Ant Group's inclusionAI open-sources Choruz, a local-first human-AI agent collaboration app
Ant Group's inclusionAI released Choruz, an Apache-2.0 licensed local-first collaboration app where humans and AI agents work together in a Slack-like space. It supports Claude Code, Codex, Grok, and webhook-driven external agents by default. The repository was published on GitHub on 2026-09-02.
China context
- Original name
- 蚂蚁集团
- Outside China
- Open weights · github.com
- Claims
- Company-reported; not yet independently evaluated
- For builders
- Builders outside China can fork and self-host Choruz to orchestrate multiple CLI-based AI agents locally, with Apache-2.0 licensing and support for Claude Code, Codex, Grok, and webhook agents.
- For investors
- Investors can monitor adoption of Choruz as an indicator of Ant Group's strategy in developer tools and open-source agent orchestration, though the project is pre-release and not yet a commercial product.
Choruz is a local-first collaboration app where humans and AI agents work together in a Slack-like space. Each agent runs a real CLI in its own workspace and can hand work to people or other agents through messages, threads, tasks, and files. Choruz supports Claude Code, Codex, Grok, and webhook-driven external agents by default. Plugins enable Pi, OpenCode, and MathCode. The project is under pre-release development, with interfaces, configuration, and data formats that may change incompatibly. Requirements include Rust, Node.js 24, pnpm 10, PostgreSQL 16, and at least one supported agent CLI. Source code and software documentation are distributed under the Apache License 2.0.
Choruz runs each agent as a real CLI in its own workspace, preserving each CLI's models, tools, and session capabilities. It provides isolated Companies, direct messages, groups, mentions, threads, and channel task boards. Agent workspaces can be dedicated directories or Git worktrees with skills, sub-agents, AI Manager, and scheduled work. The app includes an integrated terminal, file browser and editor, SSH runtime hosts, and browser-based remote control. Integration is via REST APIs, WebSocket sync, webhook agents, Slack and Telegram bridges, and optional plugins.
Developers outside China gain a new open-source tool for orchestrating multiple AI coding agents in a local-first environment, reducing reliance on cloud-based agent platforms and giving them direct control over agent workspaces and data.
Choruz could lower the cost and complexity of multi-agent orchestration for teams that prefer local execution and open-source licensing, but its pre-release status means production adoption risk remains high.
The project is in developer preview, so interfaces and data formats may change. A specific signal to check is whether the project reaches a stable release and documents migration paths for existing installations.