It definitely feels like a jump in capability. I've found that the long term quality of the codebase doesn't take nosedive nearly as quickly as earlier agentic

It definitely feels like a jump in capability. I've found that the long term quality of the codebase doesn't take nosedive nearly as quickly as earlier agentic models. If anything it's about steady or maybe even increasing if you prompt it correctly and ask for "cleanup PRs"