I'm experiencing something similar with another piece of software. ledger-cli is a boring, dependable accounting application.

The next release will be the first where the majority of commits will be made by AI, and it has definitely not gone smoothly.

After a dozen or so bug reports, it's mostly in a working state, but I worry the output is no longer reliable in subtle ways.

What specifically has gone poorly? Is a dozen bug reports high or low compared to previous versions of the software where humans are writing the code?

I don't use ledger-cli myself but I do use the similar software hledger. I don't pay very close attention to hledger's development process, but I haven't noticed any bugs that affect me in years of using it.

I think apps like this are the real risk. Who is going to get in trouble because vim* has a bug from AI generated code? Errors in your accounting software can get you into audit/compliance trouble.