I had a similar failure early on with pyrocat, but mine was instead from making up 1:10 "stock" solutions for the A and B parts to ease mixing up the final dilution. This resulted in vastly decreased shelf life w.r.t. the proper stock solutions. After that, I switched to using Pyrocat-HD in glycol as my stock solution, along with measuring pipettes suitable for the volumes involved.
Also be sure to never cross-contaminate the A part with the B part. Apparently it doesn't take but a tiny contamination with the B part to cause the A to rapidly deteriorate. For me, this means that the pipettes used to measure each part are clearly labeled and stored in different (also labeled) ziplocs when not in use.