Traces represent complete end-to-end request flows through your AI agents and workflows. Each trace contains one or more spans that represent individual operations. The Monitor pillar lets you view, filter, and deep-search traces to understand how your application behaves in production.Documentation Index
Fetch the complete documentation index at: https://www.adaline.ai/docs/llms.txt
Use this file to discover all available pages before exploring further.
View traces
In Adaline, each project represents an AI agent or workflow. All traces sent to Adaline are associated with a project. Navigate to the Monitor section of your project to see all traces:
Inspect a trace
Select any trace to view its details. Adaline provides two visualization modes:Trace tree
The tree view shows the hierarchical structure of all spans within the trace — making it easy to understand parent-child relationships between operations:
Waterfall view
The waterfall view shows spans as a timeline, revealing concurrency and sequential dependencies at a glance:
What each trace shows
When you select a trace, you can view:| Section | Details |
|---|---|
| Trace tree / waterfall | Hierarchical or timeline view of all spans in the trace. |
| Name | The name of the trace. |
| Started at | The timestamp when the trace started. |
| Ended at | The timestamp when the trace ended. |
| Status | Overall trace status (success, failure, unknown). |
| Total duration | End-to-end time for the complete trace. |
| Total cost | Cumulative cost of all LLM spans in the trace. |
| Total tokens | Sum of input and output tokens across all LLM spans. |
| Tags | The tags of the trace. |
| Attributes | The attributes of the trace. |
| Session ID | The session ID of the trace. |
| Reference ID | The reference ID of the trace. |
Filter traces
Use filters to narrow down traces across large volumes of data. You can filter by time range, status, duration, cost, tags, attributes, and session ID — and combine multiple filters to zero in on exactly the traces you need.
Export traces as JSON
Use the Export JSON button in the project Monitor view to download the current trace result set, or call the public API to export traces programmatically with cursor-based pagination. See Export Logs for the request fields, response shape, and a pagination loop example.Next steps
Analyze Log Spans
Inspect individual operations within a trace.
Analyze Log Charts
View aggregated metrics and trends.