Catenary
Spatial IDE for AI coding agents — orchestrate Claude Code, Codex, Cursor, and more with visual wires and parallel branches.
| What is it | Spatial IDE for AI coding agents — orchestrate Claude Code, Codex, Cursor, and more with visual wires and parallel branches. |
|---|---|
| Pricing | Freemium |
| Free tier | Yes |
| Platform | Desktop Application |
| Best for | orchestrating multiple AI coding agents on one canvas, running parallel experiments with isolated branches |
| Domain registered | 2026 |
Data updated Sept. 9, 2026
What does Catenary do?
Catenary is a desktop IDE built to manage multiple AI coding agents on one infinite canvas. You can drop terminals, a Monaco editor, a browser panel, and git worktrees onto the same surface, then wire them together visually. The tool supports agents like Claude Code, Codex, Cursor, Grok, OpenCode, and Pi — each running in its own isolated workspace. Your code never leaves your machine; the app is local-first, requires no account, and sends zero telemetry.
The standout feature is how it handles parallelism. The “1-Click worktrees & branches” button spins up an isolated branch, a separate working directory, and a dedicated agent for a new task — none of it touches your current code. Each task gets its own color, so you can keep track in the sidebar and on the canvas. The built-in Monaco editor shows a side-by-side diff of every change the agent makes, so you can accept or reject before anything lands. Context wires let you drag a connection from one panel to an agent, setting up automatic context relay. Maestro Mode lets one agent recruit a team of up to ten helpers; every action the team takes opens an editable approval card before it appears on the canvas.
Catenary is for developers who want to run several AI coding experiments at once without losing track. It works well for shipping multiple features in parallel, refactoring in isolated branches, or letting an agent team debug and review code under your supervision. The free plan gives you one project, two parallel tasks, and two Maestro helpers. The paid Pro plan removes those limits with a one-time purchase — no subscription, no recurring bill.
Key features
What makes it stand outWho is Catenary for?
Who benefits most from this toolPricing
Free tier available — start without a credit cardFree
- — activation
- 1 projects open
- 2 parallel tasks
- 4 of 5 (up to 3 agents) built in squads
- — your own squads
- 2 maestro mode helpers
- Canvas, terminals, editor, browser, git: Unlimited
- Connection wires: Unlimited
- Projects open at the same time: 1
- Parallel tasks: 2
- Maestro Mode helpers: 2
- Built-in squads: 4 of 5 (up to 3 agents)
- Your own squads: —
- Activation: —
- No account, no card, no telemetry
- Canvas, terminals, editor and git with no time limit
- Your own API keys, never routed through us
Pro
- Up to 2 personal computers activation
- Unlimited projects open
- Unlimited parallel tasks
- All 5 built in squads
- Unlimited your own squads
- 5 (adjustable 2–10) maestro mode helpers
- Canvas, terminals, editor, browser, git: Unlimited
- Connection wires: Unlimited
- Projects open at the same time: Unlimited
- Parallel tasks: Unlimited
- Maestro Mode helpers: 5 (adjustable 2–10)
- Built-in squads: All 5
- Your own squads: Unlimited
- Activation: Up to 2 personal computers
- Up to 2 computers at a time — move between machines whenever you like
- 1.x line updates included
- One activation and done: after it, the key is checked offline
Trust & presence
Gallery
Click any image to enlargeAlternatives in Developer Tools
Desktop app to orchestrate and run AI coding agents autonomously for hours or days on complex tasks.
A macOS desktop app that organizes multiple AI coding agents on an infinite canvas for better workflow management.
Run Claude Code, Codex, Cursor and other AI coding agents in one desktop, terminal, and iOS app
Desktop app to run multiple AI coding agents (Claude Code, Codex) in parallel on your Mac.
Native macOS app to manage Codex and Claude Code CLI sessions in one workspace with split panes, worktrees, and embedded simulators.
Open-source desktop workspace to run multiple AI coding agents in parallel, isolated in Git worktrees, with a review-first workflow.
A desktop IDE for running multiple AI coding agents (like Claude Code) in parallel with integrated terminals and worktrees.
Terminal UI to spawn, monitor, and review multiple AI coding agent sessions from one keyboard