Ask HN: Anyone else finding Claude failures almost unusable?
Its better for coding then OpenAI but it keeps silently failing that blows up my workflow. I've stopped using it this week because I don't want to end up on a dead end street spending hours trying to get back to where I was. Curious if I'm in my own world on this problem? What does a coding chatbot silent fail look like? Silent fail to make good code? Silent fail to make bad code? > because I don't want to end up on a dead end street spending hours trying to get back to where I was. Are you not using git?