Notifications

Mobius emits platform events — an incident opened, an IR action pending approval, ingest paused, a connector gone unhealthy, billing thresholds — and a routing layer decides where each goes.

Routing rules

In the notifications settings, your organization's admins configure a matrix of event types × channels: for each event type, choose whether it notifies in-app (the bell), on a connected channel, or both — and who receives it.

Under Platform rules, toggle which platform events reach you: workflow started / succeeded / failed, a new environment added, a user invited, billing events (invoice paid or failed, trial ending), and budget thresholds (85% / 90% / 100%). Each workflow's own output is routed separately, from that workflow's page.

Every event type carries a severity, which drives the color in the bell and the chip in the history: bad news never reads as routine.

Delivery channels

Under Channels, connect where notifications land beyond the in-app bell:

• Slack — paste the incoming-webhook URL for the channel where notifications should land.
• JIRA — your JIRA URL, project, email, and API token; matching events open tickets in that project.
• Email — notifications and report sharing by email. During the beta, outbound email runs in a limited sandbox, so recipients are verified with you first.

Only users with the admin role can change integrations and notification settings.

The bell and the history

• The bell shows recent notifications with severity chips; each one deep-links to the object it is about (the incident, the environment, the approval). It only carries the latest few.
• The history is the queryable archive — every notification your organization has received, newest first, with the page each one opens. Filter it by family (Ingestion, Workflows, Response, Organization, Billing, Budget), search by title or message, and page back through older ones. It is where the bell's View all notifications link takes you.

Recommended baseline

Event	Route
Incident opened (high risk)	in-app + email to the on-call
IR action pending approval	in-app + email — approvals are time-sensitive
Connector unhealthy / ingest paused	in-app + email to the environment owner
Billing thresholds	email to the org admin

Everything else can stay in-app only; the history keeps it auditable.