It is a pain point to know failure solution dependencies  missing in Power Platform pipelines after few hours post run .

I experienced same thing during implementation, and then I looked at alternatives present to get out of this bottleneck situation.

Hack is to use normal export functionality to the target environment after selecting respective solution in the source dataverse environment could be configurations/flows/plugins etc.

If there are any dependencies missing then we get direct IntelliSense from the framework to add missing components.

We can add those dependencies and in the next screen we get the option to export.

We must stop there and come back to our pipelines and trigger pipelines once again so that our deployment will be smooth.