agentpane
See every subagent at work beside the conversation: what it's doing and its tokens.
Illustration. Actual results depend on the mod, terminal and environment.
Install
/plugin marketplace add xuanji86/claude-agentpane
/plugin install agentpane@claude-agentpane
/reload-pluginsRun these commands one at a time in a Claude Code session.
claude plugin marketplace add xuanji86/claude-agentpane
claude plugin install agentpane@claude-agentpaneThen run /reload-plugins in your session, or start a new session.
Please install the Claude Code mod "agentpane" (https://github.com/xuanji86/claude-agentpane). Before installing, read the Security and Requirements sections of its README and tell me what it reads and writes, which commands it runs, whether it uses the network, and which Claude Code version, OS and terminal it needs. If everything looks fine, run claude plugin marketplace add xuanji86/claude-agentpane and claude plugin install agentpane@claude-agentpane, then remind me to run /reload-plugins in my session.Send this to Claude Code so it reads the author’s instructions before installing.
git clone https://github.com/xuanji86/claude-agentpane
cd claude-agentpane
claude --plugin-dir .Clone the repository and try it in a local session with --plugin-dir.
What it accesses
Based on the author’s documentation, checked when listed; use claude plugin validate to check for yourself.
Reads
YesReads the agent list and status, tool calls, token usage, and an agent's transcript while open.
Writes files
NoneRuns commands
NoneNetwork
NoneIntroduction
agentpane is a Claude Code mod that keeps the session's subagents in a side pane, drawn the way Claude Code draws its own rows.
- Live list: each agent's current tool call, tool count, elapsed time, model, and input/output tokens; nested agents are indented.
- Timeline and receipt: the latest batch on one time axis, then a receipt with agents, agent time, parallelism, tool uses and tokens.
- Conversations: click an agent to read its brief, replies and tool calls in the pane.
- Stop:
■ Stop, then confirm, ends a running agent through Claude Code's TaskStop. - Opens and folds itself: opens when an agent starts, folds 10 s after the last ends to a tab above the prompt; a status line and toasts cover the rest.
The author states it makes no network requests, runs no processes, reads and writes no files, stores nothing across sessions and calls no model. Requires Claude Code 2.1.287+ (tested on 2.1.288); the side pane needs /tui fullscreen, otherwise a summary sits above the prompt.
Who it is for
- People who fan work out to subagents and want to watch progress and cost
Requirements
- Claude Code ≥ 2.1.287
- The side pane needs /tui fullscreen; otherwise a summary sits above the prompt.
- It opens itself from 144 columns; /agentpane opens it at any width.