Automation was supposed to make science more reproducible: a robot runs the same steps the same way every time, without a tired hand or a bad day. It delivers on that within a single deck. Across labs, the picture is weaker, because the thing that actually determines the transfer, the liquid class, usually stays locked inside one instrument or one lab set of files. A protocol you can read but a class you cannot see is only half a method.
The hidden half of a method
Publish a protocol and you have shared the steps. But the steps assume liquid classes that are rarely published with them, so a reader can see that you transferred 20 microliters of a reagent without knowing the flow rate, air gaps, and delays that made that transfer accurate. Two labs running the same published protocol with different classes can get different results and never know why. The reproducibility gap hides in the parameters no one shared.
What openness adds
Sharing the classes, not just the steps, closes that gap. When a class is open and inspectable, a colleague can see exactly how a transfer was performed, judge whether it fits their instrument, and reproduce or adapt it with eyes open. Openness also compounds: a class one lab validated saves every later lab the work, and mistakes get caught by more people than the one who wrote them.
Open does not mean unaccountable
Sharing widely raises the value of knowing where something came from. An open class still needs its provenance, the instrument, tips, and volumes it was validated on, and its version history, so that openness spreads trustworthy knowledge rather than orphaned numbers. Open and traceable are partners, not opposites.
A protocol without its liquid classes is half a method. Share the parameters that made the transfer work, or reproducibility stops at your own deck.