Take the number of devs working in a repo, measure the average pull requests across a working week/month/etc, divide the number of dev hours by the average number of pull requests, you’ve got a very rough cycle time.
If you want to plot out each one for datapoints, you’ll need to change how you do work items.