Kandev is an AI kanban board and development environment for orchestrating multiple coding agents in parallel: assign tasks, watch agents work, review their changes, and ship pull requests from one workspace. It is open source under AGPL-3.0, self-hostable with no telemetry, and has earned just under 1,000 stars on GitHub, with active development continuing in October 2026.
Terminal agent TUIs are powerful, but reviewing and iterating on their output does not scale โ Kandev is built around closing that gap. Work is organized across kanban and pipeline views with opinionated workflows; you can assign agents from multiple providers, customize workflows, agent profiles, runtimes, and prompts, and set review gates that match your team’s process. The integrated workspace combines a file editor, file tree, terminal, browser preview, and git changes in one place, so reviewing an agent’s diff happens where the task lives. Kandev ships as a native Go binary, runs locally or self-hosted on your own infrastructure, and supports mobile remote access through Tailscale, Cloudflare Tunnel, or another private VPN.
Kandev’s stance on deployment is deliberate: because it is self-hosted and telemetry-free, teams can run agent orchestration inside their own network and keep source code, prompts, and diffs on infrastructure they control โ a requirement for many companies experimenting with autonomous coding. The native Go binary distribution keeps installation simple, avoiding container or runtime dependencies, and the project uses git worktrees so parallel agents can work on isolated copies of a repository without stepping on each other’s changes.
Kandev is free and open source under the AGPL-3.0 license. You run it yourself and pay only for the agent providers you plug into it.
Kandev is for power users and small teams running several coding agents at once who have outgrown juggling terminal windows and want a real review workflow.