[OPIK-7773] [BE] fix: correct the runbook's restart attribution and the readiness DTO builder

Two more review findings, both in scope.

The runbook's self-clearing note had the restart on the wrong path: it read "no extra restart
needed on the wrap-first path, and only the already-planned restart on the toggle-first path",
which is backwards. Both orderings spend exactly the one planned rolling restart the toggle
already requires, and neither needs another; what differs is its position, and with it what
closes the mismatch window — the wrap DDL on the toggle-first path (the restart is already
done by then), the restart completing on the wrap-first path. Reworded to say that.

HealthCheckResponse used a plain @builder, copied from the same record in
HealthCheckIntegrationTest, where the convention predates the rule. Records and DTOs take
@builder(toBuilder = true) (.agents/skills/opik-backend/SKILL.md:52, which lists bare @builder
as the anti-pattern), so the new copy uses it; the older one is left alone as unrelated to
this ticket.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

Author: thiagohora

apps/opik-backend/data-migrations/traces-local-v2-cutover/README.md

@@ -237,8 +237,10 @@ new table before the EXCHANGE. The replay matches the **full key**, not `id` alo
 > point of the check (an instance whose deletes cannot work should not serve), but it changes the cost of the window
 > from "delete-path 500s" to "no backend in rotation", so the window must sit **inside the declared maintenance window**
 > that `--confirm-maintenance` already asserts for the wrap. It is self-clearing: the probe re-evaluates continuously,
-> so rotation returns on the next successful probe after the two sides are back in step — no extra restart needed on
-> the wrap-first path, and only the already-planned restart on the toggle-first path.
+> so rotation returns on the next successful probe once the two sides are back in step. **Neither ordering needs an
+> extra restart** — both spend exactly the one planned rolling restart the toggle already requires; only its position
+> differs, and with it what closes the window: on the toggle-first path the restart comes first and the **wrap DDL**
+> closes the window, on the wrap-first path the wrap comes first and the **restart completing** closes it.
 >
 > Prefer **toggle first**, have the `--wrap-only` command ready to run the moment every backend instance is up, and
 > keep the window to seconds. Nothing in either direction corrupts data — that is what makes a short window

apps/opik-backend/src/test/java/com/comet/opik/infrastructure/db/healthchecks/ClickHouseTracesTopologyReadinessTest.java

@@ -272,7 +272,7 @@ private void execute(String sql) {
                 .flatMap(result -> Mono.from(result.getRowsUpdated()))).block();
     }
 
-    @Builder
+    @Builder(toBuilder = true)
     private record HealthCheckResponse(String name, boolean healthy, boolean critical, String type) {
     }
 }