feat(resilience): operational hardening (NEXT phase of the audit)

Acts on the audit's NEXT block — operational resilience.

Backups (N1):
- New `backup` compose service (postgres:16-alpine) runs scripts/backup-loop.sh:
  immediate pg_dump on start, then nightly, gzip, 14-day rotation into
  ./backups on the host. Configurable via BACKUP_RETENTION_DAYS /
  BACKUP_INTERVAL_SECONDS. (Offsite copy is the documented next step.)

Resource limits + healthchecks (N2):
- deploy.resources.limits.memory on postgres (2g), app (1500m), nginx (256m),
  backup (256m) so no container can starve the others (the Nginx outage was a
  reminder).
- Nginx now has a healthcheck hitting a new self-served `/nginx-health`
  endpoint on the default_server (no upstream dependency).

Chat resilience (N3):
- buildSystemPrompt() wraps its 4 Prisma queries in try/catch with safe
  defaults — if Postgres is down the assistant degrades instead of 500-ing.
- Result is cached for 60s (only on healthy builds) so we don't run 4 queries
  per message; CMS edits still appear within the TTL.
- POST fails fast with 503 if OPENAI_API_KEY is missing (instead of breaking
  mid-stream after headers are sent).
- streamText gets an onError handler that logs + persists an `error` AiEvent.

Idempotent submissions (N4):
- consultation/route.ts and operations.ts now wrap the email-tracking UPDATE
  in try/catch — the lead/signal is already saved, so a telemetry hiccup can't
  500 the request and trigger a duplicate retry. operations.ts also returns
  emailError.

Performance (N5):
- Index GlobalNode(application, isActive) — backs the case-study join on every
  application page. Migration 20260609130000_index_globalnode_application.

Verified: next build compiles (Docker parity, SESSION_SECRET unset),
TypeScript clean, prisma schema valid, golden tests 17/17,
`docker compose config` valid.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

scripts/db-backup.sh

@@ -0,0 +1,31 @@
+#!/bin/sh
+# ─────────────────────────────────────────────────────────────────────────────
+# Single Postgres backup: pg_dump -> gzip -> N-day rotation.
+# Run by scripts/backup-loop.sh inside the `backup` compose service.
+# Env: DB_USER, DB_PASSWORD, DB_NAME, BACKUP_DIR, RETENTION_DAYS
+# ─────────────────────────────────────────────────────────────────────────────
+set -eu
+
+BACKUP_DIR="${BACKUP_DIR:-/backups}"
+RETENTION_DAYS="${RETENTION_DAYS:-14}"
+TS=$(date -u +%Y%m%d_%H%M%S)
+OUT="${BACKUP_DIR}/flux_db_${TS}.sql.gz"
+
+mkdir -p "$BACKUP_DIR"
+export PGPASSWORD="$DB_PASSWORD"
+
+echo "[backup] $(date -u +%Y-%m-%dT%H:%M:%SZ) starting pg_dump -> ${OUT}"
+
+# --no-owner/--no-privileges keep the dump portable across roles on restore.
+if pg_dump -h postgres -U "$DB_USER" -d "$DB_NAME" --no-owner --no-privileges | gzip -9 > "$OUT"; then
+  SIZE=$(du -h "$OUT" | cut -f1)
+  echo "[backup] OK: ${OUT} (${SIZE})"
+else
+  echo "[backup] FAILED: pg_dump returned non-zero; removing partial file"
+  rm -f "$OUT"
+  exit 1
+fi
+
+# Rotation — drop dumps older than RETENTION_DAYS.
+DELETED=$(find "$BACKUP_DIR" -name 'flux_db_*.sql.gz' -mtime +"$RETENTION_DAYS" -print -delete 2>/dev/null | wc -l || echo 0)
+echo "[backup] rotation: kept last ${RETENTION_DAYS} days, pruned ${DELETED} old dump(s)"