· ai-devtools
Vercel Agent validates alerts before paging on Pro and Enterprise plans
Only alerts deemed Critical by the Agent will now reach your paging system, thanks to new validation capabilities in Vercel Agent.
This feature allows the Agent to examine alerts originating from Vercel monitors before they trigger an on-call page.
To activate this filtering, you must first open a monitor’s notification settings and establish or connect a webhook destination for your paging tool, like incident.io. Afterward, you select "Critical Only" next to that specified destination.
Alternatively, developers can configure these notification destinations via the Vercel CLI using the `vercel monitors notifications` command, ensuring to include the `--critical-only` flag when updating or linking a webhook destination.
The Agent assesses the potential impact by scrutinizing logs and metrics. For instance, it can distinguish between health-check errors that present no customer impact or confirm persistent 5xx errors specifically affecting checkout requests.
Furthermore, users gain the ability to inject context into individual monitors, which helps direct the Agent’s subsequent investigation process.
This alert validation functionality is accessible on Pro and Enterprise plans, provided the Observability Plus tier is active. Observability Plus includes 10 Agent investigations per billing cycle, with further investigations incurring separate charges.
Custom monitors are priced at $0.0035 per GB of data scanned, while the standard built-in alerts remain covered under Observability Plus.
