My current fave harness. I've been using it to great effect, since it is self-extensible, and added support for it to https://github.com/rcarmo/vibes because it is so much faster than ACP.
Can you shed some light on the speed difference of the direct integration vs. ACP?
I’m still looking for a generic agent interaction protocol (to make it worth building around) and thought ACP might be it. But (and this is from a cursory look) it seems that even OpenCode, which does support ACP, doesn’t use it for its own UI. So what’s wrong with it and are there better options to hopefully take its place?
I've used ACP extensively because agent-shell in emacs uses it, although the Anthropic license change means I'm not sure if I can continue to use Claude through it without getting banned. I kind of wish it integrated more tightly but also you can't really expect someone to have magit involved such that agent-shell (or the like) starts interacting with emacs directly. I'd love it if it did though.
I've started using OpenCode for some things in a big window because its side-by-side diff is great.
Yeah, ACP adds another layer of marshaling/unmarshaling (or two-one on each side) and can be slower than API calls on occasion. Like MCP, it adds JSON overhead that doesn’t really need to be there.
The best option will always be in-memory exchanges. Right now I am still using the pi RPC, and that also involves a bit of conversion, but it’s much lighter.
The better web UI is now part of https://github.com/rcarmo/piclaw (which is essentially the same, but with more polish and a claw-like memory system). So you can pick if you want TS or Python as the back-end :)
The claw version’s web UI essentially has better thinking output, more visibility of tool calls, and slightly better SSE streaming. I’ve backported some of it to vibes, but if you want to borrow UI stuff, the better bits are in piclaw. I use both constantly on my phone/desktop.
No, literally. Mistral, Gemini, opencode, everything supported by Toad, etc. I’ve tried them all. I just don’t like using either Claude Code or Codex, so I didn’t add them to agentbox and stuck with Copilot because it gives me both OpenAI and Anthropic models.
Ok, maybe we need to establish what "literally" means before we try to figure out "all of them" it seems...
I was curious about your project, but the sloppy usage of even the most basic terms kind of makes me not to want to dive deeper, how could I even trust it does what it says on the tin, if apparently we don't even have a shared vocabulary?
I think the thesis of Pi is that there isn't much special about agents.
Model + prompt + function calls.
There are many such wrappers, and they differ largely on UI deployment/integration. Harness feels like a decent term, though "coding harness" feels a bit vague.
When I hear "coding agent", I think of both the harness and the LLM as a pair. Like, Claude Opus 4.6 and Claude Code is a coding agent, or Gemini 3 Pro and Pi is a coding agent.
"Harness" is a way to reference the coding agent minus the "LLM" part.
If an agent is an LLM in a loop with tool calls, there are two components: 1) the LLM. 2) The loop with tool calls. That second part could be called the harness.
