openrig: Boot an AI Coding Agent Team at Once with YAML

Unify AI coding agents scattered across distributed terminal sessions into a single organization. Manage multiple harnesses like Claude Code and Codex within one system, define the agent team structure via YAML files, and launch the execution environment with a single command.

A lead agent receives the final objective, distributes tasks to specialized sub-agents, and coordinates their work. Each agent is assigned an independent seat to work from, while decisions requiring user input and completed deliverables are centrally delivered to the user. Track the team's progress and context in real time through a tmux-based dashboard.

This solves the problems of manually managing individual terminals and losing context between sessions, as in traditional approaches. It runs on Node.js 20+ and tmux environments, and is released under the Apache-2.0 license. It is suitable for developers seeking to simplify complex multi-agent workflows and build sustainable development environments.