The tool currently doesn't complain if the test definition reference datasets (input or output) that don't exist in the workflow. This behaviour makes it difficult to find small typos in the dataset names, and may result in tests that wrongly pass (because the output to be verified wasn't properly referenced).
The tool currently doesn't complain if the test definition reference datasets (input or output) that don't exist in the workflow. This behaviour makes it difficult to find small typos in the dataset names, and may result in tests that wrongly pass (because the output to be verified wasn't properly referenced).