App alerts you can set, and incidents for Slack and PagerDuty
Each app's Settings → Alerts now sets who hears about production's failures and how often: everyone who can see the app or chosen people, at most one email every 1, 3, 6, 12 or 24 hours, and whether jobs that give up count. Rules can hold alerts until several failures land within a few minutes, email when a development dev server won't start, and raise an alert when a Shopify app's production stops getting Shopify's webhooks. The team's webhooks get apps.production.failed when production starts failing and apps.production.recovered once it's been quiet for 30 minutes, paired by an incident id, so Slack, PagerDuty and other tools open and close one alert.
The emails don't change until you change them: every app keeps one per 12 hours to owners, admins and members who can see it. alerts: { failures: false } in settings.ib.ts still turns everything off.
The incident events follow the same rules as the emails but aren't held, and a Shopify webhook stall comes as the same pair with rule: "shopify_webhooks_stalled". Subscribe an endpoint to them under Team → Webhooks; an endpoint on all events already gets them.
More in Alerts.