Why does only Gemini CLI have "Edits preview" while using ACP?

[marcoafm26]

I've been using Zed for a few months now and I have one simple question: why does only Gemini generate the edit preview in Zed's agent panel? It is a really useful feature to have, as it saves a lot of development time.

I've tested some other agents by ACP like Claude Code, Cursor CLI and Github Copilot. All of them don't have the same behavior of Gemini CLI (rest in peace).

[benbrandt]

This is a Zed feature we implemented on top of the client filesystem apis. However, most agents don't use those, and therefore we can't provide the same diff.

We are looking into other options on the zed side, and we also decided to remove these fs apis anyway in v2 because of lack of adoption.

[marcoafm26]

Thanks for the answer. I'm excited to see what is coming!

[appleton]

Hey there, I'm working on a family of GUI clients for open source release (VSCode, Visual Studio, Desktop app) all of which share the same file versioning underpinnings and using the v1 file system APIs to capture agent edits.

If these go away in v2 I'll need to find an alternative, would you be willing to share the options you're considering for Zed?

Off the top of my head I've got three ideas but I'm not totally sure about any of them:

1. Something based around file watchers or LSP document sync but I can't see how those would differentiate between a manual user edit and an agent edit.
2. Observe diff contents as events are published to the client but I can quickly think of ways that might break down too - for a start it relies on the agent actually publishing diffs.
3. Doing some kind of syscall interceptor wrapper around agent servers to peek at the files being edited by the process. This would have the advantage of also capturing edits performed via a shell command rather than a direct edit tool.

I'd love to hear your thoughts if you're open to sharing?

[gerwitz]

On the other side, I ended up on this thread because I was looking to implement this for my agent harness. I'm sending diffs and locations which provide some nice UX, but don't see differences in those between the current v1 and v2 docs.