mvschwarz/openrig
- Source
- GitHub
- First trending
- Category
- Agents
- GitHub stars
- 1,723
- Main language
- TypeScript
This page introduces an external open-source repository. It is not an HDATF product.

What it does
A multi-agent harness that runs Claude Code and Codex as one system. The agent team is defined in YAML and started with one command, and a lead agent coordinates specialists and reports results and decisions needing attention.
How it helps ATF
When considering how Harness could connect external executors such as Claude Code and Codex, its YAML team definitions and a lead agent that brings back results and decisions are worth comparing.
License
Apache-2.0 Permissive, with a patent grant. Commercial use is allowed; keep the notices and state your changes.
More in this category
- ZJU-REAL/Easel
An open-source social media content workbench that connects the OpenClaw Agent, account profiles, content skills and media tools. It covers discovery, planning, creation, publishing and performance analysis, and publishes to seven platforms including Xiaohongshu, Douyin and Zhihu. - dataelement/dsh-desktop
A local-first desktop app for DeepSeek Harness. It manages agent presets and sessions, supports continuing sessions from a phone, and turns source material into editable PPTX decks. - NVIDIA/OpenShell
An AI agent runtime that applies policies to sandboxed file access, system calls, and network connections. Its README describes credential injection for approved endpoints and formal checks before policy changes. - modelstudioai/cli
The official CLI for Aliyun Model Studio exposes AI capabilities as structured tool calls. It includes multimodal generation, asset understanding, memory and workflow calls. - allenpeng0705/EnvoyMesh
A decentralized agent mesh using libp2p, cryptographic identities, and signed messages for chat, knowledge, and tasks. A desktop home node serves phone thin clients while model keys and repositories remain on the computer.
Only repositories in the ranked Trendshift lists are included, and the lists are used only to find candidates. We do not copy their ranks. Descriptions, licenses and star counts come from each GitHub repository. The notes are our own reading. We have not tested these projects, and a place on a trending list does not prove quality.