Gemini Skill

Gemini CLI Skill

[](https://agentskills.io/) [](https://opensource.org/licenses/MIT) [](https://skillzwave.ai/skill/SpillwaveSolutions__gemini-skill__gemini__SKILL/)

Shell out to Gemini CLI from Claude for headless AI task execution. This skill enables Claude to leverage Gemini 2.5 Pro alongside its own capabilities for parallel analysis, alternative perspectives, and batch processing.

Table of Contents

• Overview
• When to Use
• Quick Start
• Core Command Structure
• Model Selection
• Common Use Cases
• Prompt Engineering
• Workflow Patterns
• Troubleshooting
• Installing with Skilz
• Resources

Overview

This skill provides a framework for executing non-interactive Gemini tasks where content is piped to the Gemini CLI, processed with custom prompts, and results are saved directly to files. Use it to combine Claude's and Gemini's strengths for comprehensive analysis and processing.

When to Use

**Use this skill when you need:**

• Alternative AI perspective on Claude's analysis
• Parallel analysis from both Claude and Gemini
• Gemini's specific strengths for particular tasks
• Headless (non-interactive) AI execution saving to files
• Batch processing of multiple inputs
• Independent validation of Claude's work

**Don't use when:**

• Claude can handle the task directly without needing a second perspective
• Interactive conversation is required (gemini CLI is headless only)

Quick Start

Basic Pattern

cat input.md | gemini "PROMPT" --model gemini-2.5-pro > output.md

Example: