gemini

Invoke Google's Gemini AI from the command line in headless mode.

7|Updated Dec 25, 2025

Search Tags:#analysis #CLI #gemini #headless #multi-modal #code-generation #prompting

Ready to run

npx skills add https://github.com/osoleve/the-fold --skill gemini-osoleve

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: gemini
Source: https://github.com/osoleve/the-fold/tree/main/.claude/skills/gemini
Command: npx skills add https://github.com/osoleve/the-fold --skill gemini-osoleve

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill provides a straightforward way to invoke Google's Gemini AI from the command line in headless mode for AI-powered analysis, code generation, and reasoning tasks.

Core Features & Use Cases

Headless invocation: Run prompts non-interactively with optional model selection.
Output control: Retrieve results in JSON, text, or streaming formats for easy parsing.
Use cases: Quick analysis, code review, multi-modal tasks, and specialized prompts when Gemini is required.

Quick Start

Use the Gemini CLI to run a headless prompt, e.g. gemini -m gemini-3-pro-preview "Analyze this dataset for anomalies" --output-format json.

Frequently Asked Questions about gemini

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I run Gemini AI in headless mode from the command line?▼

You can run Gemini AI in headless mode by using the CLI to execute non-interactive prompts, allowing you to specify models like gemini-3-pro-preview and configure output formats for automated analysis.

Can I get JSON output from a CLI-based Gemini prompt for code generation?▼

Yes, you can retrieve Gemini CLI results in JSON format by using the output formatting options, which also supports text or streaming formats to easily parse AI-powered code generation responses.

When do I need multi-modal analysis beyond Claude's scope using Gemini?▼

You need multi-modal analysis with Gemini when a scenario explicitly requires Google AI capabilities or reasoning tasks that exceed Claude's scope, utilizing headless invocation for specialized prompts.

How to select different Gemini models for non-interactive AI analysis?▼

You can select different Gemini models for non-interactive analysis by passing the model selection flag in your CLI command, ensuring deterministic invocation tailored to your specific reasoning tasks.

Does headless Gemini invocation support configurable options for custom context?▼

Headless Gemini invocation supports configurable options for prompts and context, enabling deterministic CLI execution with selectable models and specific output formatting for targeted code review tasks.