At Horizon Insurance, the customer service team is reporting that the agent consistently routes claims...

The correct answer is C because persistent routing to the wrong subagent usually indicates semantic overlap in subagent purpose, description, or entry conditions. Agentforce uses subagent descriptions and instructions to determine which capability is relevant to the user’s intent. If “Claims Intake” and “Policy Inquiry” both describe insurance policy details, claims eligibility, or coverage questions too broadly, the reasoning engine can classify claims requests into the wrong subagent. Procedural instructions and Flows are not the normal root cause of intent classification failure. Global instructions can influence overall behavior, but they should not be used as the primary routing definition for competing subagents. Salesforce Agent Script guidance emphasizes that topic or subagent descriptions help the model understand when a topic is relevant during classification.