Claude Code Installation Guide

Last Updated: 2026-03-20 Audience: the company Employees Platform: Linux, macOS, Windows (WSL/PowerShell)

Installation at a Glance

Quick Summary - 9 Steps

Read & Submit compliance form (the company policies)
Install Google Cloud CLI (platform-specific)
Initialize gcloud with the company email
Find your GCP Project ID (team-specific)
Authenticate with GCP (application default)
Set environment variables (VERTEX, REGION, PROJECT_ID)
Persist config to ~/.bashrc or ~/.zshrc
Install Claude Code CLI
Launch from your project directory

Total Time: ~15-20 minutes

Overview

Claude Code is a CLI tool powered by the Claude AI model that enables you to code directly within your terminal. It provides intelligent assistance for:

Building features from natural language descriptions
Debugging code and resolving issues
Navigating complex codebases
Automating repetitive development tasks

Prerequisites

Required Reading & Compliance

Mandatory Form Submission

Before installing Claude Code, you MUST submit this compliance form confirming you have reviewed the following materials:

the company Policies: - Guidelines for Responsible Use of AI Code Assistants - Legal Guidelines for Using Code Assistant Tools - the company Policy on the Use of AI Technology - Guidelines on Use of AI Generated Content

Anthropic Documentation: - Claude Code Best Practices for Agentic Coding - Anthropic's Usage Policy - Claude Code User Guide

System Requirements

Operating System: Linux (RHEL 7-10, Fedora 41-42), macOS, or Windows (WSL/PowerShell)
Python: Version 3.10 to 3.14 (included in Google Cloud SDK installer for macOS)
Internet Connection: Required for installation and authentication
the company Email: Active the company email account
GCP Access: Permission to the company's GCP projects

Check Python Version:

# Linux/macOS
python3 -V

# Should output: Python 3.10.x or higher (up to 3.14.x)

Installation Steps

Step 1: Install Google Cloud CLI

Install the Google Cloud SDK based on your operating system:

Linux (RHEL 7/8/9)

Prerequisites: - the company Enterprise Linux 7, 8, or 9 - Python 3.10 to 3.14

Add Repository:

# Add the Cloud SDK distribution URI as a package source
sudo tee -a /etc/yum.repos.d/google-cloud-sdk.repo << EOM
[google-cloud-cli]
name=Google Cloud CLI
baseurl=https://packages.cloud.google.com/yum/repos/cloud-sdk-el9-x86_64
enabled=1
gpgcheck=1
repo_gpgcheck=0
gpgkey=https://packages.cloud.google.com/yum/doc/rpm-package-key.gpg
EOM

Install Cloud SDK:

# Install required dependencies
sudo dnf install libxcrypt-compat.x86_64

# Install Google Cloud CLI
sudo dnf install google-cloud-cli

Linux (RHEL 10)

Prerequisites: - the company Enterprise Linux 10 - Python 3.10 to 3.14

Add Repository:

# Add RHEL 10 specific repository
sudo tee -a /etc/yum.repos.d/google-cloud-sdk.repo << EOM
[google-cloud-cli]
name=Google Cloud CLI
baseurl=https://packages.cloud.google.com/yum/repos/cloud-sdk-el10-x86_64
enabled=1
gpgcheck=1
repo_gpgcheck=0
gpgkey=https://packages.cloud.google.com/yum/doc/rpm-package-key-v10.gpg
EOM

Install Cloud SDK:

# Install required dependencies
sudo dnf install libxcrypt-compat.x86_64

# Install Google Cloud CLI
sudo dnf install google-cloud-cli

Fedora

Prerequisites: - Fedora 41 or 42 - Python 3.10 to 3.14

Add Repository:

# Use RHEL 9 repository (compatible with Fedora)
sudo tee -a /etc/yum.repos.d/google-cloud-sdk.repo << EOM
[google-cloud-cli]
name=Google Cloud CLI
baseurl=https://packages.cloud.google.com/yum/repos/cloud-sdk-el9-x86_64
enabled=1
gpgcheck=1
repo_gpgcheck=0
gpgkey=https://packages.cloud.google.com/yum/doc/rpm-package-key.gpg
EOM

Install Cloud SDK:

# Install required dependencies
sudo dnf install libxcrypt-compat.x86_64

# Install Google Cloud CLI
sudo dnf install google-cloud-cli

macOS (Intel x86_64)

Prerequisites: - Python 3.10 to 3.14 (check with python3 -V) - macOS with Intel processor

Download and Install:

# Download Cloud SDK for x86_64
curl -O https://dl.google.com/dl/cloudsdk/channels/rapid/downloads/google-cloud-cli-darwin-x86_64.tar.gz

# Extract archive
tar -xf google-cloud-cli-darwin-x86_64.tar.gz

# Run installation script
./google-cloud-sdk/install.sh

# Restart your shell
exec -l $SHELL

macOS (Apple Silicon ARM64)

Prerequisites: - Python 3.10 to 3.14 (check with python3 -V) - macOS with Apple Silicon (M1/M2/M3)

Download and Install:

# Download Cloud SDK for ARM64
curl -O https://dl.google.com/dl/cloudsdk/channels/rapid/downloads/google-cloud-cli-darwin-arm.tar.gz

# Extract archive
tar -xf google-cloud-cli-darwin-arm.tar.gz

# Run installation script
./google-cloud-sdk/install.sh

# Restart your shell
exec -l $SHELL

Windows (WSL)

Use the Linux (RHEL/Fedora) instructions above within Windows Subsystem for Linux.

WSL Users

If you're using WSL with Ubuntu/Debian, you may need to use apt instead of dnf. Check Google Cloud documentation for Debian-based instructions.

Step 2: Initialize Google Cloud SDK

# Initialize gcloud
gcloud init

During initialization:

Sign in:
The command opens a web browser for authorization
Sign in with your the company email account (@company.com)
Grant the requested permissions
Select Configuration:
Choose to create a new configuration or reinitialize existing
Select Project:
When prompted, enter [1] to configure a default project
You'll set the actual project in Step 3

Command-Line Only Authentication

If you prefer not to use a web browser, run:

gcloud init --console-only

Verify Basic Installation:

# List authenticated accounts
gcloud auth list

# View current configuration
gcloud config list

# Check gcloud version
gcloud version

Step 3: Find Your GCP Project ID

Finding Your Project ID

Your project ID is based on your organizational reporting line.

Steps:

Open the GCP Project ID Spreadsheet
Find the GCP name in Column C that aligns with your team
Look for the project associated with your closest upline manager
Note this project ID - you'll use it throughout the installation

Important

Replace GCP_PROJECT_ID in all following commands with your actual project ID from the opsadsheet.

Step 4: Set Up Default Authentication

# Set up application default login (may show an error - continue anyway)
gcloud auth application-default login

# Set quota project (use exactly as shown - same for everyone)
gcloud auth application-default set-quota-project cloudability-it-gemini

Note

Leave cloudability-it-gemini as-is in the second command - this is static for all the company users.

Step 5: Set Environment Variables

Replace GCP_PROJECT_ID with your actual project ID from Step 3:

export CLAUDE_CODE_USE_VERTEX=1
export CLOUD_ML_REGION=us-east5
export ANTHROPIC_VERTEX_PROJECT_ID=GCP_PROJECT_ID # Replace with your actual project ID

Step 6: Make Configuration Persistent

Make these environment variables permanent by adding them to your shell configuration:

Linux (Bash)

# Add to ~/.bashrc
echo 'export CLAUDE_CODE_USE_VERTEX=1' >> ~/.bashrc
echo 'export CLOUD_ML_REGION=us-east5' >> ~/.bashrc
echo 'export ANTHROPIC_VERTEX_PROJECT_ID=GCP_PROJECT_ID' >> ~/.bashrc # Replace GCP_PROJECT_ID

# Reload configuration
source ~/.bashrc

macOS (Zsh)

# Add to ~/.zshrc (default on macOS)
echo 'export CLAUDE_CODE_USE_VERTEX=1' >> ~/.zshrc
echo 'export CLOUD_ML_REGION=us-east5' >> ~/.zshrc
echo 'export ANTHROPIC_VERTEX_PROJECT_ID=GCP_PROJECT_ID' >> ~/.zshrc # Replace GCP_PROJECT_ID

# Reload configuration
source ~/.zshrc

Step 7: Install Claude Code

Install Claude Code using the appropriate script for your platform:

Linux / macOS / WSL

# Download installation script
curl -fsSL https://claude.ai/install.sh -o claude-install.sh

# Make executable
chmod +x claude-install.sh

# Run installer
./claude-install.sh

Windows (PowerShell)

# Download and run installation script
irm https://claude.ai/install.ps1 | Out-File claude-install.ps1
.\claude-install.ps1

Step 8: Set Working Directory

Important

Claude Code can only access files within the directory where you launch it. Navigate to your preferred working directory before launching.

# Navigate to your project directory
cd /path/to/your/project

# Or create a new directory
mkdir ~/my-claude-workspace
cd ~/my-claude-workspace

Step 9: Launch Claude Code

claude

You should see the Claude Code CLI interface with a welcome message.

Optional: Install Additional Google Cloud Components

Google Cloud SDK supports additional components that may be useful:

kubectl (Kubernetes CLI)

# Install kubectl component
gcloud components install kubectl

# Verify installation
kubectl version --client

Other Useful Components

# List all available components
gcloud components list

# Install specific components
gcloud components install app-engine-python   # App Engine for Python
gcloud components install app-engine-go     # App Engine for Go
gcloud components install cloud-sql-proxy    # Cloud SQL Proxy
gcloud components install beta          # Beta commands
gcloud components install alpha         # Alpha commands

# Update all installed components
gcloud components update

Component Installation

Installing components requires write access to the Cloud SDK installation directory. On Linux with package manager installation, you may need sudo.

Configuration

Example Shell Configuration

Your ~/.bashrc or ~/.zshrc should include these entries at the end:

# Claude Code Setup

# Google Cloud SDK paths (if not using RHEL/Fedora packages)
if [ -f "$HOME/google-cloud-sdk/path.bash.inc" ]; then
. "$HOME/google-cloud-sdk/path.bash.inc"
fi

if [ -f "$HOME/google-cloud-sdk/completion.bash.inc" ]; then
. "$HOME/google-cloud-sdk/completion.bash.inc"
fi

# Claude Code environment variables
export CLAUDE_CODE_USE_VERTEX=1
export CLOUD_ML_REGION=us-east5
export ANTHROPIC_VERTEX_PROJECT_ID=your-actual-gcp-project-id # Replace with your project ID

Verification

Check Installation Status

Run the status command to verify your setup:

claude

Inside Claude Code, run:

/status

Expected output should show:

Setting	Expected Value	Status
Provider	Google Vertex AI
GCP Project	your-project-id
Region	us-east5
Model	claude-sonnet-4-5 or claude-opus-4-6

If you see "Anthropic API" instead of "Google Vertex AI", check your environment variables in Step 5.

Initialize Project Documentation

Generate a CLAUDE.md file for your codebase:

/init

This creates project-specific documentation that helps Claude understand your codebase better.

Basic Commands

Command	Description
`/status`	Show configuration and connection status
`/init`	Generate CLAUDE.md project documentation
`/help`	Display all available commands
`/commit`	Create a git commit with AI-generated message
`/clear`	Clear conversation history
`/exit`	Exit Claude Code

Troubleshooting

Quick Diagnostic Flowchart

Claude not working?
│
├─ "command not found: claude"
│ └─> Run ./claude-install.sh again
│ └─> Check PATH in ~/.bashrc or ~/.zshrc
│
├─ "command not found: gcloud"
│ └─> Reinstall Google Cloud SDK (Step 1)
│ └─> Check PATH settings
│
├─ "Authentication failed"
│ └─> Run: gcloud auth application-default login
│ └─> Check the company email is used
│
├─ "Permission denied" or "Invalid project"
│ └─> Verify ANTHROPIC_VERTEX_PROJECT_ID is correct
│ └─> Check project ID in opsadsheet
│
├─ Provider shows "Anthropic API" not "Google Vertex AI"
│ └─> Check environment variables are set
│ └─> Run: echo $CLAUDE_CODE_USE_VERTEX (should be 1)
│ └─> Reload shell: source ~/.bashrc
│
└─ Other issues
└─> Run /status inside Claude
└─> Check #help-rh-code-assist on Slack

Issue: Authentication Errors

Error: Application default credentials not found

Solution:

# Re-run authentication
gcloud auth application-default login
gcloud auth application-default set-quota-project cloudability-it-gemini

Issue: Wrong GCP Project

Error: Permission denied or Project not found

Solution:

Verify your project ID in the opsadsheet
Update environment variable:

export ANTHROPIC_VERTEX_PROJECT_ID=correct-project-id
# Also update in ~/.bashrc or ~/.zshrc

Issue: Claude Not in PATH

Error: claude: command not found

Solution:

Restart your terminal or reload shell configuration:

# Bash
source ~/.bashrc

# Zsh (macOS)
source ~/.zshrc

Issue: Provider Shows "Anthropic API" Instead of "Google Vertex AI"

Error: Using wrong provider

Solution:

Check environment variables are set:

echo $CLAUDE_CODE_USE_VERTEX # Should output: 1
echo $CLOUD_ML_REGION     # Should output: us-east5
echo $ANTHROPIC_VERTEX_PROJECT_ID # Should output: your-project-id

If not set, re-run Step 6 and reload your shell.

Issue: Unable to Access Files

Error: Permission denied or File not found

Solution:

Claude Code can only access files in the directory where you launched it. Launch from your project root:

cd /path/to/your/project
claude

Issue: Python Version Incompatible

Error: Python 3.x.x is not supported

Solution:

Google Cloud SDK requires Python 3.10 to 3.14:

# Check version
python3 -V

# If too old, install newer Python (RHEL/Fedora)
sudo dnf install python3.11

# macOS (using Homeservice-c)
service-c install python@3.11

Issue: Repository GPG Key Error

Error: GPG check FAILED

Solution:

# Re-import GPG key
sudo rpm --import https://packages.cloud.google.com/yum/doc/rpm-package-key.gpg

# RHEL 10 uses different key
sudo rpm --import https://packages.cloud.google.com/yum/doc/rpm-package-key-v10.gpg

Issue: gcloud Command Not Found After Installation

Error: bash: gcloud: command not found

Solution:

For macOS/manual installation:

# Add to PATH (if using manual installation)
export PATH=$PATH:$HOME/google-cloud-sdk/bin

# Add to shell config permanently
echo 'export PATH=$PATH:$HOME/google-cloud-sdk/bin' >> ~/.bashrc
source ~/.bashrc

For Linux package installation:

# Package should install to /usr/bin/gcloud automatically
# If not found, reinstall
sudo dnf reinstall google-cloud-cli

Getting Help

the company Internal Support

If you encounter issues:

Slack Channel: #help-rh-code-assist
Documentation: Check the Claude Code User Guide
GCP Issues: Contact your team's GCP administrator

Useful Resources

the company AI Hub: source.company.internal/projects_and_programs/ai
Anthropic Best Practices: Claude Code Best Practices
Google Cloud SDK: cloud.google.com/sdk/docs

Quick Reference

Essential Commands (Inside Claude)

Command	Description	Example Usage
`/status`	Check configuration and connection	Verify GCP project and provider
`/init`	Initialize project documentation	Generate CLAUDE.md for codebase
`/commit`	Create git commit with AI message	Auto-generate commit from changes
`/help`	Show all available commands	List all / commands
`/clear`	Clear conversation history	Start fresh conversation
`/exit`	Exit Claude Code	Close CLI session
`/model`	Change AI model	Switch between Sonnet/Opus
`/config`	View/edit configuration	Change settings

Environment Variables

# Required for the company Claude Code via Vertex AI
export CLAUDE_CODE_USE_VERTEX=1
export CLOUD_ML_REGION=us-east5
export ANTHROPIC_VERTEX_PROJECT_ID=your-gcp-project-id

# Optional: Set default model
export CLAUDE_MODEL=claude-sonnet-4-5

Launch Claude

# Navigate to project directory
cd /path/to/your/project

# Launch Claude Code
claude

# Or specify a different directory
claude --directory=/path/to/project

Verify Installation Checklist

# 1. Check gcloud is installed
gcloud version

# 2. Check authentication
gcloud auth list

# 3. Check project is set
gcloud config get-value project

# 4. Check environment variables
echo $CLAUDE_CODE_USE_VERTEX      # Should be: 1
echo $CLOUD_ML_REGION         # Should be: us-east5
echo $ANTHROPIC_VERTEX_PROJECT_ID   # Should be: your-project-id

# 5. Launch Claude
claude
# Then inside Claude: /status

Common File Locations

File	Purpose	Location
gcloud config	Google Cloud SDK settings	`~/.config/gcloud/`
Shell config	Environment variables	`~/.bashrc` or `~/.zshrc`
Claude cache	Conversation history	`~/.claude/`
Project docs	Generated by /init	`CLAUDE.md` in project root

Advanced Topics

Multiple GCP Projects

If you work with multiple GCP projects:

# List all configurations
gcloud config configurations list

# Create new configuration for different project
gcloud config configurations create project-two
gcloud config set project OTHER-PROJECT-ID

# Switch between configurations
gcloud config configurations activate default
gcloud config configurations activate project-two

# Set project for current configuration
gcloud config set project YOUR-PROJECT-ID

Using Claude Code in Different Directories

Claude Code creates a separate conversation context per directory:

# Project 1
cd ~/projects/frontend
claude # Separate context

# Project 2 (different terminal)
cd ~/projects/backend
claude # Different context

IDE Integration

Run Claude Code in your IDE's integrated terminal for seamless workflow:

VS Code: 1. Open integrated terminal: Ctrl+` 2. Navigate to project root 3. Run claude

JetBrains IDEs (IntelliJ, PyCharm): 1. Open terminal: Alt+F12 (Linux) or ⌥F12 (Mac) 2. Navigate to project root 3. Run claude

Accessibility Features

Enable screen reader support:

gcloud config set accessibility/screen_reader true

Offline Documentation

Access gcloud help without internet:

# Get command help
gcloud help

# Get help for specific command
gcloud help auth

# List all commands
gcloud help --help

Happy Coding with Claude!

For questions or issues, reach out on #help-rh-code-assist.

Claude Code Installation Guide

Table of Contents

Installation at a Glance

Overview

Prerequisites

Required Reading & Compliance

System Requirements

Installation Steps

Step 1: Install Google Cloud CLI

Step 2: Initialize Google Cloud SDK

Step 3: Find Your GCP Project ID

Step 4: Set Up Default Authentication

Step 5: Set Environment Variables

Step 6: Make Configuration Persistent

Step 7: Install Claude Code

Step 8: Set Working Directory

Step 9: Launch Claude Code

Optional: Install Additional Google Cloud Components

kubectl (Kubernetes CLI)

Other Useful Components

Configuration

Example Shell Configuration

Verification

Check Installation Status

Initialize Project Documentation

Basic Commands

Troubleshooting

Quick Diagnostic Flowchart

Issue: Authentication Errors

Issue: Wrong GCP Project

Issue: Claude Not in PATH

Issue: Provider Shows "Anthropic API" Instead of "Google Vertex AI"

Issue: Unable to Access Files

Issue: Python Version Incompatible

Issue: Repository GPG Key Error

Issue: gcloud Command Not Found After Installation

Getting Help

the company Internal Support

Useful Resources

Quick Reference

Essential Commands (Inside Claude)

Environment Variables

Launch Claude

Verify Installation Checklist

Common File Locations

Advanced Topics

Multiple GCP Projects

Using Claude Code in Different Directories

IDE Integration

Accessibility Features

Offline Documentation