# Coordinate Multi-Agent Swarms

Coordinating many agents is hard when roles, topology, and task flow are unclear. This skill provides swarm patterns, hook workflows, and coordination practices for agentic-flow.

## Install

```bash
npx skillstore add ruvnet/swarm-orchestration
```

## Metadata

- Status: approved
- Slug: ruvnet-swarm-orchestration
- Skillstore revision: r1
- Version status: missing
- Tree hash: 2fafc00e741a088759dfbd455ecb32aa477e4e08b43db0a9e651631072a43bb2
- Author: ruvnet
- GitHub username: ruvnet
- License: MIT
- Repository: https://github.com/ruvnet/claude-flow/tree/main/.claude/skills/swarm-orchestration
- Ref: 3e4b6c31a74a3bd1a291c98cf585d720cb9fbc88
- Supported tools: Claude, Codex, Claude Code
- Audit status: complete
- Agent install advisory: allowed
- Manual install advisory: allowed
- Artifact signature: available
- Audit attestation: unavailable
- Human verification: not\_verified
- Risk factors: external\_commands
- Quality score: 78
- Quality tier: bronze
- Public page: https://skillstore.pages.dev/skills/ruvnet-swarm-orchestration
- Manifest: https://skillstore.pages.dev/api/skills/ruvnet-swarm-orchestration/manifest

## Capabilities

- Shows mesh, hierarchical, and adaptive topology patterns for agentic-flow swarms.
- Documents agent spawn and task orchestration commands using agentic-flow hooks.
- Provides examples for parallel, pipeline, and adaptive task execution.
- Explains shared swarm memory for storing and retrieving coordination state.
- Covers load balancing, fault tolerance, metrics, and hook-based synchronization.

## Use Cases

- Plan Parallel Delivery Work: Break a complex build into coder, tester, reviewer, and architect agent roles with clear task flow.
- Prototype Agentic Workflows: Compare mesh, hierarchical, and adaptive swarms before choosing a coordination model for a project.
- Standardize Hook-Based Coordination: Use pre-task, post-task, and session restore hooks to keep agent work synchronized.

## Prompt Templates

### Initialize a Small Swarm

```
Use swarm-orchestration to plan a three-agent mesh swarm for [project]. Include roles, topology, and setup commands.
```

### Design a Parallel Workflow

```
Create a parallel swarm workflow for [goal]. Assign agents, define deliverables, and include timeout guidance.
```

### Add Memory and Hooks

```
Extend this swarm plan with shared memory keys and pre-task, post-task, and session restore hooks.
```

### Optimize a Resilient Swarm

```
Design an adaptive swarm for [complex system]. Include load balancing, failure recovery, metrics, and approval points for commands.
```

## Limitations

- Requires agentic-flow and Node.js to run the documented commands.
- Examples are guidance snippets, not a complete API reference.
- External npx commands should be reviewed before execution.
- It does not validate topology choices or performance results automatically.

## Best Practices

- Start with two or three agents before scaling the swarm.
- Share key context through swarm memory instead of repeating instructions.
- Review external hook commands before allowing them to run.

## Anti Patterns

- Spawning many agents before defining clear responsibilities.
- Using adaptive orchestration without timeouts or quality constraints.
- Running hook commands without understanding their workspace effects.

## Security Audit

- Audited at: 2026-07-09T08:09:21.698\+00:00
- Summary: Most static alerts are false positives from Markdown code fences. Two alerts are confirmed because the skill instructs users or agents to run unpinned npx agentic-flow commands, which execute a third-party CLI locally. No prompt injection, credential exfiltration, network beacon, or system reconnaissance intent was found.

## Stats

- Views: 235
- Downloads: 9
- Favorites: 0
- Popularity score: 0