On-Call Agent: Scheduled Service Health Sweep with Pre-Page Warnings

Runs a proactive health check across all your services on a cadence you set. Instead of waiting for an alert, the agent looks for slow-burn degradations — climbing latency, shrinking headroom, rising error rates — and surfaces them early.

Use it as a daily or hourly standup for your infrastructure when you want to catch problems before they become pages. Ideal for teams that prefer to fix things during business hours rather than during incidents.

1. 1A schedule (for example every morning) starts the sweep.
2. 2The agent reads health metrics for each tracked service from Datadog over the recent trailing window.
3. 3Logic scores each service for degradation trend and proximity to known alert thresholds.
4. 4Services that are healthy are dropped; only those trending toward trouble move forward, each tagged with a suggested preventive step.
5. 5The agent posts a ranked watchlist to Slack so on-call can act on the worst offenders first — no remediation runs automatically.