Orca: the ADE for orchestrating a fleet of coding agents in parallel
What do you do when Claude Code proposes one solution, Codex a different one, and Gemini a third? The usual move is to pick an agent up front and take whatever it produces. Orca changes that dynamic: it runs all three at once, each isolated in its own git worktree, and lets you compare the results and merge the one that won. It's an open-source ADE (Agent Development Environment)—MIT, by Stably…
Orca is an open-source Agent Development Environment (ADE) designed to run multiple coding agents in parallel, each in its own isolated worktree. This ADE, built by Stably AI under an MIT license, allows users to compare the results of different agents and merge the one that performs best. Orca supports various CLI agents such as Claude Code, Codex, Gemini, Grok, Cursor CLI, OpenCode, and 25+ more, and can be installed via Brew or AUR.
It functions on macOS, Windows, Linux, iOS, and Android devices. Orca's primary user is the agent operating the environment, while the user's role shifts to coordinating, reviewing, and deciding. The tool contrasts with traditional IDEs, which are designed for a single person to read, write, and debug code. Orca's primary function is not to replace individual agents but to manage a fleet of them, each with its own terminal, browser, and context. This approach shifts the focus from one agent solving a task to a coordinated fleet of agents.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
