ทันเอไอ

Introducing agent-manager: A TUI tool for managing multiple AI coding agents simultaneously

Yoan Wainmann launches agent-manager, a tool to control multiple AI coding agents from the terminal interface.

In modern software development, many developers have started using multiple AI coding agents simultaneously to boost productivity. However, switching windows and managing multiple sessions at once can be cumbersome. Software developer Yoan Wainmann has recently released 'agent-manager', a Terminal User Interface (TUI) tool specifically designed to solve this problem.

Developed in Go and open-sourced under the Apache-2.0 license, agent-manager acts as a central hub to manage and switch between multiple AI coding agents concurrently, such as Claude Code, OpenCode, Codex, Grok, and Gemini CLI.

Architecturally, the tool runs on top of tmux to ensure that AI agent sessions persist even if the terminal window is closed. It supports macOS, Linux, and Windows (via WSL2).

Key features include a live status dashboard categorized by project, keyboard-driven controls, and a code diff review system via the Ctrl+r shortcut. Users can add comments to specific lines of code to send back as prompts, enabling agents to fix issues with greater precision.