Multi-channel alerting features for on-call teams

Send a single alert to Telegram, voice, SMS, email, WhatsApp and Viber in one declarative policy. Channels run in parallel per step.

Chains run until someone acknowledges. Sources that send a resolved hook cancel the chain automatically; manual ack stops it from any channel.

Define an alert template once, fire it from any service with one POST /from-template/{slug}. Versioned, slug becomes a stable contract between your monitoring stack and WardenPoint. Starter library included: db-down, slo-breach, deploy-failed, backup-failed — clone, edit, ship.

API keys per integration, UUID-only public IDs, SetCurrentCompany middleware on every authenticated route. Sign in with Google, GitHub or Microsoft (Azure AD, multi-tenant). No cross-tenant leakage in tests.

Every dispatch, ack, escalation and resolve writes a JSON line. CSV export per recipient group; linked to application logs via request_id. Anomaly detector watches our own delivery rate, fires meta-alerts when credentials break, recipients become unreachable, or quota runs out — so you find out before your next real incident does. Active health probes GET each source's health-check URL every minute — a non-2xx response or timeout alerts the on-call group before a dead monitor hides a real outage. For private or VPN-only sources, dead-man heartbeat flips the model: your monitoring tool pings WardenPoint on a schedule; if the signal stops, we page you — so if your monitoring itself goes dark, we notice.

Every customer-facing service has its own 90-day uptime band updated every minute: ingest API, dashboard, each delivery channel, escalation queue, database, Redis. Incidents post live; email / Telegram / webhook subscriptions push updates the moment we change status. A drop-in JS embed shows our state on your own ops dashboard. JSON API + RSS feed for machines, postmortem pages for after-action review.

Limits running low, a failed payment, a finished migration import, a sign-in from a new device — system notifications surface in the dashboard bell and a dedicated center, grouped by category (billing, limits, import, security). Per-role visibility keeps billing and security alerts in front of the people who own them.

Give each integration a unique inbound address and point a monitor's alert email at it — the subject becomes the alert, keywords set severity, and unacknowledged alerts escalate to calls and Telegram. Closes the per-service alert-email parity gap with PagerDuty and Opsgenie.

Setup Hub walks you through a branching wizard — pick a path (inbound monitoring, outbound webhook, REST API, or channel binding), and we bootstrap recipient + group + integration + curl/python snippets + a live test event. Most users land a working webhook from registration in under a minute.

An auto-generated map of your entire alert topology — routing rules, thresholds, escalation steps, on-call schedules with a live 'who is on duty now' badge, and every fallback down to the owner-email safety net. Rendered from your live config, so it can never drift from what actually executes. Ask it "what happens to this alert?" — the simulator dry-runs your live rules and highlights the exact path, including who is on duty at any moment you pick. Edit rules, group membership, and on-call rotations right on the map — changes apply instantly and the map re-renders from the live config.