OpenRig introduces the concept of a 'rig' — a harness for harnesses — that orchestrates multiple real AI coding agents like Claude Code and Codex as a coordinated team. Instead of juggling separate tools, users define a team in YAML with seats bound to specific harnesses, a lead agent for coordination, and specialist agents for implementation and review. Notably, OpenRig uses tmux as its runtime rather than containers, because coding agents are stateful and need persistent terminal sessions. Its starter topology pairs an owner (implementer) with a checker (reviewer), enforcing separation between writing and reviewing code. Setup is transparent with dry-run plans, though it modifies tmux, Claude, and Codex configs. The project is early-stage, targets experienced users, and doubles token costs — but points toward cross-vendor agent orchestration as the next layer of the AI coding stack.
One Rig to Wrap Two Harnesses: OpenRig Puts Claude Code and Codex on the Same Team
Summary
OpenRig introduces the concept of a 'rig' — a harness for harnesses — that orchestrates multiple real AI coding agents like Claude Code and Codex as a coordinated team. Instead of juggling separate tools, users define a team in YAML with seats bound to specific harnesses, a lead agent for coordination, and specialist agents for implementation and review. Notably, OpenRig uses tmux as its runtime rather than containers, because coding agents are stateful and need persistent terminal sessions. Its starter topology pairs an owner (implementer) with a checker (reviewer), enforcing separation between writing and reviewing code. Setup is transparent with dry-run plans, though it modifies tmux, Claude, and Codex configs. The project is early-stage, targets experienced users, and doubles token costs — but points toward cross-vendor agent orchestration as the next layer of the AI coding stack.
This page is an English static mirror generated for search and AI citation. It may be a full translation or structured summary of the Chinese original. Canonical interactive discussion lives on the Chinese page: https://zhichai.net/topic/178635286