Agent Run Inspector¶
The Agent Run Inspector helps you understand what a visual agent is doing during a run.
It is available from the Design tab of saved Simple Visual Agents and Structured Visual Agents, where you can inspect a run while testing an agent.
Overview¶
The Inspector is a live debugging view for a single run. It shows the execution steps that produce the agent’s response.
Use it to:
Follow the execution of a run while it is still in progress
Identify which tools, blocks, and sub-agents were used
Inspect the inputs, outputs, and metadata of block and tool executions
Review changes to the agent state and scratchpad
Troubleshoot unexpected agent behavior
Enable the Inspector¶
To display execution events in the test chat, select Show timeline in the test window. Select an event to view more details in the Inspector panel.
Inspector views¶
The Agent Run Inspector provides:
A live graph of the run
A live timeline of execution steps in the chat
A right-side inspection panel for block and tool executions
The timeline contains:
Block rows
Tool rows
Agent rows for delegated sub-agent calls
Text rows emitted during the run
For block events, the inspection panel shows the block’s inputs and outputs, along with any changes to the agent state or scratchpad.
Sub-agents are color-coded to distinguish their activity from the current agent.
Warnings and errors¶
The graph and timeline display warnings and errors detected during a run.
When a child block fails, its error details remain available and its parent blocks are marked as failed.
When a tool raises an exception in a nested branch, the exception is propagated to ancestor blocks as a warning.
When a nested sub-agent fails, a warning appears on the sub-agent row and on the calling blocks.
Limitations¶
The Agent Run Inspector is available only in the Design tab of saved agents.
It is not available for Code Agents.
Execution reporting for Parallel blocks is not live.
The Inspector is not available in “Test Query” mode.