swarm-advanced

Distributed orchestration coordinates multiple AI agents across mesh, hierarchical, star, and ring topologies using swarm patterns. This Skill enables you to initialize agents with specialized roles, spawn them dynamically, and execute tasks in parallel or sequence across Claude Flow, scaling workflows from research and development to testing environments without manual coordination.

Swarm topologies define how agents communicate and coordinate. Mesh topologies suit research with peer-to-peer collaboration; hierarchical topologies structure development with layered authority; star topologies centralize testing through a coordinator. Each topology optimizes for different workflow patterns—research uses parallel exploration, development uses staged building, testing uses coordinated validation.

End-to-end workflow automation with multi-agent systems chains information gathering, analysis, implementation, and validation into a single coordinated process. This Skill supports defining complete automated workflows where specialized agents work in parallel on research, development, or testing tasks, then merge results into comprehensive outputs like reports or validated code.

Fault tolerance in distributed swarms enables automatic recovery when individual agents fail. This Skill implements neural pattern learning to recognize optimization patterns from successful executions and automatically recover from failures, ensuring workflow continuity across research, development, and testing environments without manual intervention.

Claude Flow is a required dependency for building distributed swarms with this Skill. It provides the MCP-based framework that enables multi-agent initialization, parallel task orchestration, memory storage, knowledge graph construction, and progress monitoring across specialized swarm architectures.

Large-scale research automation uses mesh topology with specialized agents—web researchers, academic analysts, data processors, report writers—working in parallel. This Skill provides patterns to spawn agents dynamically, orchestrate parallel information gathering and synthesis, store findings in memory and knowledge graphs, and output comprehensive reports automatically.