Skip to main content

​
Overview

The Labeling section shows all captured function calls from your SDK integrations. Use it to:
  • Review inputs and outputs from your AI functions
  • Label captures for evaluation datasets
  • Filter and search through your data
  • Build test cases from real production data

​
Viewing Captures

Navigate to Labeling in the main navigation to see all captured function calls.

​
Capture List

The list shows:
ColumnDescription
FunctionThe trace function key or agent name
TypeThe span type (llm, agent, function, guardrail, etc.)
StatusSuccess or error
DurationExecution time
CreatedWhen the capture was recorded

​
Filtering

Filter captures by:
  • Function: Select a specific trace function key
  • Type: Filter by span type
  • Status: Success or error
  • Date Range: Filter by time period
  • Tags: Filter by assigned tags

​
Searching

Use the search bar to find captures by:
  • Input content
  • Output content
  • Error messages

​
Capture Details

Click on a capture to view details:

​
Input

The exact input passed to the function: 
{
  "order_id": "order-123",
  "items": ["item1", "item2"]
}

​
Output

The output returned: 
{
  "order_id": "order-123",
  "total": 20
}

​
Span Information

FieldDescription
NameThe span name (function name or explicit name)
TypeSpan type (llm, agent, function, guardrail, handoff, custom)
DurationExecution time
Trace IDUnique identifier for the trace
Parent SpanLink to parent span (for nested captures)

​
Nested Spans

When functions call other wrapped functions, you’ll see a hierarchical view: 
Trace: agent-workflow
├── Span: validate-input (guardrail)
├── Span: fetch-data (function)
└── Span: generate-response (llm)
Click on individual spans to see their inputs, outputs, and timing.

​
Tagging

Organize captures with tags:

​
Adding Tags

  1. Open a capture
  2. Click Add Tag
  3. Select an existing tag or create a new one

​
Managing Tags

Navigate to Tags in the user menu to:
  • Create new tags
  • Edit tag names and colors
  • Archive unused tags

​
Building Evaluation Datasets

Use labeled captures to build evaluation datasets:
  1. Filter to captures you want to include
  2. Label captures with relevant tags
  3. Export as test cases for your functions

​
Best Practices

  • Review Regularly: Check captures periodically to ensure quality
  • Tag Strategically: Use tags to organize by feature, environment, or use case
  • Build from Production: Use real captures to create meaningful evaluations