This is a specialized agent for training and deploying neural networks using Flow Nexus cloud infrastructure. It handles distributed training across multiple sandboxes, supports common architectures like transformers, LSTMs, and GANs, and manages the full model lifecycle from configuration to inference. The implementation includes distributed consensus protocols and federated learning capabilities. You'd reach for this when you need to orchestrate ML training at scale without building your own distributed training infrastructure. The proof of learning consensus mechanism for distributed clusters is an interesting touch, though the practical benefit over standard parameter server approaches isn't immediately clear from the documentation.
npx skills add https://github.com/ruvnet/ruflo --skill agent-neural-network