I keep noticing my investment theses can't actually be stress-tested and so built a tool to push back and sharpen my ideas. Curious to know if you actually write theses and test them or is it just me?

This might be obvious to people here but it took me a while to see it. Most of my theses were unfalsifiable. I'd write stuff like "great moat" or "disciplined management" but felt that a specific thesis was difficult to really test beyond what I was hearing in the news or observing the occasional articles/youtube videos.

So I started forcing my theses through three questions before I trust them:

First, can I even check this?
'Berkshire is resilient' is useless. 'Berkshire won't be a forced seller because their cash + treasuries stay above the repurchase floor they committed to' is checkable. I know what to look at and where.

Second, what would prove me wrong?
Like actually try to kill your own thesis. For a cash flow call I'll say "if operating cash flow goes negative AND they're issuing a ton of debt to fund capex, that's distress and I'm wrong." Now I've written down what would change my mind before I'm emotionally attached.

Third, where do my own claims contradict each other?
If I think earnings are falling but also the business is fine, ok, which one wins when they fight? That's usually where I learn something.

Anyway, I ended up building a tool around this because I do it constantly and wanted something to push back on me and not to wait for market signals be the indicator for what I felt. I have tried to avoid making it a generic AI filing analysis tool (there are too many of them lol).

The approach I am trying to take is to not try and make a platform that gives you AI answers for your thesis but instead helps sharpen your thesis through a learning loop where it checks your reasoning, pushes back on it and finally helps you sharpen your assumptions: similar to the BRK-B thing I wrote above. My bet is that the analysis part is going to be free and commoditized soon, and the only edge left is your own judgment and how fast you can pressure test it.

It's in private beta and free, not selling anything. If you actually write theses and this resonates, I'd love a handful of sharp people to play with it for feedback. it's on my profile / you can find it through the site. Totally volunteer based, thanks for your kind time!

Let me know if I am looking at this the wrong way: do you write down what would change your mind ahead of time or do you just figure it out when a position goes against you? Curious how people here actually pressure test their own thinking.