Hello World, from the Human Side
So I bought a domain called dangerouslyskippermissions.blog. Yes, really. Here is why.
This blog will be a follow-along blog with me using Claude as my personal assistant. The idea is to share and show how different working today is thanks to AI, but also show the technical side and struggles - maybe others could learn from it, or benefit from it. And – if you ended up here, and you are not techy, it may be entertaining for you. 😄
So, the background is, that about 2 weeks ago the Hungarian internet also exploded thanks to OpenClaw. I love creating, so I installed it in a docker... And then uninstalled it really quickly, as I did not really like it all too much, and did not understand the fuss about it.
In parallel at that time I was still using Antigravity on 4 different Google hosted accounts (generous limits... :D) for a couple of fun personal projects.
Sadly, one big client project all-of-a-sudden ate upp all 4 accounts, and I was at a point where I needed LLM to keep up with the iteration speed. Since whenever I have the opportunity, I plan with Opus (and then execute with Sonnett), I decided to go with a Claude Max (the smaller, so Claude mid-Max I guess?!) subscription and Claude CLI (idea being, that Claude Cowork must be aaaaaawesooome. Not).
Now the thing is, that after working with CLI for a couple of days, I realised that there is this --dangerously-skip-permissions argument AND having it paired up with it's chrome plugin, I basically have an OpenClaw with better control over the installed mcp-s and plugins.
And boy, this changed everything... This blog will be about this change, both from my perspective as well as Claude's perspective.
