feat(deploy): run migrations as part of the release, and report schema state

Central was serving two revisions behind the code it shipped: review_contracts
did not exist there although its migration was inside the running image. There
was no migration mechanism at all — bare uvicorn CMD, nothing chart-declared —
and nothing surfaced the mismatch. The API starts happily against a schema it
was not built for and only fails when a request touches a missing table.

Adds a chart-managed Helm pre-install/pre-upgrade hook running alembic upgrade
head, weighted to complete before the API rolls. A hook rather than an init
container: init containers run per pod, so more than one replica means
concurrent alembic upgrade with no locking. Failed jobs are deliberately
retained — a migration that fails and vanishes is how this drifted in the first
place.

/state/health now reports applied and expected revisions. "unknown" is
deliberately not "ok": an instance that cannot establish agreement must not
claim it, the same principle as instance_role defaulting to unknown.

Refs STATE-WP-0083-T07

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

Assistant: claude-code
Assistant-Model: opus
Assistant-Process: 2583210@bnt-lap001
Assistant-Session: f2bff2d5-e9b2-4338-92ca-10282a927006

deploy/railiance/apps/charts/state-hub/templates/migrate-job.yaml

@@ -0,0 +1,40 @@
+{{- if .Values.migrations.enabled }}
+apiVersion: batch/v1
+kind: Job
+metadata:
+  name: {{ include "statehub.fullname" . }}-migrate
+  labels: {{- include "statehub.labels" . | nindent 4 }}
+  annotations:
+    # Run before the API starts serving, and before an upgrade swaps the image.
+    # A deployment that can serve against a schema it was not built for is the
+    # same class of defect as a projection that cannot name its source commit
+    # (STATE-WP-0083-T07).
+    "helm.sh/hook": pre-install,pre-upgrade
+    "helm.sh/hook-weight": "-5"
+    # Keep a failed job for inspection; a silent migration failure is how the
+    # schema drifted two revisions behind the code in the first place.
+    "helm.sh/hook-delete-policy": before-hook-creation,hook-succeeded
+spec:
+  backoffLimit: {{ .Values.migrations.backoffLimit }}
+  template:
+    metadata:
+      labels: {{- include "statehub.labels" . | nindent 8 }}
+    spec:
+      restartPolicy: Never
+      {{- with .Values.imagePullSecrets }}
+      imagePullSecrets: {{- toYaml . | nindent 8 }}
+      {{- end }}
+      containers:
+        - name: migrate
+          image: {{ include "statehub.image" . | quote }}
+          imagePullPolicy: {{ .Values.image.pullPolicy }}
+          command: ["/app/.venv/bin/python", "-m", "alembic", "upgrade", "head"]
+          envFrom:
+            {{- if .Values.config.enabled }}
+            - configMapRef:
+                name: {{ .Values.config.name | quote }}
+            {{- end }}
+            - secretRef:
+                name: {{ .Values.secret.name | quote }}
+          resources: {{- toYaml .Values.migrations.resources | nindent 12 }}
+{{- end }}

deploy/railiance/apps/charts/state-hub/values.yaml

@@ -51,6 +51,21 @@ ingress:
     traefik.ingress.kubernetes.io/router.tls: "true"
     cert-manager.io/cluster-issuer: letsencrypt-prod
 
+# Database migrations (STATE-WP-0083-T07). Runs as a Helm pre-install/pre-upgrade
+# hook rather than an init container: an init container runs per pod, so more
+# than one replica means concurrent `alembic upgrade` with no locking. A hook
+# runs once per release and fails the upgrade if the migration fails.
+migrations:
+  enabled: true
+  backoffLimit: 1
+  resources:
+    requests:
+      cpu: 50m
+      memory: 256Mi
+    limits:
+      cpu: 500m
+      memory: 1Gi
+
 # Classification allowed-values (CUST-WP-0067-T09). The API validates repo
 # classification against the-custodian canon; a container has no such checkout,
 # so the file travels with the release as a ConfigMap. Without it every