Understand Slack incident notifications
Learn how OpsKnight projects incident lifecycle state to Slack destinations and how updates, identity, and delivery failures behave.
OpsKnight is the source of truth for incident state. Each active service destination receives a projection. Subsequent lifecycle changes update or reconcile that projection using durable provider/message identities and idempotent delivery keys.
The message can expose only actions allowed by the shared ChatOps policy. A provider message is not an independent incident record; manually editing or deleting it does not change the OpsKnight incident.
Routine destinations and war rooms
Routine service destinations receive incident lifecycle messages. A war room is a separate incident-scoped channel with provisioning, membership, context, and cleanup. Configuring one does not configure the other.
Delivery and convergence
Provider throttling, removed bot access, deleted messages/channels, revoked tokens, or worker/database failures can delay a projection without changing the incident. Use OpsKnight delivery/operation state to diagnose and retry safely; avoid manual duplicate posts.
Identity and actions
Viewing a message does not require an OpsKnight identity link. User-attributed actions do. OpsKnight validates Slack signatures/timestamps, maps the Slack user, then applies normal product authorization and incident-state rules.
Related guides
Last updated for v2.0.0
Edit this page on GitHub