When the agent calls a tool, an expandable tool call component is added to the interface inside the thought cell.
When expanded, the tool call component shows arguments, an abbreviated snippet of the output, and the full error output if an error was thrown.
The text in the entire component is unable to be selected and therefore cannot be copy/pasted. This would be particularly useful in cases where an error is generated for the purposes of reporting the error upstream.
Screenshot showing effected component:

When the agent calls a tool, an expandable tool call component is added to the interface inside the thought cell.
When expanded, the tool call component shows arguments, an abbreviated snippet of the output, and the full error output if an error was thrown.
The text in the entire component is unable to be selected and therefore cannot be copy/pasted. This would be particularly useful in cases where an error is generated for the purposes of reporting the error upstream.
Screenshot showing effected component:
