Has Claude Code become significantly worse for you as well?
I typically use CC on a Max plan with effort 'auto' and throughout the last week I noticed a significant drop in its quality of work: sloppy mistakes, brute forcing its way through problems.
Is it only me or have things gotten worse?
I’ve mostly noticed a big jump in token usage, the 5 hour limit seems to kick in way faster than it used to, even for similar tasks.