See exactly what your agents are doing—and catch problems before they become incidents.
Dashboards
Pre-built dashboards give you instant visibility:
- Success Rate — See which workflows are failing and how often
- Execution Times — Track how long workflows take (median and worst-case)
- Error Tracking — Identify which workflows need attention
- Resource Usage — Monitor container memory and CPU
- Slow Workflow Alerts — Get notified when workflows take longer than 5 minutes
Real-Time Log Tailing
Watch what's happening right now:
pnpm logs # All logs (last 5 minutes) pnpm logs:errors # Errors only pnpm logs -- --agent=<id> # Filter by specific agent
Filter by time range: 30 minutes, 1 hour, 24 hours, or 7 days.
All Errors in One Place
Every error—whether it happens in your browser or on the server—now flows to the same place. No more hunting through different logs to piece together what went wrong.
Configurable Timeouts
Workflows can now run from 1 minute up to 24 hours. Long-running tasks won't time out unexpectedly, and you'll get alerts if something takes longer than expected.