[vessenes]

Watched the video. I like this! Really smart idea to get us one step closer to workable human workflows.

Watching you work with it, I found I wanted something that mushes aider and your stuff together. Where I am spending time in aider, I am spending time complaining about claude using bad patterns, assuming bad structure/types/etc or misunderstanding the purpose of a set of code. Where it looks like you are spending time is in fixing coding errors, and manually reviewing patches.

At the very least, it seems like automating / making easy compilation and type checking results would be nice.

Having used aider for over a year now, I know he's spent a lottt of time on prompting customization for both improved code quality and also DIFF format, and I wonder if you would benefit from some of those lessons in terms of getting better code out of different models.

Anyway, this is awesome, and I love the idea of giving some tools to the LLM to engage with the codebase, pull context and then code. Super cool.

[anonymid]

There are lsp tools available to the agent, so in theory it should be able to ask for types, references and diagnostics.

In using it I've found that the agent doesn't really make use of these tools unless I instruct it to. I think I need to do some messing around with prompts to encourage the agent to use them more.

I think the next major thing I want to add is commands to allow you to send symbol types / diagnostics to the agent before it asks for them, which should help speed up some of these workflows.

Thanks for the pointer to take another look at aider - the DIFF format sounds really interesting. Though it seems like the license is Apache 2.0 so I think I'll have to learn a bit about what that means since I made mine MIT licensed.