Documentation

Introduction

AgentsMesh is an AI Agent Fleet Command Center that enables teams to orchestrate AI coding agents like Claude Code, Codex CLI, Gemini CLI, and Aider at scale. Manage remote AI workstations (AgentPod), coordinate multi-agent collaboration (AgentsMesh), and track progress with integrated task management — all from a single platform.

Supported AI Agents:

  • Claude Code (Anthropic)
  • Codex CLI (OpenAI)
  • Gemini CLI (Google)
  • Aider
  • OpenCode
  • Custom agents via MCP protocol

What You Can Do

AgentsMesh gives your team superpowers for AI-driven development.

Orchestrate Multiple AI Agents

Coordinate Claude Code, Codex CLI, Gemini CLI, Aider and more working together through channels and pod bindings.

Remote AI Workstations

Launch AgentPods — cloud-based development environments with real-time terminal access, Git worktree isolation, and full AI agent capabilities.

Task-Driven Development

Create tickets, bind them to AgentPods, and track progress on a Kanban board. From requirements to execution to delivery.

Self-Hosted & Secure

Deploy runners on your own infrastructure. Your code and data never leave your environment. BYOK model for full cost control.

How It Works

AgentsMesh separates the control plane from the data plane. Orchestration commands travel through gRPC with mutual TLS, while terminal I/O streams through a globally distributed Relay cluster — so you can operate agents on any machine, from any device, securely.

Client Layer

Access from anywhere — browser, phone, or tablet

🖥️ Web
📱 Mobile
📟 Tablet
HTTPS / WebSocket
Cloud
AgentsMesh Cloud

Orchestration, monitoring, collaboration & security

Orchestration
Monitoring
Collaboration
Security
Control Plane
Pod lifecycle, agent commands, status sync
🔒 gRPC + mTLS
Data Plane
Terminal I/O, real-time streaming
⚡ Relay Cluster
Execution Layer
Self-Hosted Runners

Deploy on your own infrastructure — your code never leaves your environment

🖥️ macOS
🐧 Linux
☁️ Cloud VM
PTY + Sandbox + Git Worktree
Agent Layer

AI coding agents running in isolated Pods

🤖 Claude Code
🤖 Codex CLI
🤖 Gemini CLI
🤖 Aider
🔐
Security Architecture — Dual-Path Design

Control signals (pod creation, agent commands, status updates) travel through gRPC secured with mutual TLS (mTLS) — each Runner holds a unique X.509 certificate issued by the platform PKI. Terminal data streams (keystrokes, output) flow through the Relay cluster, keeping latency-sensitive I/O on a separate, optimized path. This dual-path design ensures you can securely operate AI agents on any machine, from any device, anywhere in the world.

Quick Links