Replying to @tosanjayr @aramh
If you meant what tools am I using, is TLA+ a “proof assistant”. This video is a good introduction youtu.be/l3UdQy98sQ8
you write the Spec and run a cli to check it exhaustively using a massive exploration space.
I’m not sure I got your question right tho?
Week 2 trying to replace Agents "markdown specs":
Model TLA+ -> run proof -> graph in .dot -> generate tests -> implement code.
This design can run and get verified, and it is much more readable than any wall of text.
Replying to @codingedgar @JurajMalenica
the problems are:
- i cannot click
- i cannot interact
- i cannot reply to any box or add to chat
- it holds no links
- limit graph support by mermaid render, etc
pretty bad experience in @ChatGPT , graphs should be first class citizen
Replying to @codingedgar @JurajMalenica
small examples, not my favorite by any means, but i'll keep taking screenshots to destill something and have more data for this idea.
I' would look impossible, but this forecast takes into account a max time to work per day to catch up, and distributes the load across the month.
It also takes into account all the other projects I have.
The best version of my Cumulative Line Graph so far.
It packs a lot of info!
This gives me a clear view of what I have to do to catch up with my project and reach my goals.









