aider

Claude Code is fairly notoriously token inefficient as far as coding agent/harnesses go (i come from aider pre-CC).

Intent alone is not generally sufficient for a conviction is the US, save for some cases of special liability (attempts, or some forms of aider and abettor or conspiracy).

So you can configure aider or vscode or whatever you're using to connect to chatgpt to point to your local models instead.

Things like aider use prompt toolkit and loses the layout when you resize window.

I briefly dabbled with Aider some months back but never got any real work done with it.

Am I misremembering cause I could have sworn that Aider started out as a nvim plugin??

The best survival rate I could find was 30%, which is where CPR is given immediately by a trained first aider.

Could you say more about the differences between Aider and OpenCode?

Any agent is in theory supported, but it has specific support for Claude, Gemini, codex, aider, and opencode.

Seems like it'd work with aider, OpenCode or even vibe.

This replaced Aider for me a couple months back.

But having even a mediocre version of the same available as software for each smartphone will only improve what the casual user or even first-aider is capable of.

Aider did an "architect-editor" split where architect is just a "programmer" who doesn't bother about formatting the changes as diff, then a weak model converts them into diffs and they got better results with it.

And token spending dropped considerably: Works today with Claude Code, Codex, Aider, and Gemini CLI.

It's a local proxy (npx @sliday/tamp) that sits between your coding agent (Claude Code, Aider, Cursor, Cline, etc.) and the upstream API.

Can anyone clarify how this compares with Aider?