In an effort to become even more agentic, I took all of @catehall's posts and created a Claude Project I can ping. My guess is that it will make Claude provide me with slightly more strategic advice on how to approach a situation. Any other resources I should add?
This is valuable so far I think cc @catehall
It’s referencing: https://open.substack.com/pub/...
@JacquesThibs @catehall This would be fun to benchmark. Does including all of Cate Hall's posts in context increase LLM time horizon? (Or improve benchmark scores at all?)
@chi_t_williams @catehall Hah, I did have this thought before when I was considering doing some kind of 'agency' benchmark
@JacquesThibs @catehall CFAR handbook? Sequences? TOCs of them? I haven't done this much. Opus 4 is good enough to highlight concepts of which some are useful in a situation, but Opus doesn't understand them deeply enough to walk me through e.g. goal factorization.
@JStoehler Overall I just want to nudge it out of internet and average common perspective slop so think there’s a way to at least make it slightly better
@JacquesThibs @catehall I'm also curious how you scraped the posts -- did you use Claude Code or something?
@chi_t_williams @catehall Nah, was faster to just copy-paste them


