Jean

Jean is an open-source AI agent desktop harness and orchestrator developed by coollabsio, the team behind Coolify. It provides a visual and terminal-aware workspace built on Tauri, Rust, and Bun that bridges GUI canvases with underlying AI CLI tools and backends. In daily development, Jean orchestrates multiple AI backends—including CLI-based engines like Antigravity (`agy`), CommandCode, and model-specific providers. It wraps these backends into cohesive task pipelines supporting dedicated execution modes such as plan, build, and yolo (autonomous execution). Interactions are coordinated through chat sessions, visual worktree canvases, and structured output parsing over JSON streams. What distinguishes Jean is its native integration with developer workflows and worktrees. Instead of functioning merely as a terminal chat shell or a generic wrapper, Jean provides visual worktree management, canvas-based smoke testing, MCP (Model Context Protocol) configuration discovery, and automated schema enforcement via `--json-schema` prompts for consistent agent tooling. Jean is an active open-source project hosted on GitHub, distributed under an open-source license, allowing developers to self-host and run it locally with their own API keys and local CLI tools without subscription lock-in.

Tags: agent harness, ai agent, cli, coding agent, developer tools, OpenSource

Visit Jean

Key features

  • Multi-Engine CLI Orchestration — Connects to and drives agent backends like Antigravity CLI (agy) and CommandCode with streamlined authentication and model listing. - Worktree Canvas & Smoke Testing — Manage Git worktrees visually with the ability to execute smoke tests directly from the worktree canvas. - Execution Modes — Supports distinct operation modes including plan, build, and yolo for varying levels of autonomy and verification. - MCP Discovery — Discovers Model Context Protocol configurations automatically to integrate external tools into agent workflows. - Structured Output Streaming — Uses --json-schema enforcement and streaming JSON parsing for reliable magic prompts and agent dispatch. - Cross-Platform Tauri Core — Built on Rust and Tauri with a Bun-powered UI for low memory footprint and high responsiveness.

Use cases

Use Case: Orchestrating Autonomous Coding in Isolated Worktrees A developer needs an agent to refactor a subsystem without dirtying the current branch. Jean manages a dedicated Git worktree on its visual canvas, launches the agent backend in plan or yolo mode, and lets the developer inspect diffs and execute canvas smoke tests before merging.

Use Case: Managing Multi-Agent CLI Backends A team uses different CLI-based AI tools for planning versus implementation. Jean acts as the central harness, handling onboarding, authentication, and MCP tool discovery across backends like Antigravity CLI and custom tools from a single interface.

Use Case: Structured Agent Task Execution with Schema Enforcement A developer wants predictable, schema-validated modifications across a codebase. Jean drives CLI backends using magic prompts formatted against explicit JSON schemas, ensuring that agent tool calls and diff outputs conform to exact expectations.

Who it is for

Jean is designed for software engineers and teams who want a visual, worktree-aware harness to orchestrate and test CLI AI coding agents across multiple backends. It is less suited for developers seeking an in-editor VS Code extension or a pure zero-config single-model chat terminal.

Install guide

Installation Clone the repository and build the desktop application using Bun and Tauri: bash git clone https://github.com/coollabsio/jean.git cd jean bun install bun run tauri dev Alternatively, download the precompiled binary releases from the GitHub Releases page.

First Run 1. Launch Jean and complete the onboarding flow. 2. Ensure your chosen AI CLI backend (such as Antigravity CLI agy) is installed and authenticated on your system. 3. Configure your preferences and MCP servers in the Jean settings toolbar.

Example Session Open a project repository inside Jean and initialize a new worktree task: bash # From the Jean UI canvas, select your branch and switch to Plan mode # Submit your task prompt: "Analyze the user auth module and draft a plan to migrate token verification to RSA256." # Switch to Build or Yolo mode to execute code changes and run smoke tests directly from the canvas.

Part of CLI Coding Agents

Categories: AI Coding Agents, Code Assistants, Developer Tools

Related CLI agents

  • SWE-agent — SWE-agent is an open-source autonomous software engineering agent harness built and maintained by researchers from Princeton University and Stanford University.…
  • Agent Runbook — Agent Runbook is an open-source contract-based multi-agent skill framework developed by KnoxOps. It acts as an orchestrator and compiler that…
  • Weaver — Weaver is an open-source agent harness and orchestration framework maintained by sean35mm on GitHub. Built on top of the Bun…
  • Context Bridge — Context Bridge is an open-source CLI agent harness and context orchestration utility developed by serdardb. It is built to facilitate…
  • Hivelore — Hivelore is an open-source agent harness and contextual memory orchestrator created by Doucs91. It provides persistent project memory, decision tracking,…
  • Agentic Engineering Framework — Agentic Engineering Framework is an open-source orchestration harness and governance layer developed by Dimitri Geelen. It is designed to constrain,…
  • AI Tools
  • Categories
  • Industries
  • CLI Coding Agents
  • MCP Servers
  • MCP Categories