at weave we have metric called code output where 1 output = 1 hour of expert engineering hour pre ai. output doesnt fully encapsulate the full value from a pr. i had an experiment to try and have agents interact on an exchange to try and price the utility/impact of a pr in relation to the output

lmk if you guys have any thoughts!