Claude YOLO

A wrapper for the Claude CLI that ALWAYS enables --dangerously-skip-permissions mode AND bypasses Docker and internet permission checks so that you can run in danger mode anywhere, anytime.

⚠️ SECURITY WARNING: This wrapper bypasses important safety checks! This completely bypasses the "human in the loop" checks, this could delete your data, leak your secrets and even brick your computer. Use at your own risk.

Installation

npm install -g claude-yolo

The first time you run claude-yolo, you will be presented with a consent prompt explaining the security implications. You must explicitly agree to continue.

Your consent choice is remembered for future runs.

GitHub: https://github.com/eastlondoner/claude-yolo

Usage

claude-yolo [options]

All arguments and options are passed directly to the Claude CLI.

This wrapper:

1. Checks for and automatically installs updates to the Claude package
2. Displays "🔥 YOLO MODE ACTIVATED 🔥" warning in yellow text
3. Creates a modified copy of the Claude CLI code to bypass permission checks
   • Replaces all getIsDocker() calls with true
   • Replaces all hasInternetAccess() calls with false
   • Adds colorful YOLO-themed loading messages
4. Leaves the original Claude CLI file untouched (won't affect your normal claude command)
5. Adds the --dangerously-skip-permissions flag to command line arguments
6. Imports the modified copy of the CLI

New in Version 1.6.1

• Runtime Consent Check: Now requires explicit user consent on first run
• Consent Persistence: Remembers user consent for future runs
• Fixed Global Installation: Consent prompt will properly show for both local and global installations

New in Version 1.6.0

• Installation Consent Prompt: Added explicit user consent during installation
• Enhanced Security Warnings: Clear explanations of the security implications
• Installation Abort Option: Users can cancel installation if they don't agree with the security implications

New in Version 1.5.0

• YOLO Mode Warning: Displays a "🔥 YOLO MODE ACTIVATED 🔥" warning in yellow text
• Colorful Loading Messages: Adds fun YOLO-themed loading messages with colorful text
  • "Thinking (safety's off, hold on tight)" in red
  • "Computing (all gas, no brakes, lfg)" in yellow
  • "Clauding (yolo mode engaged)" in magenta
  • "Processing (dangerous mode! I guess you can just do things)" in cyan

Features

• Auto-update: Automatically checks for and installs updates to the Claude package at runtime
• Non-destructive approach: Creates a separate modified copy of the CLI file instead of modifying the original
• Safe for global installations: Your regular claude command will work normally even after installing claude-yolo
• Debug mode: Set the DEBUG=1 environment variable to see detailed logs about the modifications

Why?

Sometimes you just want to YOLO and skip those pesky permission checks. This tool lets you do that without modifying your original Claude CLI installation.

Debugging

If you encounter any issues, you can run with debug output:

DEBUG=1 claude-yolo

This will show additional information about:

• Claude package update checks
• Current and latest available versions
• When updates are being installed
• Modifications being made to the CLI file

Auto-Update Feature

Claude YOLO automatically checks for updates to the Claude package each time it runs:

1. When you run claude-yolo, it checks for the latest version of @anthropic-ai/claude-code on npm
2. If your installed version is outdated, it will:
   • Update your package.json with the latest version
   • Run npm install to get the newest version
   • Notify you that an update was applied
3. This ensures you're always using the latest Claude CLI features

Important Security Disclaimer

This is an unofficial tool and not supported by Anthropic. Use at your own risk.

SECURITY WARNING:

• This tool bypasses safety mechanisms intentionally built into the Claude CLI
• The --dangerously-skip-permissions flag was designed for use in container environments
• By using this tool, you acknowledge that:
  • Important safety checks are being bypassed
  • Claude may access files it normally would not have permission to access
  • You accept full responsibility for any security implications

Anthropic designed these safety checks for good reason. Only use claude-yolo if you fully understand and accept these risks.