View Agent Reasoning

For every AI Messaging Agent conversation in Agent Studio, you can now view an inline reasoning timeline that shows exactly how the agent arrived at its response. This includes which prompt block was active, which knowledge sources were cited, which tools were called, which guardrails fired, the agent's confidence level, response latency, and a verbatim summary of its reasoning.

For Who? Users of Invoca AI SMS Agents. Currently available to customers in the US.

Availability & Implementation: Live now. No setup required. The reasoning timeline is available directly within Agent Studio.

‍

• Self-serve audit of any agent response in seconds, without engineering involvement
• Eliminates common questions and concerns like, "Can I trust and audit this agent?" and "Where did the agent get this information from?"
• Full transparency into agent decision-making, from knowledge sources to tool calls