Hi @zocket, I was having a look at the example data provided and tried to keep the non-duplicate values using the Merge and Item Lists node but couldn’t come up with a working approach that would only let you keep unique values.
So unless someone else has a nice idea for this that has not come to my mind I’d write a quick helper function counting the occurrences of each ID. You could then use the IF node to filter those with only one occurrence.
Example Workflow
Does this help you?