I build and run multi-agent systems on my own hardware, then write down what breaks.

These are operational notes — failure modes, measurements, postmortems. Things I hit running a conductor/worker orchestration stack across several machines, written down because nobody else had.

Recent notes

• Nothing published yet.