Alerts

Alerts watch a Spanly metric and notify you when it crosses a threshold. Rules are scoped to an environment and can target a single server, a single client, a specific tool, or the project as a whole.

Metrics

• Error rate: share of failed requests in the window.
• p95 / p99 duration: request duration percentiles.
• Volume drop: traffic falls below a floor. Use as a deadman switch for "production went quiet".
• Volume spike: traffic exceeds a ceiling. Useful on sensitive tools.

Notification channels

• Email: one or more addresses per channel.
• Slack: paste a Slack incoming-webhook URL.
• Webhook: POST a signed JSON payload to any HTTPS endpoint for routing into PagerDuty, Opsgenie, or your own incident system. The signing secret is generated when you create the channel.
• In-app: surface the alert in the Spanly dashboard.

Notification channels require the Pro plan or above.

Typical rules to set up

• Error rate above 5% over 5 minutes on a critical server.
• p95 request duration above 2 seconds on tools/call.
• Volume drop on a production client (deadman switch).
• Volume spike on a destructive tool such as delete_file.

Availability

Alert rules and all four channel types are live today. You can manage them from the Alerts tab in your project, or programmatically through the Spanly MCP server. If you need a rule type that isn't there yet, email support@spanly.com.