KanBots has launched an open-source desktop Kanban app that runs parallel AI agents on every card, available for macOS, Linux, and Windows. The app, released under the MIT license, allows users to drop a folder and instantly generate a board where multiple AI agents like Claude Code or Codex can work simultaneously on tasks, automating workflows efficiently (kanbots.dev).
Users can dispatch these AI agents on as many cards as needed, each operating in its own isolated worktree. The app also features an autopilot mode where different AI personas split tasks, run in parallel, and self-verify their outputs. This setup enables continuous progress on multiple fronts without manual intervention, streamlining project management and development processes.
This innovation is significant in the productivity and AI collaboration space, combining Kanban’s visual task management with AI’s parallel processing capabilities. By integrating AI agents directly into task cards, KanBots offers a unique approach to automating complex workflows, potentially reducing bottlenecks and increasing efficiency. Its open-source nature and pay-what-you-can donation model may encourage widespread adoption and community-driven improvements.
KanBots plans to maintain free access to the desktop app while exploring cloud options for team collaboration. Future updates may focus on enhancing AI persona capabilities, improving integration with existing development tools, and expanding platform support. Users and developers can track progress and contribute via the GitHub repository.