End a diet engagement and capture the closure summary

Twelve weeks of the cut done. You hit the target weight on Friday and you're switching to maintenance Monday. Before activating the next plan, you close out the cut engagement so the closure summary captures what happened — adherence, weight delta, refeed compliance, the whole arc as one read.

What the agent needs to hear

Behind your sentence, the agent works backwards to two pieces: confirmation that you’re ending the engagement and, optionally, a closing note — « hit the target », « calling it early — too aggressive », « switching to maintenance ». The active engagement is implicit; there’s only one running at a time, and the agent already knows which.

Before flipping anything, the agent renders a closure summary on a preview card — adherence percentage across the window, weight delta start to end, refeed compliance, average daily macros — and waits for a nod. If the numbers look honest, « yes, close it » settles it.

What gets captured at closure

The closure summary freezes a handful of read-once metrics against the engagement window. The adherence percentage is computed across every day from start to close. The weight delta reads your start weight against your end weight. The refeed compliance captures how often you hit the scheduled refeed days if a protocol was set. The average daily macros roll up across the engagement — calories, protein, carbs, fat. And the closing note, if you offered one, attaches as the engagement’s last word.

Once saved, the summary is immutable — the engagement moves to history with everything frozen as of close. Nothing about the underlying meal logs changes; they stay where they were, readable as past data, unaffected.

What ending the engagement does vs leaving it open

Leaving a finished engagement open is the silent failure mode. The cut keeps reading as « active » in your daily view, the daily targets keep drawing against the cut macros instead of whatever you’re actually eating, and the adherence percentage corrupts as the off-plan days pile up.

Ending the engagement freezes the window cleanly. Adherence is read once and stored. The daily view stops drawing against the cut. The engagement appears in your engagement history, available for comparison against future cycles. The plan itself isn’t deleted — it stays in the library, ready to re-activate later for a future cut.

When the agent gets it wrong

The preview is the chokepoint — three failure modes show up there. The agent ended the wrong engagement: « no, the cut — the one active right now, not the maintenance from last quarter ». The agent moved to delete the plan instead of just ending the engagement: « keep the plan in the library, just close this engagement ». The closure date is wrong: « I finished Friday, not today — backdate the close ». Each correction reshapes the card; you confirm again before it commits.

What makes a clean engagement close

Three things decide whether this engagement closes usefully: the closing summary captures the engagement’s actual story (a 12-week cut closed at the wrong date misreads the adherence window by days), the plan stays in the library and the engagement goes to history (deleting the plan along with the engagement strips your ability to re-run or compare later), and the closing note carries the intent (« hit the target » reads differently than « bailed at week 6 — too aggressive »; both are valid signals for future plan design). A clean closure makes the next plan’s targets more informed — the engagement history is what the system reads when you ask what worked.