will learn a great deal — and if the answers convict your city on several counts, that is not an indictment. The one city fully on the record here, built carefully by a governance-minded founder, closed its survey with twenty-two findings and sixteen contradictions (Engagement 01 record, E.1). The diseases are properties of the workforce, not of you.

What the questions cannot do

A survey runs three levels, in strict order. The first walks the streets — what the city shows anyone willing to look. The second reads the blueprints: the code, the records, the guards, checked against what the streets showed. The third interviews the mayor — you — and here is the part that cannot be self-served: the interview's value is not your answers. It is the gap between your answers and what the first two levels already found. In the first city's interview, the owner was asked whether he could verify, himself, any claim a worker makes. The answer on the record is one word: "No." Asked what he accepts on faith because he cannot check it, the answer was one word again: "Everything." (Engagement 01, F-010). Fifteen separate times in one sitting, the answer to a question about his own city was some form of "I would need a tool to track this" (Engagement 01, F-021) — and more than once, the tool existed, current and waiting, and nobody had told the mayor it was there.

You cannot find that gap yourself, for the same reason you cannot proofread your own letter the moment you finish writing it. The highest-value finding in the whole instrument is the contradiction — two levels of the city disagreeing with each other — and seeing one requires two levels and a stranger.

What a survey is

Fourteen districts — from the workers themselves, through the walls, the lamps, the records hall, the firehouse — each walked at those three levels, never out of order. Every claim graded by what enforces it, because a rule with no named enforcement is decoration. Every finding delivered with its receipt, in an owner's language, ending not in a lecture but in a work order: what to fix, in what sequence, and what each item costs to defer.

And one more thing, because it is the whole reason this book gets to exist: the instrument is governed the way it tells you to govern. During the first engagement, the surveyor's own work produced twenty-one confident predictions that all died on contact with evidence, and a self-test that caught one planted defect out of two (Engagement 01, R3.1–R3.3). Both are published in the engagement record, graded with the same scales. An inspector who hides their own misses is running disease #3 as a business model.

If you want a stranger to walk your city

The survey exists as a service. If this book convicted your city on three tells or more, that is the signal worth acting on — before the busy season, before the next hire, before the buyer's diligence call, whichever deadline your city answers to.

Bring the questions you could not answer. That list is the beginning of the map.