severity 57527 important
merge 47764 57527
severity 53802 important
merge 47764 53802
severity 56466 important
merge 47764 56466

Found a few apparent duplicates (they are all about a &nar-error, read-int and process-stderror).

On second thought, I don't think the patch I referred to #56466 will help here, though it could hardly harm here.

For now, you can try working-around by retrying "guix pull".

There are two hypotheses on the cause in

The second seems most plausible to me, and the first seems the simplest to test (and even if it's not the cause, it would still add some robustness).