How do I use Gemini CLI in a headless environment like Claude Code?

You must use the headless mode by providing a prompt via the -p flag (e.g., gemini -p 'prompt') or by passing data via stdin to ensure it functions correctly in non-TTY environments.

Which Gemini model is recommended for daily coding tasks?

The gemini-3-flash-preview model is the default and recommended choice for standard reviews and batch operations due to its speed and efficiency.

What is the recommended timeout for complex code analysis?

For deep architectural or security analysis, especially when using the Gemini 3 Pro model, a timeout of 600,000ms (10 minutes) is recommended.

Can I parse the results of a Gemini analysis programmatically?

Yes, by using the --output-format json flag, the skill returns a structured response that can be easily parsed with tools like jq for automation scripts.

Gemini CLI Integrator

Name: Gemini CLI Integrator
Author: creator-hian

bycreator-hian

•

開発者ツール

Empowers Claude Code with Google Gemini capabilities for high-performance code analysis, automated reviews, and validation workflows.

概要

The Gemini CLI Skill bridges the gap between Claude Code and Google's Gemini models, enabling developers to perform deep code analysis, architectural reviews, and security audits directly from their terminal. By supporting both the rapid Gemini 3 Flash and the sophisticated Gemini 3 Pro models, this skill offers flexible execution modes including headless operation for CI/CD pipelines and structured JSON output for automated scripting. It is an essential tool for developers who require specialized Gemini-driven insights to augment their local development environment and validate complex codebases with high precision.

主な機能

Cross-directory analysis with the ability to include multiple context directories
Support for dual models: Gemini 3 Flash for efficiency and Gemini 3 Pro for complex logic
Advanced session management including the ability to save and resume analysis states
2 GitHub stars
Structured JSON and JSONL output formats for reliable script parsing and automation
Headless execution mode for seamless integration with Claude Code and CI/CD pipelines

ユースケース

Automating security audits and bug detection within deployment pipelines
Conducting deep architectural reviews across multi-repository projects
Generating structured code metrics and documentation via automated scripts

概要

主な機能

Cross-directory analysis with the ability to include multiple context directories
Support for dual models: Gemini 3 Flash for efficiency and Gemini 3 Pro for complex logic
Advanced session management including the ability to save and resume analysis states
2 GitHub stars
Structured JSON and JSONL output formats for reliable script parsing and automation
Headless execution mode for seamless integration with Claude Code and CI/CD pipelines

ユースケース

Automating security audits and bug detection within deployment pipelines
Conducting deep architectural reviews across multi-repository projects
Generating structured code metrics and documentation via automated scripts