Everything I could decide myself is decided. What's left needs
you. Each one carries the recommendation of the session that raised it — if you
agree, pick it and press Done. A model's opinion, where there is one, appears on the card
as a receipt line; cards with no receipt were never put to a model. "I'm not sure" is a real answer and becomes work for me.
You don't need to open the session — pressing Done sends your answer back and it picks up.
1An earlier ruling said to wait for PR #9767's re-land before repairing 5 duplicate customer records, because the cross-property redirect guarantee was unproven. #9786 has since merged and closes that gap, and the proof obligation now passes 6 of 6 against merged main while still failing 3 of 6 against pre-merge main (non-vacuity control). #9767 is prospective-only, operates on the mint path only, and has not been started. Is waiting for #9767 still required before repairing the 5 pairs?
In plain terms. Should I restore five customers' missing notes tonight, or hold off? Each of them was accidentally created twice and the team keeps opening the blank copy, so their page looks empty; pointing the blank copy at the real one makes the notes visible again and can be undone, while holding off leaves those five looking empty until a teammate rewrites a fix they pulled tonight and has not restarted.
The session raising this recommends: Restore their notes tonight; it can be undone Both decider passes agreed on this and it escalated only because the question brushed a commitment-to-a-person class gate. The condition we were waiting on existed to guarantee the card follows the record it is merged into across properties; that is now measured directly on merged main, 0 of 6 pairs failing, while the same test still fails 3 of 6 against pre-merge main so it is a real measurement and not a test that stopped checking. The change we are waiting on only prevents NEW duplicates at creation time, so it cannot change how an already-repaired pair behaves. It has no branch and no pull request 1h15m after being pulled, so waiting is open-ended. Caveat I am verifying now: merged is not deployed, and the repair must not run until the cross-property fix is live on the production sync worker. FABLE was asked, and escalated. Receipt f55c89db2 — NEEDS HUMAN. ANSWERED BY ASTRA, NOT FABLE — Fable did not answer because it could not be reached.
Both passes agreed on 'No longer required — the compatibility the condition protected has been directly measured, so repair the 5 pairs now using the tombstone direction blank-into-rich', and it is being escalated anyway: this question turns on money or a commitment to a person (matched 'guarantee'), which is not a model's to settle however confident it sounds.
Paid for on 2026-08-07: the unit-607 renewal rate came back RESOLVED from both passes at high confidence and was wrong — Fable found the standup ruli…