Source: www.producthunt.com/products/mao-multi-agentic-orchestration-tool
MAO lets developers add Claude Code, Codex, Gemini, Grok or OpenCode as live terminal agents on a shared canvas. Connecting two agents with a wire lets them ask each other questions, delegate tasks and return results. A lead agent can split one instruction across the others, and the user steps in only when an agent needs help.
Each agent can be given a role, such as the built-in Builder and Reviewer or a custom role written in Markdown. Agents can run on the user's own API keys, existing logins or a custom endpoint, and several keys can be pooled and switched from within an agent. MAO also ships skills that teach agents to message, delegate and report back on the canvas.
According to its Product Hunt listing, MAO is written in Rust on GPUI, the framework behind Zed, and runs natively on macOS, Linux and Windows. The app has a 7-day free trial that needs no card, after which it costs $20 paid once with no subscription.
Why it matters if you build agents
Developers who run several coding agents in separate terminals can coordinate them on one canvas instead of copying context by hand between windows. Because keys and workspaces stay on the local machine, the setup suits teams with strict data handling, though the agents themselves still call their providers' APIs.
Get stories like this every weekday
One short email with the day's agent tooling news.