Agentforce Observability & Tracing FAQs

Question 1

What is the Session Tracing Data Model (STDM)?

Accepted Answer

STDM is the standard Salesforce data model used to track Agentforce interactions, capturing detailed records of sessions, interaction turns, execution steps, and messages.

Question 2

What are the prerequisites for using this skill?

Accepted Answer

You need a Salesforce org with Data 360 and Agentforce Session Tracing enabled, as well as configured JWT or ECA authentication for Data 360 access.

Question 3

How does this differ from the Agentforce testing skill?

Accepted Answer

While testing focuses on running formal regression loops, this observability skill is designed for extracting and analyzing telemetry from actual sessions to find root causes for unexpected behavior.

Question 4

Why does this skill use Parquet and Polars?

Accepted Answer

Parquet provides efficient storage for large telemetry datasets, while Polars allows for high-performance, lazy analysis of complex session trees that would be difficult to process in standard JSON.

Question 5

Can I analyze GenAI Trust Layer records with this?

Accepted Answer

Yes, this skill can incorporate GenAI Trust Layer audit records to help debug content quality and generation-related issues within the session timeline.

Agentforce Observability & Tracing

Key Features

Use Cases

Agentforce Observability & Tracing

Key Features

Use Cases