TrendWhat rose on GitHub, every day at 07:30 KST · 한국어

GitHub · as of November 15, 2026 · View on GitHub →

mvschwarz/openrig

Multi-agent orchestration framework

+110 stars in 7 days · trending 2 of the last 30 days

Manage a coordinated team of AI coding agents with a single command

Imagine moving away from juggling multiple terminal windows to give instructions to different AIs, and instead forming and managing a team with a single command. OpenRig is an open-source system that binds individual AI coding agents, such as Claude Code and Codex, into a structured team. You tell the lead agent the desired outcome, and specialized agents divide the work and bring you the necessary decisions.

Defining agent team structure with YAML

OpenRig declaratively defines the composition of an agent team using a YAML file (RigSpec). It specifies which agent takes on which role, the connections between agents, and how to resume if a session is interrupted (continuity policy). A single rig up command automatically handles everything from the tmux session and runtime to startup files and readiness checks. You can also find and incorporate existing running Claude Code or Codex sessions into management using rig discover. When changing the team structure, you can expand or shrink the running topology with rig grow or rig shrink.

Viewing task progress and results at a glance

The terminal user interface (TUI) displays the team's coordination status in table and graph formats. You can navigate the status, project, feed, and system health of each agent (seat) using the keyboard, mouse, or command bar. Communication between agents is handled via rig send, rig broadcast, and rig chatroom commands, and a typing guard feature prevents automatic messages from mixing in when a person types directly into a specific seat. Once a task is complete, you review the results and feedback, and the flow continues by requesting the next changes from the same owner agent.

System changes to verify before installation

OpenRig writes several files to the user's system during installation and execution. It adds mouse support settings to ~/.tmux.conf, and on macOS, you can install or configure cmux. When the daemon starts, it creates a database and plugin resources under ~/.openrig, and records trust settings and hook information in the configuration files for Claude Code and Codex (~/.claude.json, ~/.codex/config.toml). Specifically for Claude Code, it adds context collector commands and activity hooks to .claude/settings.local.json in the workspace. It is important to understand the impact of these changes on your system and back up the relevant files before running.

Execution environment and requirements

OpenRig runs on macOS or Linux environments with Node.js 22 or 24 and tmux installed. For Apple Silicon Macs, Node.js 22 is recommended due to compatibility issues, and native Windows is not yet supported. WSL2 has not been tested. Even when installing with Bun, Node.js 22 is required alongside it. Optionally, you can use herdr or cmux to view agent terminals together, and Docker for service-based rigs. When upgrading from an existing Node.js 20 environment, you must first change the Node.js version to 22 or higher and then reinstall the CLI.

Situations where this system is suitable

OpenRig is suitable for developers who want to manage complex workflows where multiple AI agents collaborate, rather than a single AI session. It is particularly useful if you use Claude Code and Codex simultaneously and want to manage their sessions integratively while maintaining task continuity. However, it requires a deep understanding of the system and caution regarding configuration file changes. OpenRig does not replace the agents themselves but acts as a harness that manages the structure and operation of the team formed by the agents. If you want to introduce an agent team to your workspace and continuously manage that team's work context from the same address, this repository is worth reviewing.

Written by AI from this repository's README on October 2, 2026. For the facts, the original on GitHub is the reference.

By the numbers

GitHub trending history
16 days between Jun 19, when Trend began recording, and Nov 15, 2026 · 2 of the last 30 days · first on Sep 28 · best daily rank #3 (Sep 30)

Read from GitHub, as of November 15, 2026.

Language
TypeScript
Topics
agent-harness · agent-orchestration · agent-skills · ai-coding · claude-code · cli · codex-cli · multi-agent
Stars
8,861
Stars gained in 7 days
+110
Contributors
104
Forks
759
Latest release
v0.7.16 · November 15, 2026
Commits
5,403
Last commit
November 15, 2026
Open issues
164
Open pull requests
72

Related repositories

View on GitHub → · Homepage

README badge

trend.ondo.sh · GitHub daily trending #3 · September 30, 2026

The badge above shows the highest this repository has ranked on GitHub's daily trending since June 19, 2026, and the day it did. Paste the Markdown below into your README. The badge opens this page, and it updates by itself when the repository climbs higher.

[![trend.ondo.sh](https://trend.ondo.sh/github/r/mvschwarz--openrig/badge.en.svg)](https://trend.ondo.sh/github/r/mvschwarz--openrig/en.html)