CodeSense - AI Programming Tools & Models Platform

Title

Install via NPM

Cross-platform

Install Gemini CLI using npm package manager

# Install globally
npm install -g @google/generative-ai-cli

# Verify installation
gemini --version

Install on Linux

Linux

Linux installation with proper permissions

# Install with npm
sudo npm install -g @google/generative-ai-cli

# Or use local installation
npm install @google/generative-ai-cli
npx gemini --version

Install on macOS

macOS

macOS installation using npm

# Using npm (recommended)
npm install -g @google/generative-ai-cli

# Verify installation
which gemini && gemini --version

Download from GitHub

Source

Build from source code

# Clone repository
git clone https://github.com/google/generative-ai-cli.git
cd generative-ai-cli

# Install and build
npm install && npm run build

API Key Configuration

Set API Key

Configure your Google AI API key for authentication

# Set API key
gemini config set api-key YOUR_API_KEY_HERE

# Test authentication
gemini auth test

Environment Variables

Alternative method using environment variables

# Linux/Mac
export GEMINI_API_KEY="your-api-key-here"

# Windows PowerShell
$env:GEMINI_API_KEY="your-api-key-here"

# Verify
echo $GEMINI_API_KEY

Common Commands

Text Generation

Generation

# Basic generation
gemini generate "Write a Python function"

# With parameters
gemini generate --model gemini-pro \
  --max-tokens 500 "Explain AI concepts"

Code Review

Code Analysis

# Review a file
gemini review ./src/main.py

# Review with context
gemini review --context "security" ./app.js

Interactive Chat

Interactive

# Start interactive session
gemini chat

# Chat with file context
gemini chat --file ./README.md

Batch Processing

Batch

# Process multiple files
gemini batch --input ./docs/ \
  --output ./summaries/

# Custom prompt for batch
gemini batch --prompt "Summarize" ./files/

Advanced Usage

Custom Configuration

Advanced configuration options for power users

1# Set custom model parameters
2gemini config set model.temperature 0.8
3gemini config set model.max-tokens 2048
4
5# Configure output format
6gemini config set output.format json
7gemini config set output.directory ./ai-outputs/
8
9# Set rate limiting
10gemini config set rate.limit 30/minute

Scripting Integration

Use Gemini CLI in shell scripts and automation

1#!/bin/bash
2# Automated code review script
3
4FILES=$(git diff --name-only HEAD~1)
5for file in $FILES; do
6  if [[ $file == *.js || $file == *.py ]]; then
7    echo "Reviewing $file..."
8    gemini review "$file" --format markdown > "reviews/$file.md"
9  fi
10done
11
12echo "Code review complete!"

Troubleshooting

Installation Problems

Common installation and permission issues

# Fix npm permissions
sudo chown -R $(whoami) $(npm config get prefix)

# Clear cache and reinstall
npm cache clean --force
npm install -g @google/generative-ai-cli

# Verify installation
gemini --version

Authentication Errors

API key and authentication troubleshooting

# Test authentication
gemini auth test

# Reset and reconfigure
gemini auth reset
gemini config set api-key YOUR_API_KEY

# Check configuration
gemini config list

Command Not Found

PATH and environment setup issues

# Check if installed
npm list -g @google/generative-ai-cli

# Add to PATH
export PATH=$PATH:$(npm bin -g)

# Use npx as alternative
npx @google/generative-ai-cli --help

Rate Limiting

Handle API rate limits effectively

# Check current limits
gemini status

# Adjust rate limiting
gemini config set rate.limit 20/minute
gemini config set rate.delay 3000

# Monitor usage
gemini config get rate.current

Tips & Best Practices

Pro Tips

•Use gemini config list to view all current settings
•Set up environment variables for API keys to avoid exposing them in commands
•Use --help with any command to see available options
•Test authentication before running batch operations
•Monitor your API usage to avoid unexpected rate limits