Explorer¶

After an application is created, the system aggregates each Agent / LLM request as a log event into the Explorer. You can search, filter, and analyze Session and Trace data for a single application or all applications.

The top of the page supports filtering by application, allowing you to view data for all applications or a specified one.

Viewing Modes¶

Session List¶

Displays data by Session dimension by default. The list fields include Session ID, Input Tokens, Output Tokens. The quick filters on the left support filtering by dimensions such as Token range, model provider, request model, response model. Supports various analysis views like list, Top List, Time Series, pie chart, treemap, grouped table chart.

Click on a row to enter the Session details page:

Summary: The left side displays the Session Trace waterfall chart, and the right side displays the preview and attribute information.
Call Analysis: Displays model call proportion, tool call proportion, Skill call proportion, and the list of Traces associated with this Session.

Trace List¶

Switch to the Trace List Tab to view data by Trace dimension. Fields include Trace ID, associated Session, Trace duration, Input Tokens, Output Tokens, Trace status.

Click on a row to enter the Trace details page:

Summary: Continues to use the waterfall chart display, with Span details below.
Call Analysis: Displays tool call records and Skill call records.

Details Page¶

Upon entering the details page, the left side displays a single Span and Trace generated by the application, and the right side displays the corresponding preview and attribute information.

Options¶

Search and locate based on the node name source.
Choose whether to display metrics and scores.

What are metrics and scores?

Metrics: Display of duration time.
Scores: Model quality evaluation scores, descriptions, and other information.

Preview¶

Includes node type, node name, execution duration, Token consumption, execution status, execution details.

Status Display¶

Execution Status (`status`)	Possible Causes
Success: Success
Error: Error	Agent / LLM call timeout, tool returning exceptions, etc.
Warning: Warning	Output content triggered sensitive word detection but did not interrupt execution, etc.

Execution Details¶

Displays the Input and Output within the application. The former contains all elements used to construct the final request sent to the model, while the latter displays the result generated based on the input.

Attribute Information¶

Relevant field information contained in the current Span or Trace.

Associated Views¶

Using the + button in the upper right corner, you can bind built-in views to the current details page to achieve specialized display of Agent / LLM data.

Click the add button.
Select the associated field.
In the pop-up Bind Built-in View panel, select the view you want to associate.
After successful binding, the newly added tab will be displayed in the tab bar.