hive-mind-advanced

Multi-agent coordination uses a queen-led hierarchical system where a central coordinator directs specialized agents (architect, coder, tester, reviewer) toward shared objectives. This Skill implements role-based orchestration with collective memory and consensus mechanisms to ensure agents align on decisions and share context across the workflow.

Byzantine fault tolerance is a consensus method that produces reliable decisions even when some agents provide incorrect or corrupted information. This Skill applies it to guarantee robust outcomes in multi-agent coordination, protecting against faulty agent behavior without losing overall system integrity.

Persistent agent memory uses LRU caching and SQLite with WAL (write-ahead logging) to maintain a shared knowledge base accessible to all agents in the swarm. This Skill manages collective memory automatically, enabling continuous learning and context preservation across sessions.

Yes. This Skill is built for orchestrating swarms of specialized agents—architects, coders, testers, reviewers—through design, implementation, and validation phases. Each agent holds a distinct role; the system distributes tasks, maintains shared memory, and achieves consensus on critical decisions.

This Skill supports majority-vote, weighted, and Byzantine consensus modes. Choose majority for simple agreement, weighted to prioritize certain agent opinions, or Byzantine to tolerate faulty agents while reaching reliable consensus across complex workflows.

Automatic task distribution assigns work to agents based on role and capacity; performance monitoring tracks agent execution, latency, and outcomes in real time. This Skill uses these metrics to optimize task flow, identify bottlenecks, and ensure efficient collective work across the swarm.