Every major editor now ships some version of "describe the change, get a diff." The marketing language is nearly identical across all of them. What actually differs is where the agent's context comes from: some re-read the whole open file on every request, some maintain a persistent index of the repo, and a few quietly send far more of your codebase to the model than their docs admit.
The other real difference is failure behavior. When an agent gets something wrong, does it silently commit a bad edit, or does it stop and ask? We found real spread here — the tools that ask more questions feel slower turn-by-turn but produce fewer edits you have to manually revert later.