Revisiting My Investing Scripts
Stock screens, Greek property-listing screens, watchlists, and the recurring need to make judgment visible.
By April 2026, the investing work had grown into something broader than a few scripts. There were stock screens, valuation notes, persona files, cyclical chart scans, a runbook for a dedicated investing workspace, and even a real-estate opportunity finder for Greece: a browser-assisted workflow for collecting property listings and scoring them with the same discipline as stocks.
The common thread is obvious to me now: I keep trying to make judgment visible. Markets are noisy. Property listings are noisy. My own thoughts are noisy. Code helps when it turns a vague feeling into something I can inspect, rank, challenge, and rewrite.
That property tool is a good example. Browser-assisted capture, site adapters, extracted listings, price per square meter, distance to sea, tourism language, renovation upside, reports in JSON, CSV, and Markdown. Not glamorous. Very useful. A small machine for reducing blindness.
The equities work has the same nervous system. Fetch what can be fetched. Cache what should not be fetched repeatedly. Check official sources where possible. Keep yfinance useful without pretending it is sacred. Add valuation notes. Track cycles. Let the visual cortex do what it does well, then force the numbers to complain.
Longer term, this whole layer belongs inside Benji. In many ways it already does: the investing scripts, notes, screens, and opportunity tracking are exactly the kind of resource-and-decision work Benji should swallow and delegate. The backend direction is there. The UI still needs a lot of improvement before it feels as calm and powerful as the idea deserves.
I still like the idea of an AI council that disagrees by design: Buffett, Munger, Lynch, Pabrai, Damodaran, and a boring credit officer whose entire personality is ruining beautiful stories. That last one may be the most useful member of the band.
Looking back, I can see a pattern that goes beyond finance. When something matters, I eventually try to build a tool around it. Not because the tool replaces judgment. Because the tool gives judgment a place to stand.