Monitor Skill

/armor:monitor

You: Monitor the orders table hourly

Claude: Created freshness schedule:

Table: warehouse.public.orders
Check Interval: hourly
SLA Threshold: 2 hours
Alert Severity: warning

You'll be notified if the table doesn't update within 2 hours.

You: Monitor customers with a 30-minute SLA

Claude: Created freshness schedule:

Table: warehouse.public.customers
Check Interval: every 15 minutes
SLA Threshold: 30 minutes
Alert Severity: critical

Tight SLA configured. Consider if this is production-critical.

You: Set up full monitoring for my gold schema

Agent: Done. Here's what I set up:

Freshness (8 tables):
- gold.fact_orders: hourly, 2h SLA
- gold.fact_payments: hourly, 2h SLA
- gold.dim_customer: daily, 24h SLA
... and 5 more

Schema monitoring: enabled on all 8 tables

Want me to set up a Slack alert for freshness violations?

You: Yes, send to #data-alerts

Agent: Done. Created:
- Slack destination: #data-alerts
- Alert rule: freshness violations (warning + critical) -> #data-alerts
- Alert rule: schema changes (all severities) -> #data-alerts

Your gold schema is fully monitored.

You: What monitoring should I set up for the orders table?

Claude: Based on the orders table update history:
- Updates every ~45 minutes during business hours
- Recommended: 30-minute check interval, 90-minute SLA
- This balances alert noise with detection speed

Want me to create this schedule?