Agent Board plugin
A shared board for Claude Code, Codex and other agents: a question an agent stops on rises to the top within seconds.
- 58 ★ on the repository
- Last commit
- MIT
Extended access
Why this level
- Setup changes agent hook settings, and a board reply is fed to the agent as its next input
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add plugins/agent-boardDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- Setup changes agent hook settings, and a board reply is fed to the agent as its next input
This is third-party code. Review the repository files before installing.
What it does
Agent Board gathers reports from several agents on one local board. When an agent stops and asks a question, that question lands on top with a button for each option the agent could read, or you write your own. Agents keep running in their own terminals and the board reads what their hooks report. Data sits in one Markdown file, `progress.md`, which you can read and edit. The page listens only on `127.0.0.1`, no account is needed, and the board itself makes no model calls. The `setup` command shows what it will change and asks once before writing, and `remove` takes out what it wrote and leaves your settings alone.
Who it is for. Developers who keep several Claude Code, Codex or Cursor sessions open and miss questions in the windows.
Good fit when
- Five agents work in different terminals and you need to see at once who is waiting for an answer
- An agent asked which option to take and it is easier to press a button on the board
- Progress of several agents has to be followed through one readable file
- After updating agent settings the connection must be removed with one command
Not a fit when
- Cost and token metrics are needed, since the board tracks questions and statuses, not spending
- Agents run only in desktop apps or IDEs, where a board reply does not yet reach the agent
- A shared team board for several people is needed rather than a local file on one computer
Example request
Open the Agent Board and show which agents are waiting for my answer right now.Limitations
Node.js 22 or newer is required. A board reply reaches Claude Code when the agent is free, and is handed to Codex and Antigravity from the next turn, one per turn. In Cursor, OpenCode, Grok Build and Hermes Agent the reply is pasted by hand. In desktop apps and IDEs a board reply does not yet reach the agent. Setup changes agent hook settings after confirmation. The guide describes what the Claude Code plugin may send off the machine. The install version is tagged `@next`.
How to disable. Run `npx @wenyuchiou/agent-board@next remove`, review the plan and repeat with `--apply`. Delete `progress.md` by hand if it is not needed.
FAQ
Which agents are supported?
Claude Code, Codex, Antigravity, Cursor, OpenCode, Grok Build and Hermes Agent. In a terminal a reply reaches Claude Code, Codex and Antigravity without pasting.
Where is data stored?
In one `~/.agent-board/progress.md` file for a fresh install. The format is in `docs/FORMAT.md` and the file can be read and edited by hand.
An agent is missing from the board, what now?
Start a new agent session since one already open is not connected. For Codex the README has a separate trust check with a preview and the `--apply` flag.
An agent shows up twice, why?
It is probably connected both by the Claude plugin and by `setup`. What to remove is described in the troubleshooting section of the guide.
Is an account or model needed?
No. The page is local, there is no account, and the board itself calls no model.
Discussion
No questions or reviews yet.
Related
PluginsA plugin for Claude Code and Codex: a shared task wall, project memory, reports and a dashboard that shows agents' work across sessions
A pixel-art office in the browser: each Claude Code or Codex session sits at a desk, types and raises a hand when it needs you.
A CLI for AI agents sharing a repository: a task board, a spec and acceptance by a second agent against a specific commit
A local bridge linking Claude Code and Codex in one session so the agents message each other and review each other's work
A local whiteboard and MCP server for research agents: every claim is tied to the run and the code that produced it
Local meeting rooms for Codex and Claude Code agents to talk to each other while a person reads and posts from a browser.
Comparison with similar tools
Cards that share the main search query, and how they differ.
| Tool | Official | Repository stars | Main difference |
|---|---|---|---|
| Agent Board (this card) | no | 58 | |
| Agent Kanban | no | 502 | manages agent tasks on a kanban board; Agent Board collects questions and replies from session status |
| Claude Squad | no | 8,790 | launches parallel sessions; Agent Board only watches and takes replies |
README in short
The board puts questions ahead of logs. Everything is kept in one file, and install and removal require confirmation.