See Skill releases here (zipped, ready for import): github.com/oaustegard/c...

Unzipped code in the repo

Note that these skills are by no means one-shot endeavors: it represents 111 commits to the repo, 25000 lines added, 6700 removed -- github.com/oaustegard/c...

Let me know if it’s helpful— or how it could be improved. I went for broke and shot for all three environments; CCotw required a fair bit of feedback to refactor appropriately; a bit like playing LoC accordion

Ready for real world testing and ahem _iterating_

github.com/oaustegard/c...

I’m currently vibing on a similar concept, work (very much) in progress here: github.com/oaustegard/c...

Exhibit B (non-AI): bsky.app/profile/aust...

Hmm, based solely on feel, like the others above, given their “grounding with Google” I’d put it next to Claude.

right, I was using it for a paper summary and it said something about a 4 week timeline and a $60 budget. Which it had derived from a prior paper and now twisted into MY requirement. That was the straw that made me turn it off.

@gracekind.net scores

Oh I meant just to train and release a competitive base model then later add thinking

Isn’t that pretty much what DeepSeek did a year ago? Then later when they dropped the thinking model the stock market did a mini crash?

Presumably this is the paper? www.sciencedirect.com/science/arti...

It comes to the **exact opposite conclusion** to the Economist subheading:

“we advise recruiters to be cautious about paying too much attention to these images when inferring candidates' personality traits.”