loaditout.ai
SkillsPacksTrendingLeaderboardAPI DocsBlogSubmitRequestsCompareAgentsXPrivacyDisclaimer
{}loaditout.ai
Skills & MCPPacksBlog

cc-swarm

MCP Tool

chrismole310/cc-swarm

Multi-machine Claude Code orchestration. Connect 3-10 CCs into a coordinated production swarm.

Install

$ npx loaditout add chrismole310/cc-swarm

Platform-specific configuration:

.claude/settings.json
{
  "mcpServers": {
    "cc-swarm": {
      "command": "npx",
      "args": [
        "-y",
        "cc-swarm"
      ]
    }
  }
}

Add the config above to .claude/settings.json under the mcpServers key.

About

CC Swarm

Connect multiple Claude Code instances into a coordinated production swarm.

Turn 3-10 machines running Claude Code into a unified AI workforce. Shared state, task queues, real-time messaging, file coordination — all through native CC tools.

---

The Problem

Claude Code is powerful on one machine. But production work happens across many:

  • A render farm for video/audio generation
  • A server for backend services and deployments
  • A posting machine for social media automation
  • A dev box for coding and testing

Each CC works in isolation. No shared context. No task coordination. No way to say "CC1, render this — CC3, post it when done."

The Solution

CC Swarm adds 13 native tools to every Claude Code instance:

You: "Check the swarm"
CC:  ⚡ 3 nodes online — CC1 (engine), CC2 (render), CC3 (posting)
     📋 2 tasks pending, 1 in progress
     💬 5 unread messages

You: "Assign the video render to CC2 and tell CC3 to post it when done"
CC:  ✅ Task assigned to CC2 — Mac Studio
     💬 Message sent to CC3 — "Post video when CC2 completes render"
Quick Start
1. Start the Hub (any machine)
npx @atlas/cc-swarm-hub
# ⚡ CC Swarm Hub running on port 7777
# Dashboard: http://localhost:7777
2. Connect Each CC

Add to ~/.claude/settings.json on each machine:

{
  "mcpServers": {
    "cc-swarm": {
      "command": "npx",
      "args": ["@atlas/cc-swarm-mcp"],
      "env": {
        "SWARM_HUB": "http://192.168.1.100:7777",
        "SWARM_NODE_ID": "my-machine",
        "SWARM_ROLE": "render-farm"
      }
    }
  }
}
3. Done

Every CC now has swarm tools. Ask it to check status, assign tasks, message other nodes, share files, or coordinate workflows.

Tools

| Tool | Description | |------|-------------| | swarm_status | Dashboard view — nodes, tasks, messages | | swarm_nodes | List all connected CCs with roles | | swarm_assign_task | Create and assign work to a specific CC | | swarm_claim_task

Tags

ai-agentsai-workflowanthropicautomationclaudeclaude-codedevtoolsmcpmcp-servermulti-machineorchestrationswarm

Reviews

Loading reviews...

Quality Signals

0
Installs
Last updated19 days ago
Security: AREADME

Safety

Risk Levelmedium
Data Access
read
Network Accessnone

Details

Sourcegithub-crawl
Last commit3/28/2026
View on GitHub→

Embed Badge

[![Loaditout](https://loaditout.ai/api/badge/chrismole310/cc-swarm)](https://loaditout.ai/skills/chrismole310/cc-swarm)