How-to4 min read
Claude Code vs. OpenCode vs. Codex: running all three side by side
Three agent CLIs are installed on every machine. Nothing stops you from using them at once.
Every machine ships with more than one agent CLI installed: Claude Code, OpenCode, and Codex. They are all terminal programs on the same box, so nothing stops you from running them at once — on different tasks, or on the same one to compare.
Three CLIs, one machine
Because they share the machine, they share the machine's context: the same repositories, the same toolchain, the same files. Switching between them is switching terminals, not switching environments.
Keep them in separate panes
The persistent workspace is the natural home for this. A pane each keeps their output from tangling:
- One agent on a feature branch while another reviews a different one.
- The same task handed to two, so you can read both diffs before choosing.
- A long agent run in one pane, a shell you keep working in beside it.
One caution
Two agents editing the same files at the same time will step on each other, the same as two people would. Give each its own branch or its own directory and let them work in parallel without a collision.
More posts
How-to4 min read
How to run Claude Code in a browser, on any device
The agent CLI is already installed and running on the machine — you reach it through the same tab as everything else.
How-to4 min read
How to keep an AI agent running 24/7 after you close the tab
Detach from a session instead of ending it, and a long job keeps running while you are gone.
How-to5 min read
Running Docker on a cloud dev machine, with root on a real VM
You have root on a real VM, which means Docker runs the way it does anywhere else.