Ask HN: Has Claude Code quality dropped recently for anyone else?
I’m a daily user of Claude Code (Pro subscription) and it’s been my primary tool for the last few months.
Over the past week or so, I’ve noticed:
More shallow reasoning
Ignoring parts of context
More confident-but-wrong answers
Slight regression in structured refactors
This is mostly in real-world coding tasks (mid-size projects, not toy prompts).
Could just be my workload getting more complex — but it feels different.
Has anyone else noticed a shift in quality recently? Or is this just variance + perception bias on my end?
Would love to hear if others are seeing similar patterns (or not).
No comments yet.