I typed six lines into Claude Code last week on Aschenbrenner’s latest 13F and had the comparison to Ackman, Einhorn, and Pabrai.

Two minutes and ten seconds later, four hedge fund 13F filings sat in my vault as clean markdown tables, plus a one-page comparative memo.

The most-watched 13F of the quarter had just dropped, Leopold Aschenbrenner’s Situational Awareness LP. Headlines were already framing it as a big bet against the chip complex. I wanted the raw data and a clear read on how it compared to what the value crowd was holding.

So I opened Claude Code and wrote:

It’s 13F season. Aschenbrenner’s Situational Awareness fund just filed yesterday and the market is talking about his chip shorts. Pull his latest 13F. Also pull Ackman, Einhorn, and Pabrai for context. Save each one in raw/13f/ as a clean markdown table. Then in output/, write me a one-page memo on what Aschenbrenner is doing and how it compares to what the value guys are holding. What’s he betting against? What’s he betting on? What changed since last quarter?

That was it. Six lines. No URLs. No source list. No instructions on format or tools.

Claude Code picked the sources. It used Firecrawl. It decided what a “clean markdown table” for a 13F should look like. It structured the memo without ever being shown an example. Two minutes and ten seconds later the files were there.

I didn’t touch a single formula or write a single line of scraping code.

I recorded the full session. The scary (and exciting) part is how little I had to specify. Just a clear end goal.

That’s when these systems stop feeling like fancy autocomplete and start feeling like actual leverage.

I’m still figuring out where the ceiling is on this. But every time I test it on something live, the gap between what I have to do and what actually gets done keeps widening.

What live filing or event are you thinking of running something like this on?