Show HN: Kanban-md – File-based CLI Kanban built for local agents collaboration (github.com)

A new project titled **kanban-md** has been unveiled, designed as an agents-first, file-based Kanban tool that aims to streamline collaborative workflows among AI agents. Unlike traditional project management software, which often focuses on human interactions, kanban-md allows AI agents to execute commands and manage tasks directly using a fast, single binary command-line interface (CLI). This approach eliminates reliance on databases or servers by leveraging plain Markdown files to represent tasks, making it accessible and highly efficient for both agents and human supervisors. Significantly, kanban-md introduces features such as atomic claim-and-move operations, enabling cooperative task assignment without clashes among multiple agents. This is facilitated through a mechanism for automatic claim expiration and cooperative locking, ensuring agents can work concurrently without conflicts. The system includes built-in agent skills, making it easier for AI to understand and navigate the Kanban board. With zero runtime dependencies, rapid task updates, and optional terminal user interface (TUI) for visualization, kanban-md positions itself as a future-proof solution that could reshape AI-driven project management.