Comment by nurettin
14 hours ago
Claude is also good at that. I made a habit of asking "are you sure?" after a complex task. It usually says it overlooked something.
14 hours ago
Claude is also good at that. I made a habit of asking "are you sure?" after a complex task. It usually says it overlooked something.
I find both to be true. I use Claude for most of the implementation, and Codex always catches mistakes. Always. But both of them benefit from being asked if they’re sure they did everything.