Getting started

octomux documentation

octomux is a local dashboard for running many coding agents in parallel. Each agent works in its own git worktree; you get one inbox for every permission prompt, a live grid of the whole fleet, and in-app diff review. It runs on your machine — no cloud, no telemetry, MIT.

# install, initialise, and start in your repo
cd your-repo
npm install -g octomux && octomux start   # → localhost:7777

What it does

Four acts, one window — from prompt to merged PR.

① DispatchType a task, or paste a Jira / Linear / GitHub link, or a whole list. Each one gets its own worktree, branch and agent. Pick the model and harness per task.

② Run unattendedAttach a verify command and the agent loops in fresh context until it exits 0. Put it on a cron and it runs at 3am. Hand a goal to the orchestrator and it plans and splits the work.

③ Watch, or don’tEvery agent’s live terminal on one grid, the diff as it grows, and every “allow this tool?” prompt collected in a single inbox rather than scattered across panes.

④ Review & shipAn agent drafts a walkthrough and inline comments against the real diff. You read it in-app and publish as one batched review when you accept it.

Crash, reboot, close the lid — octomux start restores every task, branch and session.

Where to go next

QuickstartInstall, requirements, and your first task in about a minute.
Read →

Core conceptsTasks, worktrees, harnesses, workflows and integrations — the words the rest of the docs use.
Read →

PluginsAdd a harness, a tracker or a scheduled job as an npm package.
Read →

CLI referenceEvery command, for scripting and for driving octomux from an agent.
Read →

A note on status

Some of what octomux.com describes is still being built. These docs mark anything not yet released with coming. Anything marked shipped works in the version on npm today. If you find a gap between the two, that is a bug in the docs — open an issue.