I have a similar, although not quite the same setting. In your circumstances, what I am doing will work as follows: job A stores the test results (say junit xml) with the key of its build identifier and runs C through Parameterized Trigger Plugin , passing it the location of the test results. Job C can then either simply post these tests or do some extra processing. Job B does the same as job A, with regard to his tests and getting started with C. Then all your results are aggregated in job C.
Additional processing, in which task C may include saving test results A to a temporary location, and then processing them later with the results of task B. This is not automatic, but still much easier than writing a completely new plug-in. You can also configure it with any way.
source share