Show HN: SleepyQuant – a 12-agent crypto quant running on one Mac

SleepyQuant is a local-first crypto quant experiment running 12 specialized AI agents on a single Apple M1 Max Mac.

It uses MLX models, local storage, and a routing layer to handle spot and futures paper trading without cloud dependencies.

The system adds risk controls, a failure vault, and weekly post-mortems/newsletters to track what breaks and why.

The author is asking whether the multi-agent setup is worth the overhead, and how to manage MLX memory efficiently on Apple Silicon.