During a Tableau Next implementation, a Tableau Next Consultant is explaining to the admin team...

Agent Scoping is specifically designed for situations in which Tableau Agent receives a question without clear analytical context. Salesforce defines Agent Scoping as the administrator-controlled list of semantic models that Tableau Agent can search when context cannot otherwise determine which model should answer the user's question.

Typical examples include Tableau Agent being used in Slack or embedded on a page that does not contain a Tableau Next dashboard, metric, or other asset capable of supplying context. The configured semantic-model list serves as a fallback list. Tableau Agent can then determine which approved model should be used to resolve the user's request.

Agent Scoping does not run indiscriminately for every question, which eliminates B. When Tableau Agent is invoked from an analytical surface that already provides clear semantic context, that existing context guides the interaction. C is also overly restrictive; Agent Scoping is not limited to metric-details pages or scenarios involving multiple models attached to a metric.

Administrators can configure approved fallback semantic models for active Analytics and Visualization agents using the Data Analysis subagent.

References/Topics: Agentic Experiences - > Tableau Agent - > Agent Scoping - > Fallback Semantic Models - > Context Resolution.

===========================================================================