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.
1OK to build 'one conversation per person per company'? Today a conversation is keyed per person per home; Western Slope (one company mailbox/line, 70 homes) had a prospect email land on the company front-door record, match no thread, and sit unmatched for an hour. Proposal: every engagement from a person flows into one conversation; home becomes metadata on the inquiry. Ships dark behind a per-company setting, default off. Company remains the join boundary -- no cross-company merges. Single-building customers (Camellia, Yale 25 Station) unaffected. Rollout: lab harness (merged) -> Western Slope (70-home test co) -> Willows. Design: https://docs.propflowai.co/a/agent-loop-decoupling-2026-09?s=person-first-design
In plain terms. Should we let all of one person's messages at a multi-home company land in the same conversation, instead of a separate one per home?
At Western Slope — one shared inbox and phone number covering 70 homes — a prospect's reply landed on the wrong record and sat unread for an hour, because today each home keeps its own separate conversation. The fix would combine a person's messages into one conversation per company, turned on customer by customer, off by default. Single-building customers like Camellia wouldn't change. It's already been proven in a test harness; next is a live trial at the 70-home test company, then a bigger one.
What do you want to do?
The session raising this recommends: Build it -- ship off, turn on per customer Stop-gap fixes are live; this is the lasting fix so it doesn't happen again at other multi-home companies. Tried first, unsuccessfully: Design doc written and reviewed, lab harness for the person-first grouping already merged; Fede reviewed and recommended building it, then handed the decision to Gera in Slack instead of deciding it himself.