Fingers crossed for this - it seems like Zed is kinda “going back to the roots” (of working on collaboration) and leaving this in place to disrupt the agentic IDE category (and make themselves not have to spend time on competing in it).
Curious to see how adoption among cli agents will go (it’s nice to see Gemini cli already in).
The level of competition in the LLM and coding assistant market is always nice to see, and this only helps to make costs of switching between offerings even smaller.
I really hope that APC will get the necessary traction from the community, even if I suspect that many standards that are currently proposed in the LLM field will be abandoned soon.
I ask AI to write a feature (or fix a bug, or do a refactoring) and then I read the commit. If the commit is not to my liking, I "git reset --hard", improve my prompt and ask the AI to do the task again.
This way, there is no interaction between my coding environment and the AI at all. Just like working with a human developer does not involve them doing anything in my editor.
Because there is no ai hype in LSP. AI is literally going through the same phase as early javascript libraries did. An explosion of tools, completely ignoring/bypassing previous accumulated knowledge, solving the wrong problems with the wrong tools.
And this will pile up the bloat stack, similar to js libraries, because new problems will rise from the wrong solutions, requiring even more wrong solutions to be incorporated in the stack.
Your response is an exact proof of what I said in my comment. Thanks
Edit: Ok elaboration for the well intended. There is a HUGE overlap of LSP and ACP. LSP solves the problem of "oh we have many IDEs and many language integration plugins, we have an NxM problem, let's make it a N+M". ACP goes to do exactly that but for agents (which is an implementation detail), which means they could very well piggy back on top of LSP
This kind of arguments don't make sense because of scope creep. LSP is for programming languages, why would the maintainers of LSP consider AI agents as a valid language server? And ACP is useful beyond code editors that supporting LSP just for the sake of it makes even less sense.
Claude, come up with a protocol for communicating between AI agents and IDEs/editors. Create node, python and rust libraries. Create a website with a landing page
MCP also started as JSON-RPC over stdio. With solutions like GitHub Codespaces, devcontainers, or "background agents", I wonder if we'll see the development of JSON over SSE.
Currently, my environment uses Claude Code on bare metal, and my application runs in a container, and the agent can do "docker compose exec backend" without any restrictions (YOLO).
My biggest obstacles to adopting workflows with git worktree are the need to share the database engine (local resource constrains) and the initial migration time. Offloading to cloud might be interesting for that.
I don’t see why we need so many protocols. In such a greenfield tech, many are eager to define rules. There’s already a protocol called AG-UI that does the similar thing, but even its purpose isn’t entirely clear to me.
Rather than rushing to create standards, I think the focus should be on building practical implementations, AI-centric UI components that actually help developers design more AI-friendly interfaces. Once the space matures and stabilizes, that’s when standardization will make more sense. Right now, it feels too early.
I don't know about other protocols but for this one it's because AI vendors decided to ship their agents in the form of TUIs, which are not programmable. They need to be wrapped into something to be build on top of.
I actually prefer something like opencode which is one TUI to rule them all (albeit still lacking in some area), it also provides an an SDK that abstracts away the underlying model.
I see they've also caught the RFC2119 bug. This "MUST", "SHALL", "MAY" thing is a linguistic blight on our standards landscape and should be eradicated. HTML5 is written without this unnecessary level of linguistic pretense and works fine.
If your proposed spec is full of "SHALL", "MUST", and "MAY", I'm going to dock you ten points of credibility from the outset. It's a signal you've set out to imitate the vibes of seminal RFC specs without independently considering the substance.
Why don’t they focus on IDE first? I still can’t use their IDE as my daily. They also need to hire someone for UI. It took me 10 minutes to figure out how to bring up their AI chat window.
I really wish Anthropic will adopt this for Claude Code, and I hope it provides at least the kind of integration they have with VS Code. Would be nice if it had access to at least diagnostics from the editor.
I love this idea; I hope it gains traction. One thing that is not clear to me is file search vs unsaved files. It's common for agents to use, e.g., ripgrep to search the file system. But if the communication protocol includes read/write access to unsaved files, there is a desync in terms of accuracy.. rg can't search unsaved files.
So much negativity in this thread. After seeing the recent demo of gemini cli working in zed, I'm actually pretty excited for standardization in this space.
Helix has been my favorite editor ever since I started using it, but the lack of integrated AI features has been becoming more of a deal-breaker for me lately. Though, I can see why the maintainers are hesitant to build for it, because there hasn't been a standardized protocol. I hope ACP changes that.
I’ve been writing a tool to enable Claude Code to use ACP (because I use CC and Zed a lot) and so far I’ve been pretty successful (using the Claude Code SDK and the ACP Client library) although there are some rough edges. I will polish it a little bit and publish it tomorrow I guess.
35 comments
[ 1.6 ms ] story [ 60.1 ms ] threadCurious to see how adoption among cli agents will go (it’s nice to see Gemini cli already in).
The level of competition in the LLM and coding assistant market is always nice to see, and this only helps to make costs of switching between offerings even smaller.
You need to define an interface for Ai to click buttons? Or to create keyboard macros that simulate clicking buttons? We are all doomed!
https://xkcd.com/927/
I ask AI to write a feature (or fix a bug, or do a refactoring) and then I read the commit. If the commit is not to my liking, I "git reset --hard", improve my prompt and ask the AI to do the task again.
I call this "prompt coding":
https://www.gibney.org/prompt_coding
This way, there is no interaction between my coding environment and the AI at all. Just like working with a human developer does not involve them doing anything in my editor.
And this will pile up the bloat stack, similar to js libraries, because new problems will rise from the wrong solutions, requiring even more wrong solutions to be incorporated in the stack.
The question doesn't even make sense, there is no overlap between LSP and ACP
Edit: Ok elaboration for the well intended. There is a HUGE overlap of LSP and ACP. LSP solves the problem of "oh we have many IDEs and many language integration plugins, we have an NxM problem, let's make it a N+M". ACP goes to do exactly that but for agents (which is an implementation detail), which means they could very well piggy back on top of LSP
Currently, my environment uses Claude Code on bare metal, and my application runs in a container, and the agent can do "docker compose exec backend" without any restrictions (YOLO).
My biggest obstacles to adopting workflows with git worktree are the need to share the database engine (local resource constrains) and the initial migration time. Offloading to cloud might be interesting for that.
Rather than rushing to create standards, I think the focus should be on building practical implementations, AI-centric UI components that actually help developers design more AI-friendly interfaces. Once the space matures and stabilizes, that’s when standardization will make more sense. Right now, it feels too early.
I actually prefer something like opencode which is one TUI to rule them all (albeit still lacking in some area), it also provides an an SDK that abstracts away the underlying model.
If your proposed spec is full of "SHALL", "MUST", and "MAY", I'm going to dock you ten points of credibility from the outset. It's a signal you've set out to imitate the vibes of seminal RFC specs without independently considering the substance.
https://news.ycombinator.com/item?id=45038710
https://github.com/zed-industries/zed/blob/main/crates%2Fage...
https://www.npmjs.com/package/@zed-industries/claude-code-ac...
Even though that doesn't seem to be mentioned on the site yet.
Helix has been my favorite editor ever since I started using it, but the lack of integrated AI features has been becoming more of a deal-breaker for me lately. Though, I can see why the maintainers are hesitant to build for it, because there hasn't been a standardized protocol. I hope ACP changes that.