Skip to main content

Expanding and collapsing Runchat nodes

A Runchat node with its More menu open showing Expand workflow
You can publish any Runchat workflow as a tool and then use it as a node within another workflow. These are Runchat nodes. After you run a Runchat node, the internal state of the workflow (all input and output parameters) is saved on that node. You can view this internal state, called a subflow, to see exactly how data flowed through the workflow. Open the node’s More (…) menu and choose Expand workflow. Choose Collapse workflow to fold it back into a single node.
An expanded Runchat node showing its internal workflow: a note and a Code node with its last inputs and output
Expand workflow only appears in the menu once the node has run and has internal state to show.

Expanding and collapsing Agent nodes

You can also expand an Agent node to view the internal state of any tools it used to answer your prompt. Click Tools on the Agent node’s toolbar to add tools, then run the node. If a tool was used, Expand workflow appears in the node’s More menu.

Internal state

The expanded view shows the internal state of the tool when it was last run. If you run the parent node (Agent or Runchat) again, the inputs and outputs of the child nodes update to reflect the new state. This is a useful diagnostic tool for detecting bugs in workflow logic or flaws in prompts.

Editing subflows

You can make changes to nodes in a subflow (for example to Code or Agent nodes) and these are saved as a custom override to the tool. This is useful for quickly exploring variations on prompts or fixing bugs.