This agent reconfigures your multi-agent system's network topology on the fly to optimize communication patterns. It analyzes current performance, generates candidate topologies (hierarchical, mesh, ring, star, hybrid), and uses multi-objective optimization to select the best structure for your workload. You'd reach for this when agent communication becomes a bottleneck or when your swarm's workload profile changes significantly. The genetic algorithm approach to agent placement is solid, and the migration planning helps you avoid naive topology switches that cause downtime. Fair warning though, the benefits really depend on having actual communication hotspots to optimize. If your agents barely talk to each other, the complexity overhead isn't worth it.
npx skills add https://github.com/ruvnet/ruflo --skill agent-topology-optimizer