Usage & Monitoring#

Nebula tracks execution metrics for every agent invocation — models used, tokens consumed, estimated costs, and errors. This data is available in the Settings overview.

Usage Dashboard#

The Settings → General tab shows an org overview with usage statistics.

Usage dashboard

Org Overview#

• Total agents
• Active tasks
• Total messages
• Total executions

Usage Metrics#

Displayed for both 30-day and all-time windows:

• Total executions
• Total tokens consumed
• Estimated cost
• Top models by execution count
• Top agents by execution count

Error Tracking#

The Settings → System tab shows recent execution errors.

Error log

Each error entry includes:

• Timestamp
• Agent name
• Model used
• Error message

You can dismiss individual errors or clear all errors at once. Errors are paginated (50 per page, max 200).

What Gets Tracked#

Every agent execution logs:

• Which model was used
• Token counts (input + output)
• Estimated cost (based on model pricing)
• Success or error status
• Agent and conversation context

This happens automatically — no additional configuration required.