The river at GET / is an aesthetic choice and stays. The surface that corresponds to upstream's /status is the field report, and that is where parity is owed -- so it now carries upstream's status information set rather than the handful of process-local counters it grew from: repository rollup with percent complete (BackfillStats), live cursors (LiveStats), retention with configured lookback and oldest retained seq (CursorLookbackStats), and timestamp-import job state and phase (ImportInfo), alongside the lifecycle phase and backfill duration already added. Absence stays distinguishable from zero throughout, which is the failure mode this surface is most prone to: no repository state does not render as "0 of 0 complete", and disabled cursor replay is reported rather than shown as no retention. A test pins the whole set so it cannot quietly shrink back. This closes the row honestly rather than by argument. The previous claim that our shape was "different by design" did not survive checking: git log shows the page began as da53e98, "homepage: an ascii river at GET /", an ASCII-art feature that later grew stats. It was never designed against upstream's status surface at all. Status summary/collections/segments and status hosts/accounts partial -> verified. zig build test (311), ReleaseSafe, status-contract. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>