Notifications¶

Stay on top of what your platform is doing without watching the screen. Every user configures their own channels from the bell menu in the header (Notifications modal).

What triggers a notification¶

Job outcomes — a deploy, rebuild, backup or any other operation finishing (successfully or not) — and alerts. Alerts come in two levels: critical (host unreachable, instance down, disk critical…) and warning (a container of an instance down, errors in an instance's log, a declined subscription payment…). Failed deploys and rebuilds are treated as severe and always delivered immediately.

An open warning that turns critical is announced again, on every channel, to everyone whose level covers critical alerts.

Your preferences¶

Level — decides both which job outcomes and which alerts reach you:

Level Jobs Alerts

All every outcome critical and warning

Failures and warnings failed only critical and warning

Failures only (default) failed only critical only

None none none

The daily digest follows the same level. - Muted projects — silence everything from projects you don't care about. - Notifications only cover what you can already see: project visibility rules apply to every channel.

Channels¶

In-app (always on)¶

Toasts for finished jobs and critical alerts, plus the alert badge in the header. Real-time — no configuration needed.

Email¶

Toggle it on and pick a mode:

Immediate — one email per job outcome or alert.
Daily digest — a single summary email every morning. Severe failures (deploy/rebuild) and critical alerts still arrive immediately.

Telegram¶

Create a bot with @BotFather and paste the bot token.
Send any message to your bot, then click Detect chat ID — the platform finds your chat automatically.
Send test to verify. Done: job outcomes and alerts arrive as Telegram messages, following your level.

Webhook¶

Point the platform at any HTTPS endpoint to integrate with Slack bridges, on-call tools or your own automation. Two event types arrive as JSON POSTs — job_state_change for successful job completions, and alert for failures and infrastructure alerts:

{
  "event": "job_state_change",
  "job_id": 123,
  "job_name": "Deploy Instance",
  "state": "done",
  "severe": false,
  "host": "host-1",
  "instance": "production",
  "log_url": "https://…/cloud/log/123",
  "timestamp": "2026-07-12 10:00:00"
}

{
  "event": "alert",
  "alert_id": 45,
  "code": "job_failed",
  "level": "critical",
  "message": "Deploy Instance on production failed",
  "host": "host-1",
  "instance": "production",
  "job_id": 123,
  "log_url": "https://…/cloud/log/123",
  "timestamp": "2026-07-12 10:00:00"
}

When an alert clears automatically (the monitor that raised it observes recovery), the same payload shape is sent again with "event": "alert_resolved" — consumers can close the incident they opened on the alert event. Dismissing an alert by hand in the panel does not send this event: the operator silencing it is already looking at it.

If you set a secret, every request carries an X-IncubaCloud-Signature: sha256=<hex> header — the HMAC-SHA256 of the raw body. Verify it with a constant-time comparison before trusting the payload.

Backup usage alerts¶

Separately from these notifications, each backup backend can email a warning when bucket usage crosses a threshold (default 80%). Configure it on the backend detail page.

Level	Jobs	Alerts
`All`	every outcome	critical and warning
`Failures and warnings`	failed only	critical and warning
`Failures only` (default)	failed only	critical only
`None`	none	none