projects/ has the stuff that works and that people use. This is the rest: explorations, simulations, questions I keep dragging around.

A friend of mine has a bingo card for whenever I tell him "I have an idea": graph, hexagons, emergent, tests. He wins every time.

That said, laying these projects side by side, the common thread is simpler than that. What keeps coming back is emergence: lots of agents, simple rules, and watching what happens. The Commons Arena↗ as the game-theory version, hexmap as the weather version, lightfall as the game version.

Some of these projects touch domains where I have zero background, hexmap's geophysics typically. The code is written by Claude. If I can't write the code, how do I know it isn't wrong?

My rule: being able to tell when it's wrong. Tested invariants (if a water total drifts in hexmap, a test fails), measurements instead of impressions, and a logbook that also records the mistakes and the pivots.