nftechie/stonkfly

A fly-connectome simulation that can operate a crypto trading account. Actual neural output, actual Coinbase integration. Profitable learning has not been demonstrated. How it works: Public Coinbase prices become an RGB chart. It stimulates 3,335 brightness inputs and 811 R8 color inputs in the retained MaleCNS v1.0 graph: 166,700 neurons, 25.6 million connections. A fixed neural readout proposes buy, sell or hold. A custom Coinbase AgentKit ActionProvider checks limits and places spot orders through Coinbase Advanced. Positive portfolio P&L stimulates 15 identified PAM11 dopamine cells; negative P&L stimulates two PPL101 aversive dopamine cells. A candidate memory rule changes existing KC-to-MBON connections. These are engineered reinforcement signals, not modeled pain receptors. Synaptic changes do not establish that it learns to trade profitably. Model and evidence. Python 3.11, a C++17 compiler, macOS/Linux. Allow several GB for the dataset and dependencies; 16 GB RAM recommended. Default: paper trades, real public BTC-USDC data, $100…