Claude Code multi-agent
Jmux
A tmux environment built for running coding agents in parallel — with a persistent sidebar that shows every session, what's running, and what needs your attention.
Install
git clone https://github.com/jarredkenny/jmux.gitFrom the README
jmux From ticket to shipped feature, in the terminal you already use. Define the workflow stages you actually think in, then run your whole backlog through a fleet of coding agents on them — Claude Code, Codex, or any agent, each in its own isolated session. jmux shows you who's working, who finished, and who's blocked on you. You're still the pilot. ▶ Watch the walkthrough — real agents, nothing staged. Ticket in, MR out — Linear · GitHub · GitLab, self-hosted GitLab and GitHub Enterprise included. Get running - Install — the command above needs no runtime; it installs a self-contained binary. Also on Homebrew (brew install jarredkenny/tap/jmux) and npm (bun install -g @jx0/jmux, which needs Bun 1.3.8+ and is the only option on Alpine/musl). - Requirements — tmux 3.2+. jmux offers to install it on first run. - Just want to look around? jmux --demo runs with mock data — explore every feature, no credentials needed. - New to tmux? Start with the Getting Started guide. Features Define your own workflow Your tracker has a lot of statuses — Ready for Development, Awaiting QA Sign-off, Pending Release Approval — most of them named for someone else's process. You don't think in 25 states. You think in five. So jmux has you define your own workflow stages and sit each one on top of one or many of your tracker's statuses: A stage shows up as a tab in the info panel, but that's how it *appears*, not what it *is* — it's one rung on the ladder you actually work by, and its place in the list is its priority. Every status then has exactly two settings: which stage it belongs to, and whether it parks. That second one is what keeps a fleet manageable: - Park what you've handed off. A teammate has it in review, QA has it, it's blocked on someone else — that session…
Details
- Repository
- jarredkenny/jmux
- Website
- jmux.build
- Language
- TypeScript
- License
- AGPL-3.0
- Last updated
- 29 August 2026
- Topics
- agentsclaudeclaude-codecodexidepiterminaltmux
FAQ
What is Jmux?
A tmux environment built for running coding agents in parallel — with a persistent sidebar that shows every session, what's running, and what needs your attention. It is a multi-agent for Claude Code, from jarredkenny/jmux on GitHub.
How do I install Jmux?
Clone jarredkenny/jmux from GitHub and follow the setup steps in its README.
Is Jmux popular?
jarredkenny/jmux has 84 GitHub stars and 13 forks. It was last updated on 29 August 2026.
More multi-agent & worktrees
Codex Orchestrator
91alexzh3/codex-orchestrator
Claude Code plugin that orchestrates GPT Codex agents for implementation, monitoring, and independent reviews.
OSA
79Miosa-osa/OSA
An AI agent for your terminal, built on signal theory: every output is signal, not noise. Codes, runs multi-step work with subagents, and works with any model, including your Claude or ChatGPT subscription.
Pumasi
85fivetaku/pumasi
Claude as PM + Codex CLI as parallel outsource developers — for large parallel coding tasks
Omux
101Happenmass/omux
Orchestrate AI coding agents (Claude Code, Codex) as parallel subagents over tmux — a loop-engineering runtime with auto-continue, execute-then-review, and cross-session memory.
Claude Agents Coordination
87ilyasibrahim/claude-agents-coordination
Multi-agent coordination system for Claude Code with institutional memory
OrbitDock
93Robdel12/OrbitDock
Run, review, and orchestrate AI coding agents from anywhere
Browse the most popular Claude Code mods, all multi-agent & worktrees, or the skills and MCP server directories.