@mcumming was using our tech to build containers but wanted to power detections based on the digest of the base image. Initially, I suggested that he read the ContainerLabel items after the publish to find the base image and use that - however we do not actually create the matching Item, we only synthesize it during the image creation. We should ensure that this is added to the MSBuild environment when we create the rest of the CreateNewImage Task's Outputs.

In addition, he bound to the _PublishSingleContainer Target because that target contained the structure with the generated configuration, which he could then dig into. We should instead ensure that the PublishContainer target Returns the same structures that _PublishSingleContainer and _PublishMultiArchContainers does so that the same 'entry point' can be consumed regardless of publishing style.