Stdlib AI Agent
I've spent the past few months working on little projects mostly coded via LLMs. Used it to build my own little AI agent in python attempting to keep it stdlib-only, and a little blog engine with MFA. Things I would never have previously been able to do as a sysadmin/infrastructure but not developer. The struggle I'm finding with it is how to go from good enough for me to good enough for production. After being stuck in QA/review hell trying to use LLMs to assess the code, I've started trying to unpick it myself. I can't help but feel that the time spent learning the vibe-coding approach wouldn't have been better spent just learning software development properly. While that's been frustrating, most my fun has been on the non-tech side. To get a break from the computer last year, I ran the length of Japan (around 3,400 km). And now I've been helping my partner edit the videos and stick them on YouTube: https://www.youtube.com/watch?v=50oQQThXWnc&list=PLVFWiVG-hZ...