Action Nodes
- Every attribute is passed through* but in order to use an attribute in a specific action node it must be mapped in to it as an input parameter.
- *although there can be exceptions such as the converge action,
- Once a mapping is configured you it will reused in all instances of the actionflow. This is because the action nodes sit within the actionflow itself and only the data coming into and out of an actionflow changes (for each instance)
Action nodes are the building blocks of Actionflows, they are represented as a circle with an icon. They perform specific tasks, e.g. updating data, performing calculations, and API integrations, and are wired together.
Action nodes can be dragged from the toolbar onto the canvas to create a new action, and some can be clicked to see existing objects, e.g. click 
Open Screen to list existing screens, then drag from the Repository onto the canvas.
Most nodes run once per record they receive on their driving connection point. However, some nodes only trigger once after receiving the final record or once per flow, e.g. Open Screen, 
Start Phase and 
Open URL nodes.
Usually, action nodes have 
Driving connections, which pass data through to the next action node. Some action nodes have a 
Lookup connection, which send out a request for data, that is then returned to the action node data.
Nodes must be connected sequentially starting from the 
Driving interface connection point. Records will then pass through each node and be manipulated accordingly.
For most action nodes, data flow is configured when attributes are mapped to the input connection points. For 
Gateway nodes, the data flow can be controlled by mapping to the output connection points. This is because the expressions in a gateway provide the logic for which path the data flows, and different paths may require different data to pass through the gateway.
