agent-deck
A terminal control center for managing multiple AI coding agents in one place.
Install & Use
Copy this prompt and send it to your AI assistant (Claude / Cursor / TRAE / Codex / WorkBuddy etc.) to auto-install:
Help me install this AI Skill: agent-deck. It is used for: A terminal control center for managing multiple AI coding agents in one place. Full Skill content: https://321skill.com/skills/agent-deck/raw/index.md Read that page and install it.
The prompt includes a link to the full Skill content. You can also view the full content.
When developers run multiple AI coding assistants (e.g., Claude Code, OpenCode) across different projects, constantly switching terminals, managing session states, and tracking costs becomes cumbersome. Agent Deck addresses this pain point by providing a unified terminal interface, allowing you to monitor the status of all AI agents, switch sessions, manage project context, and track spending—all within a single view.
Using Agent Deck is straightforward. After installation, simply run the agent-deck command in your terminal to launch it. The tool automatically discovers and integrates your active AI coding agent sessions, presenting them in a clear grid view. You can quickly switch between sessions using keyboard shortcuts, create session groups, search history, and even utilize Git worktrees to isolate code environments for different branches.
This tool is ideal for developers juggling multiple projects or frequently using different AI coding assistants. Whether you're a full-stack engineer switching between front-end and back-end projects or an agent developer debugging and orchestrating multiple sub-agents, Agent Deck significantly improves the efficiency and clarity of multi-task management.
For first-time users, we recommend starting with one project and one AI agent to familiarize yourself with the interface and shortcuts. For team collaboration, you can leverage its session grouping and state-sharing features. Note that Agent Deck primarily manages the session interface and state of AI coding agents; it does not replace the agents' own code-generation capabilities.
Key Features
Unlike single AI coding tools, Agent Deck is a "meta-management tool" at its core. It does not generate code but instead unifies and visualizes all active AI coding agent sessions, providing a centralized control panel akin to a task manager, with native support for Git worktrees and cost tracking.
Limitations
Not suitable for users who only use a single AI coding tool or primarily work in graphical (non-terminal) interfaces.
FAQ
Which AI coding agents does Agent Deck support?
It is designed to manage AI coding agent sessions running in the terminal, such as Claude Code, OpenCode, etc. Please refer to the project documentation for the specific supported list.
Does using Agent Deck affect the functionality of my existing AI agents?
No, it is solely a management and monitoring layer and does not impact the code generation or interactive capabilities of the agents themselves.
Installation guide for AI assistants
If your AI coding assistant (Claude Code, Cursor, TRAE etc.) can see this page, send it this message to auto-install:
Visit https://321skill.com/skills/agent-deck/raw/index.md to read the original Skill definition (Markdown format) for agent-deck, and install it according to the instructions.
Raw Markdown URL for AI: /skills/agent-deck/raw/index.md