The control room for your coding agents
Tens of coding agents in one Zellij or tmux room, with state, task, context health, and live cost at a glance.
curl -fsSL https://raw.githubusercontent.com/rimio-ai/rimz/main/scripts/install.sh | shbrew install rimio-ai/rimz/rimzcargo install --locked rimz
The sidebar triages the fleet on the left; agents work in their own panes.
- Claude Code
- Codex
- Pi
- OpenCode
- Antigravity
- Copilot
- Droid
- Cursor
- Amp
- Kiro
- Qwen
- Kimi
- Grok
The agents you already run, stock: your flags, your config, and the official apps keep working. Claude Code and Codex are the daily drivers; the lighter marks are alpha or experimental. Per-agent detail
Why RimZ?
One lightweight binary inside the terminal you already use, and the whole fleet reads at a glance.
See the fleet
- Attention, routed
- The cockpit line reads the whole fleet, the column below arrives already triaged, and one click drops you into the pane that is waiting.
- Live agent cards
- Working state and current task, model and effort, context health and compactions, live token stats, and the subagent tree.
- Know your pace
- Spend and token insight by day, week, and month, with plan and budget-window bars for every provider that exposes them.
- Extremely lightweight
- A single binary that hooks the agents you already run, inside the Zellij or tmux you already use. Same keybinds, same terminal.
Drive the work
- Teams, cross-model
- Pair a planner on one model with a coder on another and launch them as one team, each role on the model best at its job.
- A worktree per agent
- Open agents side by side in an isolated worktree with a dynamic layout: your editor, an agent, and a shell in one tab.
- Handles and messages
- Every agent answers to a handle. Park at the turn boundary, steer the live turn, or schedule for later; agents talk to each other.
- Scriptable, end to end
- One headless grammar for every agent, with exit codes, JSON output, streaming, and the full transcript kept for scripts and CI.
Keep it running
- Loops on a clock
- Calendar, interval, cron, or a check-guarded watchdog that runs a command and wakes an agent on the result.
- Auto-continue
- A rate-limit pause resumes the moment the budget window resets, and transient overload retries on a backoff ramp.
- Local or remote
- Start on your laptop or a server, close the lid, and reattach from anywhere. The link heals itself every time you reconnect.
- Steer from your phone
- When an agent stops to ask, the question reaches the official Claude and ChatGPT mobile apps, and your answer lands in the session.
Everyday moves
These run from any pane in the room, and from any script or CI job that reaches it. They compose: a profile becomes a team, the team lands in a worktree, the worktree’s agents take messages, and a schedule fires the whole thing while you sleep.
Every supported agent joins the room the same way: type its own command into any pane. When you want more than the default, rimz agents adds permission modes, reusable profiles, layouts, and isolated worktrees: , splits, + tiles, / stacks.