Swarm Orchestration

Orchestrates multi-agent swarms with mesh, hierarchical, and adaptive topologies for parallel task execution.

1|1|Updated Nov 28, 2025
One-click install
npx skills add https://github.com/33may/robotics --skill swarm-orchestration-33may
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: Swarm Orchestration
Source: https://github.com/33may/robotics/tree/main/humanoid/.claude/skills/swarm-orchestration
Command: npx skills add https://github.com/33may/robotics --skill swarm-orchestration-33may

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires agentic-flow.

What problem does it solve? Coordinating multiple AI agents on complex tasks requires manual management of task distribution, communication, and failure handling, which becomes unmanageable as agent counts grow. ## Core Features & Use Cases - Multiple Topologies: Configure mesh, hierarchical, or adaptive swarm structures to match task complexity and coordination needs. - Flexible Execution Modes: Run tasks in parallel, as sequential pipelines with dependencies, or let the swarm auto-orchestrate based on goals and constraints. - Shared Memory & Resilience: Share state across agents via swarm memory, with load balancing, retry policies, and fault-tolerant task reassignment. - Use Case: Building a REST API with a team of agents—spawn a coder, tester, and reviewer, orchestrate them in parallel, and share the API schema through swarm memory so every agent works from the same context. ## Quick Start Initialize a mesh swarm with five agents and orchestrate the task of building a REST API with tests in parallel mode.

Frequently Asked Questions about Swarm Orchestration

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

FAQPage Schema
How do I orchestrate multiple AI agents in parallel?

Initialize a swarm with npx agentic-flow hooks swarm-init, spawn agents with agent-spawn, then run task-orchestrate with --mode parallel. The swarm distributes tasks across agents and aggregates results automatically.

What swarm topology should I use for multi-agent workflows?

Use mesh for peer-to-peer collaboration among equal agents, hierarchical for centralized coordination with a queen agent directing workers, and adaptive when you want the swarm to switch topology based on task complexity.

What are the requirements for running agentic-flow swarms?

You need agentic-flow version 1.5.11 or higher and Node.js 18 or later. Familiarity with distributed systems concepts is helpful but not required.

How do agents share state in a swarm?

Agents share state through swarm memory using swarm.memory.store to write data and swarm.memory.retrieve to read it. This lets all agents access shared artifacts like API schemas without direct messaging.

Why are my swarm agents not coordinating properly?

Coordination failures usually stem from missing memory access or disabled hooks. Verify agents can read and write swarm memory, and enable pre-task and post-task hooks for automatic synchronization.

How does a swarm handle agent failures during task execution?

Configure resiliency with swarm.setResiliency to set retry policies with exponential backoff and fallback behavior. Failed tasks can be automatically reassigned to other agents after the configured retry attempts.