Kit 24 · Roster Reset · deliverable · designed and built by I Wayan Usada — instructional design portfolio
Planning teams audit AI scheduling proposals as an auditable practice: reason-log quality up, silent accepts and reflex overrides down, in ops review.
Teams either rubber-stamp the AI proposal or override it reflexively — both unlogged, both unimprovable; judgment leaves no trace.
| # | Behavior | What it looks like | Where the artifact trains it |
|---|---|---|---|
| 1 | Find the hidden constraint violation | violation located before any decision | proposal audit |
| 2 | Choose accept / edit / override for a reason | every decision logged with a stated reason | reason log |
| 3 | Correct the board rather than win the week | corrected board with stated reason outranks luck | debrief |
It will not teach the scheduling tool — the unit is the human-in-the-loop audit habit, tool-agnostic.
Debrief scores decision correctness AND reason quality — judgment is the artifact, not the board.
Prefer the source? Download the raw markdown.