Status: 🟡 ATTENTION (infra clean; coverage gaps)

Production aguiarinjurylawyers.com — Cloudways app 6360875 on Vultr ATL server 1615235 — is healthy.

Cloudways — ok

• Server saguiar-prod-vultr-atl (Vultr ATL, 155.138.233.18), running, debian12, 128 GB volume.
• Disk: 97.86 GB free of 128 GB (~76% free). 24h trend -0.6 GB/day (benign).
• Services: apache2, nginx, php8.2-fpm, mysql, redis, varnish, memcached, imunify360 all running. newrelic-sysmond stopped (non-core).
• Backups: backup_frequency=1 (daily). last_backup_at not exposed by server_get → recency not verifiable this run.
• Copilot insights: skipped — 403 "AI Insights optin is required".
• Origin/edge health: origin 200 (0.38s), edge 200 (0.37s), Cloudflare serving, Varnish HIT. ✅
• Note: legacy DO server 1469735 (nyc3) still running with old apps; not production.

Cloudflare — skipped: oauth-pending

Both cloudflare-dns-analytics and cloudflare-observability MCPs expose only authenticate (OAuth not completed; requires interactive session). Cache-hit-ratio / DNS analytics not collected. Cloudways CF add-on not configured at app level (prod uses Autonomous CF Enterprise at platform tier).

Sentry — skipped: oauth-pending

sentry MCP exposes only authenticate. Error counts not collected.

Anomalies

None critical. Watch: newrelic-sysmond stopped; disk -0.6 GB/day (~160 days headroom); 3 monitoring blocks unavailable (recurring).

Drift / self-heal (2 low-severity, proposal queued — not auto-applied)

1. monitoring_app_summary requires type param (bw|db) — pivoted to monitoring_server_summary type=db + disk graph. HIGH confidence.
2. server_get no longer exposes last_backup_at — reported backup_frequency, flagged gap. MEDIUM confidence.

Proposal: infra-daily-pulse/proposals/2026-06-02-v1.md. Mirrored to Codex QA queue.