ai-maestro
AI Agent Orchestrator with Skills System - Give AI Agents superpowers: memory search, code graph queries, agent-to-agent messaging. Manage Claude Code, Codex, G
What is ai-maestro?
AI Agent Orchestrator with Skills System - Give AI Agents superpowers: memory search, code graph queries, agent-to-agent messaging. Manage Claude Code, Codex, Grok Build or any AI Agent from one dashboard. Move Agents between computers and locations
Overview of ai-maestro
What it is
AI Maestro is an open-source AI agent orchestration tool with a skills system. It provides a dashboard to manage multiple AI agents, such as Claude Code, Codex, and Grok Build, across multiple machines. It includes persistent memory, agent-to-agent messaging, and a peer mesh network without a central server.
Key features
- One dashboard to see and manage all AI agents, with guided creation, smart naming, and four deployment modes (tmux, Docker, AWS EC2, AWS ECS Fargate).
- Multi-machine support via a peer mesh network, allowing agents to run on different computers (Mac, Linux, cloud) and be managed from one place.
- Agent Messaging Protocol (AMP) for email-like communication between agents, with priority levels, message types, cryptographic signatures, and push notifications.
- Gateways to connect agents to Slack, Discord, Email, and WhatsApp, with smart routing and content security against prompt injection.
- Persistent memory for agents: long-term memory refreshed nightly, including an entity graph of relationships, so agents retain knowledge across sessions.
- Works with existing agents: Claude Code, Codex, Grok Build, Pi, OpenClaw, GitHub Copilot CLI, and any terminal-based agent.
Use cases
Running a multi-agent team for software development, where each agent owns a repository or process and communicates directly with others. Managing agents across different machines for specialized builds (iOS on Mac, Docker on Linux, heavy compute in cloud). Coordinating multiple research, writing, or DevOps agents without manual copy-paste between terminals.
Getting started
bashcurl -fsSL https://raw.githubusercontent.com/23blocks-OS/ai-maestro/main/scripts/remote-install.sh | bashRequirements: Node.js 18+, tmux. For Windows, install WSL2 first. Dashboard opens at http://localhost:23000.
License and pricing
MIT licensed. The license is open source with no stated pricing; no account, telemetry, or hosted tier are required.
Based on the project's public README and repository metadata.
