KAGAN - Terminal AI Orchestration Layer – AI-driven Kanban for managing coding agents

Summary: KAGAN is a terminal-based Kanban TUI that orchestrates AI coding agents like Claude Code and OpenCode to autonomously or collaboratively complete development tasks. It converts natural language descriptions into actionable tickets, runs tasks in isolated git worktrees, and integrates code review for human oversight before merging.

What it does

KAGAN organizes AI agent workflows as discrete Kanban tickets within the terminal, supporting two modes: PAIR for real-time collaboration via tmux sessions and AUTO for parallel background execution. It breaks down complex requirements into scoped tasks and ensures safe integration through isolated git worktrees and review agents.

Who it's for

It targets indie developers and teams experimenting with AI-assisted workflows who need to manage multiple agentic CLI tools without losing context or manual orchestration.

Why it matters

KAGAN addresses the challenge of coordinating AI agents and human judgment by providing a structured interface that reduces context switching and streamlines collaboration on coding tasks.