Since Python 3.14 will now allow me to usesys.monitoring for branch coverage, the last obstacle (for me!) remains this package.

I've found a decent workaround: marking every test that renders a template that I want to get coverage on and then running those without the environment variable, separately from the others, and combining the coverage files from the two test runs. Given the timings from my experimentation, this seems well worth the effort.

So this is really just a question: is there anything that can be done in this package or in Coverage.py (or I suppose, in my project) that can work around this in a more general way, without me needing to make this somewhat painstaking change to my codebase?