swarm-advanced

Orchestrates multi-agent swarms for research, development, testing, and analysis workflows.

1|Updated Jun 3, 2026
One-click install
npx skills add https://github.com/KentwareDemo/RuView --skill swarm-advanced-kentwaredemo
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: swarm-advanced
Source: https://github.com/KentwareDemo/RuView/tree/main/.claude/skills/swarm-advanced
Command: npx skills add https://github.com/KentwareDemo/RuView --skill swarm-advanced-kentwaredemo

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires claude-flow.

What problem does it solve? Coordinating multiple AI agents across complex tasks like research, full-stack development, and testing requires manual orchestration, which is error-prone and hard to scale. This Skill provides structured patterns for initializing swarm topologies, spawning specialized agents, and running parallel or sequential workflows with Claude Flow. ## Core Features & Use Cases - Swarm Topology Patterns: Configure mesh, hierarchical, star, or ring topologies matched to research, development, testing, or pipeline workflows. - Parallel Task Orchestration: Execute independent tasks concurrently across specialized agents with monitoring, memory persistence, and state snapshots. - Fault Tolerance & Learning: Apply auto-recovery strategies, error analysis, and neural pattern training to improve coordination over time. - Use Case: Spin up a hierarchical development swarm with an architect, backend and frontend developers, testers, and a reviewer to design, implement, test, and deploy a REST API in coordinated phases. ## Quick Start Initialize a mesh swarm with Claude Flow and spawn researcher and analyst agents to investigate a topic in parallel and synthesize a report.

Frequently Asked Questions about swarm-advanced

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I set up a multi-agent swarm with Claude Flow?

Install claude-flow via npm, then call swarm_init with a topology such as mesh or hierarchical, spawn agents with agent_spawn specifying type and capabilities, and orchestrate work using task_orchestrate or parallel_execute. A CLI fallback like npx claude-flow swarm also works.

Which swarm topology should I use for research vs development?

Use mesh topology for research and analysis since agents communicate peer-to-peer, hierarchical for structured development with a coordinator, star for centralized testing, and ring for sequential pipeline processing.

Can Claude Flow swarms run tasks in parallel?

Yes, parallel_execute runs independent tasks concurrently across assigned agents, while batch_process handles multiple test suites or items. Dependent tasks should use sequential strategy with declared dependencies.

How do I persist swarm state across sessions?

Use memory_usage with namespaces and TTL values to store findings, memory_persist for cross-session persistence, and state_snapshot with context_restore to create and recover checkpoints during long workflows.

What should I do when swarm agents fail to coordinate?

Verify the chosen topology fits the workflow, check memory namespace configuration, and enable swarm_monitor for real-time status. Apply daa_fault_tolerance with auto-recovery and use error_analysis to diagnose failure patterns.