Up and running in two minutes.

1 · Install

Grab the macOS app, or install the CLI with one command:

The CLI and the desktop app share one engine — use whichever you like.

2 · Connect your agents

From the repo you want to coordinate, install the adapters once:

regente install --scope repo --tools codex,claude-code,gemini

This adds Regente's MCP server and session hooks to each agent's config — so Claude Code, Codex, Gemini CLI, and Cursor auto-join the workspace when they start. Already-running sessions just need a restart.

3 · Work like you already do

Start your agents in the repo — nothing changes about how you run them:

cd ~/acme-app
claude          # joins as it starts
codex           # joins too
cursor-agent    # and Cursor

Give each a lane with a role (optional but powerful):

regente role --agent claude --role frontend
regente role --agent codex  --role backend

4 · Watch the fleet

Open the desktop app (or run regente serve) to see every agent, the files each is editing, a live collision band the instant two overlap, and the auto-handoff queue that resolves it:

# the moment a second agent reaches for a claimed file:
⛔ blocked · pre-merge — claude vs codex on public/index.html
⤳ codex queued #1 · auto-handoff on release

# ...the instant claude releases:
↳ auto-handed off public/index.html to codex (was waiting)

The blocked agent isn't stuck — it's queued and handed the file automatically when the holder is done. On the desktop app you'll also get a native notification, even with the window closed. That's the whole point: no two agents edit the same file — and nobody waits idle.

From the terminal

Everything the app shows is available from the CLI:

regente status        # fleet, claims, conflicts
regente claim src/api.ts --agent codex
regente roles         # who owns which lanes
regente metrics       # activation, cross-tool, collisions caught