The behavioural loops

# The behavioural loops

| Loop | Trigger | User value | Product implication |
|---|---|---|---|
| Orient | material change or milestone | relief from monitoring | interrupt on changed meaning, not activity |
| Judge | a consequential uncertainty | agency without operating the studio | ask for one small, legible decision |
| Steer | a question becomes worth reading | momentum and a live read to decide against | point a Position at the question — a priced Reading with a monitor |
| Learn | a Position re-prices on evidence | the read sharpens legibly | show how the read moved, its reasoning, and the evidence |
| Refine | the expert can be corrected | ownership and compounding | make later behaviour show what was taught, and keep the track of how the read moved |

Human-in-the-loop does not mean approving everything. Research can run autonomously while Merge folds on
the domain's cadence and **a re-priced read awaits the user's review** (the machine re-prices the
read; the human decides what to do with it). Trust expands by consequence:

```text
observe → suggest → stage → apply with review → apply by default → audit exceptions
```

The boundary must remain inspectable, reversible, and granular.

Raw Markdown source