InspectDatadog
Start automations from Datadog monitor alerts and warnings.
Connect Datadog
Connect Datadog in Customize → Connectors with an API key and an application key. The Datadog monitor trigger stays unavailable until Datadog is connected. The keys must be able to read and update the monitors you choose and to manage webhooks.
Add the trigger
Select Datadog monitor under Add trigger. Under Datadog monitor, search for a monitor by name, or paste its monitor URL and select Use this monitor URL. Datadog site follows your Datadog connection or the pasted URL; change it before searching if the monitor is on another site.
Under Run for, choose which monitor transitions start a run:
| Option | Transitions |
|---|---|
| Alert | Triggered, Re-Triggered, Renotify |
| Warning | Warn, Re-Warn |
| Alert and warning | All of the above |
Recoveries and other transitions do not start runs.
How Inspect connects to the monitor
When you save, Inspect checks that the monitor exists and that the keys can access it. It then creates an Inspect webhook in Datadog and adds its @webhook-inspect_… mention to the monitor's message. The trigger shows Setup pending while this happens and Active once the monitor is ready. If setup fails, the trigger shows the error; select Retry setup after fixing it.
Leave the @webhook-inspect_… mention in the monitor's message; removing it stops the monitor from starting the automation. Inspect removes the mention when you delete the trigger.
Reply in the Datadog Slack thread
Datadog monitors often notify a Slack channel too. Keep Reply in the matching Datadog Slack message thread selected, and set the automation's Default Slack channel to that channel, to keep the investigation next to the alert. Inspect looks for the monitor's Datadog notification in that channel and replies in its thread. Clear the option to post as a new message instead.
What starts a run
Each matching monitor notification starts one run. The agent receives the monitor name and ID, transition, status, message, tags, the time it occurred, and a link to the event in Datadog. Repeated deliveries of the same notification do not create another run.
Example
Pair a Datadog monitor trigger with a Production Triage agent: on each alert, it reviews the relevant logs, traces, and recent deploys, then replies in the monitor's Slack thread with its assessment and a suggested next step.