I’m working on a similar project (DepsHub) where LLMs are used to make major library updates as smooth as possible. While it doesn’t work in 100% cases, it really helps to minimize all the noise while keeping your project up to date. I’m not surprised Slack decided to go this way as well.