What it does
Watches the support queue on a schedule and raises the alarm before an SLA target is missed, not after. It reads your actual Zendesk SLA policies rather than assuming a threshold, pulls the per-ticket metrics that say how much time is left, and separates the tickets that need a nudge in a channel from the few that justify waking someone up. The hard part of a watchdog is not detection, it is not being ignored. This one tags every ticket it has already alerted on and excludes tagged tickets from its next search, so a queue running every fifteen minutes produces one alert per ticket rather than ninety-six. Anything it raises carries the evidence: which policy applies, what the target is, how much time is left, and who the ticket is assigned to. It paces itself deliberately. A first-reply target slipping on a low-priority ticket goes to Slack. Paging a human is reserved for an urgent ticket that has already breached or is minutes from it, and the rate limit is set low enough that a broken integration cannot turn into a pager storm. It never replies to a customer, never changes a ticket's priority, status, or assignee, and never resolves anything. It writes exactly one thing into Zendesk: an internal note recording that it alerted, and the tag that stops it from alerting twice.
Example prompts
- Check the Support queue for tickets about to miss their SLA and alert the channel.
- Which urgent tickets have already breached their resolution target?
- Run the queue check and page on-call only if something urgent has actually breached.
Before you connect
The credentials this agent will ask you for — the full setup is on the Setup tab.
Needs 3 credentials (2 required) to connect. See setup
