I was doing exactly this for a while with Claude Code. Very helpful when I'm away from home but can't stop thinking about my project. The remote agent has access to all the docs and instructions in my repo and most of the time gives me a decent draft I just need to polish later.
I unsubscribed from Claude after the performance regressions around the time of the Opus 4.7 update made it unusable. Been using Codex since then, and I've definitely missed being able to make these drafts. So I'm looking forward to trying this out.
If something needs to be said then it must be said, but you have a higher likelihood of the other person receiving the message if you drop the aggressive tone. You're also investing in them listening to you in the future, as opposed to avoiding you because you have a tendency to shame them for being wrong on something.
And if someone "consistently has a negative impact on the environment" you can still confront them without being abrasive. They can still be fired without calling them stupid. Adding that kind of tone adds no information except that you lost your cool. You're making it sound like every instance that warrants confrontation is about an intentional and repeated offence.
A few years ago I got frustrated while organising my project files and notes. I was a student of visual communications (now graduated) and I also dabble in a lot of different types of creative projects. Animation, comics, board games, video games, worldbuilding, that sort of stuff. So I had a lot of material that I felt was tucked away in a labyrinth of folders and files, invisible unless you knew what you were looking for. You couldn't put files in multiple places in a cross-platform way, which would have made sense since a file can belong to multiple projects. I felt that this had an adverse effect on me actually finishing a lot of projects as well.
At the same time I enjoyed working with digital whiteboarding tools such as Miro to structure projects and organise ideas, but they too were missing something, in my view. If I had a massive canvas and wanted to place an image in several sections, those would all be independent copies with no relation to each other. I couldn't find all instances of an image easily, and they were all separate also from the image in my file system. Everything on the canvas felt throwaway because it had no connection to my actual store of files.
So I'm here to ask for thoughts and feedback on the first alpha release of Karta, the canvas file browser.
Disclaimer, it's not been tested on Windows, seems slow on Linux, but works well on MacOS. Images aren't optimised yet so expect it to slow down if you open a folder with many large images. Let me know if you have issues.
How it works is you select a folder for Karta to access, and then it opens it as a graph on a canvas. You can move and resize any folder or file, and you can connect any of them together. Connected nodes always appear together, so you can start making associations and then explore your files in a more freeform and intuitive way. There are also text nodes you can create to add notes and headings, which won't exist in your file system.
The app has gone through many iterations and took this long because I wasn't sure about what I was making, and had to learn most of the required skills as I went. Now after being in a bubble for too long I finally have something that works, so I'm looking for honest feedback about whether this project is useful or at least interesting. I oscillate between thinking this has a lot of potential and this is pointless, so some contact with the real world would be nice :)