This project provides a command-line interface (CLI) for orchestrating multiple AI coding agents in parallel, leveraging Git worktrees for isolated development environments. It's designed for developers who need to scale AI-assisted coding tasks, such as generating code, refactoring, or fixing bugs, across a codebase efficiently.

How It Works

Uzi manages AI agents by creating separate Git worktrees for each agent, ensuring isolation of dependencies and code changes. It integrates with Tmux for managing terminal sessions per agent and handles development server setup, including port allocation using a configurable port range. The CLI facilitates starting agents with prompts, monitoring their progress in real-time, broadcasting instructions, and merging completed work back into the main branch via Git commits.

Quick Start & Requirements

• Install via Go: go install github.com/devflowinc/uzi@latest
• Ensure $GOBIN is in your PATH: export PATH="$PATH:$HOME/go/bin"
• Prerequisites: Git, Tmux, Go, and an AI tool (e.g., Claude, Codex).
• Configuration: Create uzi.yaml for devCommand and portRange.

Highlighted Details

• Parallel execution of multiple AI coding agents.
• Automatic Git worktree management for isolated development.
• Tmux session management for each agent.
• Real-time monitoring and status updates.
• Automatic handling of agent prompts and confirmations.

Maintenance & Community

No specific community links or notable contributors are mentioned in the README.

Licensing & Compatibility

The README does not explicitly state the license.

Limitations & Caveats

The project requires specific tools like Git, Tmux, and Go to be installed and configured. The devCommand in uzi.yaml must include all necessary setup steps for each agent's isolated environment.