Terminal tabs are a bad interface for running multiple AI coding agents. Catenary is an attempt to fix that with an infinite 2D spatial canvas that replaces the tab pile entirely.
Built by indie developer Giorgio Nícolas, Catenary launched this week on Product Hunt. The core idea is that agent workflows are visual by nature: you have models handing off context to other models, editors sitting next to terminals, browser previews running alongside tests. A flat list of tabs does not map to that. A spatial canvas does.
What It Includes
- Visual Cables: Wire agent terminals together to pipe context between them and chain multi-model setups, such as a heavy reasoning model feeding results to a fast test runner.
- Task Islands: Spin up isolated visual workspaces in one click without losing your current layout.
- All-in-One Canvas: Monaco editor, native PTY terminals (WebGL
xterm.js), and local browser previews on the same surface. - Local-First: Zero telemetry. Runs entirely on your machine. Your code and API keys never leave your device.
️ Pricing and Availability
Catenary is free to download. It runs on macOS (Apple Silicon and Intel), Windows, and Linux. The download is at thecatenary.app.
If you run Claude Code or any other terminal-based agent and you have hit the wall of overlapping windows, this is worth a look.
