Automations: Best practices
The following best practices help you design reliable, efficient automations and avoid common setup issues. Use them as guidance when building and refining automation nodes to ensure they run as expected.
Use constraints at the right level
You can add constraints either in the trigger node or in a condition node, depending on how you want the automation to behave.
Trigger constraints: add constraints at the trigger level when you want the automation to run only for specific cases. If the constraint is not met, the automation won’t start at all. This helps avoid unnecessary automation runs.
Condition constraints: add constraints in a condition node when you want the automation to start for all cases, but branch into different paths based on whether the condition is met. This allows you to define actions for both the “yes” and “no” paths.
Order nodes carefully
Automations run sequentially, and if a step fails, the automation stops at that point.
To avoid unintended results (such as empty entries), pay close attention to node order.
For example, when creating SEO metadata in a new referenced entry, place the AI Action before the “Create Entry” action. If the AI action fails, no empty entry is created.
Use clear, descriptive node names
You can rename all nodes except the trigger. Use simple, descriptive names so it’s easy to understand what each node does.
For example, when translating multiple fields, rename AI Action nodes to their operation:
“Translate name”.
“Translate description”.
“Translate content”.
This makes it easier to reference the correct outputs later in the automation.
Use Patch (or Edit) multiple AI results in a single node
When using AI actions, use an Edit Entry (Patch entry) action to define where AI-generated content should be written.
If multiple AI Actions generate content for the same entry, you can use one Patch entry node with multiple operations instead of adding a dedicated Edit entry node for each AI Action.