Raine Virta

> I have a very expansive allowedTools in my user configs Won't it still ask the initial confirmation that can claude code read and modify the code in this new...

This thread and [wtp](https://github.com/satococoa/wtp) as inspiration I've also made my own tool to suit my workflow. The big difference to others is the tight coupling to tmux. Each worktree gets...

It's not well documented, but it seems like you can write promptfoo configs in TypeScript. This seems to work: ```ts // promptfooconfig.ts import type { UnifiedConfig } from 'promptfoo' import...

It appears to be a side effect of history suppression. If I do `suppress_history: false` I don't see the issue. ``` session_name: helloworld windows: - window_name: dev suppress_history: false start_directory:...

`CLAUDE.local.md` still appears to work

Desperately needed. Option to disable them altogether would be sufficient for me as well. I only click trusted links anyway, so no need for telegram's babysitting.

I mean by trusted link also that the source is trusted to only send legitimate links. All I'm saying is that if the implementation cost of trusted domains is too...

[My use case](https://github.com/raine/workmux?tab=readme-ov-file#agent-status-tracking) is that I need to run a script when user submits their prompt, so that I can show in a tmux status bar that an agent is...