Skip to main content

Gemini CLI Developer Guide

What is this about?

Gemini CLI is an open-source terminal agent from Google for coding and general shell-assisted work. This guide explains how Gemini CLI is shaped, how to set it up, how its runtime and permission model fit together, and where to contribute.

Checked against primary sources

This guide is based on the official Gemini CLI docs and the official google-gemini/gemini-cli repository, checked on June 26, 2026.

In this section​

1. What Gemini CLI is​

The shortest accurate description is:

Gemini CLI is Google's open terminal agent for running Gemini-powered workflows close to the shell, filesystem, and developer toolchain.

That puts it in the same broad family as Codex CLI, Claude Code, Qwen Code, and Crush.

2. The mental model​

Think of Gemini CLI as four cooperating layers:

LayerWhat it does
terminal UXMain interactive command surface
agent runtimePlans and executes file and shell work
Gemini provider layerConnects to Google's models and auth flow
extensions and docsBroaden use cases and contributor entry points

3. Why developers would choose it​

  • It is the official Google-native CLI for Gemini models.
  • It is open source, so teams can inspect how the agent behaves.
  • It is built for terminal workflows, which suits engineers well.
  • The docs emphasize not just usage, but also development, configuration, and trust.

4. When it fits best​

Gemini CLI is strongest when you want Gemini reasoning close to:

  • codebases,
  • shell commands,
  • local files,
  • repeatable terminal routines.
  1. this overview,
  2. install and quickstart docs,
  3. authentication and configuration docs,
  4. permission and trust docs,
  5. the repo structure after the runtime model is clear.

Sources​