At the moment we just have some endpoints returning download URLs, and we leave users hanging: * They have to do the wget calls themselves * The requests aren't taken into account by the rate limit algorithm * The request outcomes aren't cacheable