When a monitor's webhook destination is configured as “Critical Only,” Vercel says the agent sends onward only alerts it classifies as Critical. A webhook can point to a paging tool such as incident.io.
The agent checks logs and metrics during its assessment. Vercel gives two examples: health-check errors that do not affect customers, and sustained 5xx errors affecting checkout requests. Teams can attach context to an individual monitor to guide the investigation. The setting can be configured in monitor notification settings or through vercel monitors notifications with --critical-only when linking or updating a webhook destination.
Alert validation is available to Pro and Enterprise customers with Observability Plus. That package includes 10 Agent investigations per billing cycle; additional investigations are billed separately. Custom monitors cost $0.0035 per GB of scanned data, while built-in alerts remain included with Observability Plus.
The announcement does not state the classifier's accuracy, investigation latency, price for additional investigations, or the rate at which alerts are suppressed or escalated.