iCode Agent · FAQ
Questions, answered.
How do I install iCode Agent?
Install Node.js 22 or newer, then run npm install -g @icodehq/icode-agent. See Get started.
How are iCode Workbench and iCode Agent different?
iCode Workbench (in early access) is the desktop ADE: it runs Claude Code, Codex, and OpenCode in parallel, each in its own Git worktree, with tasks, review, and Mira across projects. iCode Agent is the focused terminal agent described in this FAQ. They are separate products, and either works on its own.
Do I need Workbench to use Agent?
No. iCode Agent is a standalone CLI — install it with npm and run it in any repository. Workbench does not currently wrap or launch iCode Agent.
Which models can I use?
iCode Agent speaks the openai and anthropic protocols, so it works with the official APIs and any compatible endpoint. Configure one with /model.
Do I need Bun?
No. Bun is the development runtime for working on iCode Agent itself. The published package runs on Node.js 22 or newer.
Which platforms are verified?
Cross-platform compatibility is implemented for Windows, macOS, and Linux. Automated verification is performed on Linux; native Windows and macOS verification is still pending.
Does iCode Agent upload my code?
It runs locally and talks only to the model endpoint you configure, using the API key you provide. Your source is sent to that endpoint as context for the tasks you run; the provider's own data policy applies.
What does it cost to run?
iCode Agent itself is free and open source (Apache-2.0). You pay only your model provider for the tokens you use — and iCode Agent is built to keep that number down on long tasks.
How does iCode Agent keep token usage down?
Context is managed at every boundary: output is capped as it is produced, older tool results are reduced only when a request would exceed its budget, and overflowing history is summarized while keeping a recent tail. See how context is managed.
Does reducing context lose information?
A withheld result is replaced by a marker that names where the full output still lives, so the agent can read it back. Failure evidence and errors are never withheld, and reduction only changes the request that is sent — the saved session keeps every result.
Where is my work saved?
Sessions are saved automatically under your config directory (~/.icode-agent/). Resume the latest with --continue, a specific one with --resume <id>, or manage them with /session.
Which agents or editors does it work with?
iCode Agent is its own terminal agent — you use it directly, not through another agent. Its configuration and sessions are read by the CLI and TUI only.
Is it ready for production use?
iCode Agent is in early experimental development and changes quickly. See the changelog for what each release includes.