The problem is with the inputs. Run FastQC and you'll see the sequence problems. The data has extremely high ambiguous base calls (NNNs) for all 6 inputs in the working history.
In short, there is not enough non-ambiguous base calling content have anything to assemble. This also means that trimming will not be enough to produce a successful assembly. Reads on average only have about 8-10 non-ambiguous bases of content. It is unique content, but very short.
Note: FastQC checks just the first 200k sequences (a sample of the data). So, I also ran it on the last 100k sequences just to make sure there wasn't a problem at the start of the sequencing run (not uncommon with certain technologies). But, that data also produced the same type of QC results.
Hopefully, this helps to explain the current problems (empty results).
Thanks! Jen, Galaxy team