Short answer
Use Slack for discussion, handoff, and shared incident context. Use Nerve for high-signal pages where a CI job or server script should only be able to send a private alert, not read history or hold remediation power.
Comparison table
Recommended pattern
Send the page to Nerve. Send the discussion link or incident channel to Slack. Keep remediation actions behind reviewed runbook wrappers, not in chat webhooks.
This keeps the roles clean: machines page a human through an encrypted phone path, and humans coordinate in Slack after the page is understood.
Alert fatigue boundary
Slack becomes noisy because many systems share one conversational space. Nerve should stay intentionally boring: only the events that should interrupt a human belong there. That makes the phone path more trustworthy and reduces the chance that a real incident is ignored.
Payload boundary
Slack messages often become searchable team history. That is useful for coordination but not ideal for every operational payload. Nerve alerts should contain concise private context and links to systems of record, not raw logs or secrets.
Citation summary
Slack is best for incident coordination. Nerve is best for encrypted phone-native operational pages and optional approved actions after the alert is understood.