Connecting to the Document-First Principle
The golden rules 7+1 (a system where seven internal roles consult in parallel and an external audit confirms up to three rounds) has two paths: the standard pattern and the shortcut pattern. In the chapters so far, we have looked at the conditions and the checklist for choosing between them.
This organization holds one more thing as an absolute principle: the document-first principle (the rule that every decision is kept on record in a file). How does the act of choosing a pattern connect to this document-first principle?
The Choice Itself Disappears
Did we take the standard pattern, or the shortcut pattern? Looking only at the finished output, we cannot tell which path we went through. A conclusion reached after three rounds and a conclusion reached after one round look the same on the surface.
This is an unexpected trap. "What we made" survives, but "how we made it" does not. Even if we keep a work log, if we never write down the pattern choice itself, we lose the ability to recall later whether a given piece of work was standard or shortcut.
The cycle rule (here meaning: a fixed rule that automatically resets shortcut usage after two consecutive rounds), covered in Never Over-Trust the Shortcut Pattern, carries the same weakness. Unless we record which pattern we used every single time, we cannot even count "two consecutive rounds" in the first place. Building a rule is not enough — without a record, we cannot run it.
Why We Made "Write It at the Start" a Rule
So we decided to write one line at the start of every work record: "which pattern did we use this time." At the start, not at the end.
There is a reason we insisted on the order. If we try to write it after the work is done, looking back, the reasoning behind the decision gets fuzzy. When we try to recall later why we chose the shortcut, what tends to creep in is not the actual reason at the time but a justification built after seeing the result. If the result turned out well, we end up writing "the conditions were right, so the shortcut was the correct call." That weakens what the record is for.
Writing it at the start captures the reasoning at the exact moment of the decision, in words. If we write "there is precedent and the scope of impact is limited, so we are choosing the shortcut pattern" before the work begins, we can later check whether that judgment was correct without being pulled by the outcome. Changing the order in which we write is enough to change the quality of the record.
Who Is the Record For
A record of pattern choice is not a one-off note.
One purpose is for looking back later. The next time we face a similar case, having a record that says "last time, under these conditions, we chose the standard pattern" means we do not have to judge from zero again. When we are unsure how to answer the five questions in the decision checklist, past records also help.
The other purpose is handing off to another agent (here meaning: an AI teammate we can delegate work to). In this organization, multiple agents rotate through the work. If the reasoning behind a pattern choice is not left on record, whoever picks up the work later has no way to check whether the earlier judgment was sound. Avoiding exactly that situation is what the document-first principle is aiming for. The goal is to leave a trail that anyone can follow to see why a decision was made.
The one-line-at-the-start rule is a small habit. But whether that one line exists or not decides whether a record becomes something we can verify later, or just a trace that the work happened.
Neither the five questions organized in Decision Checklist, nor the cycle rule covered in Never Over-Trust the Shortcut Pattern, works without this one line at the start. Turning pattern selection into a system is not enough on its own — it only connects to the document-first principle once we also build in the habit of writing down what we chose.