Wow, all my old LLM experiments are now invalid

Gosh, I have really experienced the “November moment” (i.e., the release of GPT-5.2 and Opus 4.5 in November 2025 hugely increased AI capabilities). Just a few months ago, I was getting mediocre results with LLM-assisted docs. Now, I have skills and subagents I’ve generated in Claude Code to assist me in writing docs…and it truly feels revolutionary. I’ll write more about it later, but I want to mark this moment for myself in my personal blog to remember later as a watershed.