ATL-D part 2 — the org map's record line, on real stage data
The focused org's map with the live head read; a failed read now marks one number, not the whole org.
27 Sept 2026 · branch ui-atl-d-part2-could-not-check · captured against propflow-stage · at 11d7590858
Data note. Every name and balance below comes from propflow-stage, which is anonymized by construction (ADR-0097/0110 — phones in the impossible +1000 NPA, emails @example.test). The identical capture against prod would carry real tenants' names and balances and is not safe to attach anywhere, which is why the generator refuses to run against one.
1 · One focused org, the map, the record's answer
The TEST company, opened from the Organizations grid, in the Map view. The line under the org name is the per-tree count the projector builds from the head read — the rendered capture #8491 owed for part 1, taken on this branch against propflow-stage.
What this does not show: the new could not check state. A healthy table does not fail a point read, and no fault injection was added to stage one. That copy is pinned by hand-written literals in atlas-org-map-record-copy.test.tsx, and the route's per-number failure by atlas-detail-route-contract.test.ts.
Also visible, and out of this PR's scope: one phone card carries no record state — a code-routed number the read scope (built from the property rows' numbers only) never looks up. That is #8491's still-open reviewer finding.
Asserted in the captured DOM — Fairhaven Residential (TEST) · One tree per org · phone number looked up here · Read-only — nothing on this view writes.. The count itself is measured, not pinned — stage's head rows move.
