GemCLI - Consumer Gemini AI for Code Completions, System Automation, and Image Generation

A professional Python command-line interface that leverages Google’s consumer Gemini AI for unlimited code completions, autonomous file operations, system automation, and AI image generation - all directly from your terminal.

GemCLI provides a high-fidelity terminal experience with real-time markdown rendering, customizable themes, and browser-based authentication.

---

PRIVACY & SECURITY This is a CLIENT-SIDE application with NO SERVER component. Your data remains on YOUR machine. All processing is local. Session tokens never leave your system.

---

Key Features

• Modern Terminal Interface: Built with Rich library for beautiful, responsive command-line experience
• Four Operating Modes: Chat, Semi-Agent (AI Coding Assistant), Agent (Autonomous), and Image Generation
• Browser Cookie Authentication: Automatic session extraction from Chrome, Edge, or Firefox - no API key needed
• File Operations: Read, edit, search, and create files through AI commands in Agent/Semi-Agent modes
• System Command Execution: Control system operations - open applications, adjust brightness/volume, launch file explorer, manage media playback
• Git Integration: Automatic commit, push, and AI-generated commit messages for version control
• Diff Viewer: Preview code changes before applying with VS Code or terminal-based diff view
• Six Theme Profiles: Customize with Cyan, Pink, Gold, Green, Purple, or White color schemes
• Asynchronous Architecture: Non-blocking API communication built on asyncio
• Complete Privacy: Client-side only - your session tokens never leave your local machine

---

Modes & Capabilities

GemCLI offers four distinct operating modes, each tailored for specific use cases:

Mode	Ask/Response	Read Files	Edit Files	Workspace Search	Autonomous	System Cmds
Chat	✓	✗	✗	✗	✗	✗
Semi-Agent	✓	✓	✓	✗	✗	✓
Agent	✓	✓	✓	✓	✓	✓
Image Gen	✓	✗	✗	✗	✗	✗

Mode Descriptions

Chat Mode

Basic conversational AI interface. Ask questions, receive answers, and engage in natural conversations with Gemini.

Semi-Agent Mode (AI Coding Assistant)

Enhanced mode where you specify files to work with. Gemini can:

• Read files you specify

• Suggest code modifications

• Apply changes with your approval

• Show diffs before committing

• Integrate with git workflow

• Execute system commands:

• Open/close applications (Chrome, Notepad, Calculator, etc.)

• Adjust brightness and volume

• Launch file explorer

• Control media playback

• System shutdown (when explicitly requested)

Agent Mode (Autonomous)

Fully autonomous coding assistant. Gemini can:

• Search your entire workspace for relevant files

• Read any files it determines necessary

• Make coordinated changes across multiple files

• Work independently without file specifications

• Handle complex multi-file refactoring

• Execute system commands:

• Open/close applications

• Adjust brightness and volume

• Launch file explorer

• Control media playback

• System shutdown (when explicitly requested)

Image Generation Mode

Create AI-generated images from text descriptions with customizable save locations.

---

Getting Started

Prerequisites

• Python 3.8 or higher installed on your system
• Active Gemini account - Must be logged into gemini.google.com in Chrome, Edge, or Firefox
• Internet connection for API communication

Installation

Option 1: Install from PyPI (Quick Start)

Standard pip installation:

pip install gemcli

After installation, use one of these commands to start:

# If PATH is configured:
gemcli

# If gemcli command not found (use this if above fails):
python -m gemini_cli

Option 2: Install with pipx (Recommended - Automatic PATH Setup)

pipx automatically handles PATH configuration so gemcli works immediately:

# Install pipx if you don't have it
pip install pipx
pipx ensurepath

# Install GemCLI
pipx install gemcli

Close and reopen your terminal, then start GemCLI:

gemcli

Option 3: Install from Source

Navigate to the project directory:

cd GemCLI

Set up a virtual environment (recommended):

python -m venv venv

# Activate on Windows:
venv\Scripts\activate

# Activate on macOS/Linux:
source venv/bin/activate

Install required dependencies:

pip install -r requirements.txt

Run GemCLI:

python gemini_cli.py

# OR install locally and use the gemcli command:
pip install -e .
gemcli

First-Time Setup

When you launch GemCLI for the first time:

Automatic Authentication: GemCLI will automatically detect and extract your Gemini session from your browser (Chrome, Edge, or Firefox). Ensure you are logged into gemini.google.com in one of these browsers. 1.

Mode Selection: You will be presented with a menu to select your operating mode:

• Chat - Basic conversational AI
• Semi-Agent - AI coding assistant with file operations
• Agent - Autonomous coding assistant
• Image Generation - Create AI images

Ready to Use: Start typing your queries or commands. GemCLI will respond with formatted output directly in your terminal.

---

Usage Guide

Starting GemCLI

Launch the application:

gemcli

Available Commands

Command	Description
/help	Display help with mode capabilities and commands
/exit or /quit	Safely exit the current mode
/clear	Clear the terminal screen
/mode	Switch between operating modes
/status	Show git repository status (Semi-Agent/Agent modes)
/commit	Commit changes with AI-generated message (Semi-Agent/Agent modes)
/push	Push commits to remote repository (Semi-Agent/Agent modes)

File Path Autocomplete

In Semi-Agent and Agent modes, type / to trigger workspace file path autocomplete. Example:

You: Read /gemini_cli.py

---

Settings

Access via main menu Settings option:

Theme Customization

Choose from 6 professional color schemes:

• Cyan
• Pink
• Gold
• Green
• Purple
• White

Git Integration

Configure automatic version control:

• Enable/Disable: Toggle git integration on or off
• Commit Mode: Choose immediate or on-exit commit behavior
• Auto Push: Automatically push commits after committing
• Branch Prompts: Configure branch creation and selection prompts

Diff Viewer

Preview code changes before applying:

• VS Code: Open diffs in VS Code editor
• System Default: Use system default diff tool
• Terminal Only: Display diffs directly in terminal

Image Generation

Configure image output settings:

• Save Path: Specify directory for generated images
• Filename Pattern: Customize image naming convention

---

Project Structure

CliTool/
├── gemini_cli.py      # Core application logic and UI
├── requirements.txt   # Python package dependencies
├── README.md          # Main documentation
├── QUICKSTART.md      # Quick reference guide
├── GIT_INTEGRATION.md # Git workflow documentation
└── INSTALL.md         # Detailed installation instructions

---

Example Workflows

Coding with Semi-Agent Mode

You: Read /src/app.py
Gemini: [Reads and analyzes file content]
You: Add error handling to the main function
Gemini: [Suggests changes with diff preview]
You: [Review and approve changes]
Gemini: [Applies modifications to file]

Autonomous Agent Mode

You: Refactor the authentication system to use JWT tokens
Gemini: [Searches workspace for relevant files]
Gemini: [Reads necessary files and plans changes]
Gemini: [Makes coordinated modifications across multiple files]
Gemini: [Displays summary of all changes made]

System Commands (Agent/Semi-Agent Mode)

You: lower the brightness
Gemini: Setting brightness to 30%
[Executes system command]
Command executed successfully

You: open chrome
Gemini: Opening Chrome browser
[Executes system command]
Command executed successfully

You: increase volume
Gemini: Increasing system volume
[Executes system command]
Command executed successfully

Available System Commands:

• Open/close applications (Chrome, Notepad, Calculator, File Explorer, etc.)
• Adjust brightness and volume levels
• Control media playback
• System shutdown (when explicitly requested)

For detailed system command documentation, refer to SYSTEM_COMMANDS.md

Image Generation

You: Create a futuristic cyberpunk cityscape at night
Gemini: [Generates image and saves to configured directory]

---

Troubleshooting

Command Not Found: ‘gemcli’ is not recognized

If gemcli shows an error after pip installation, you have three options:

Quick Fix (No configuration needed):

python -m gemini_cli

Permanent Fix (Recommended):

# Uninstall current installation
pip uninstall gemcli

# Reinstall with pipx (handles PATH automatically)
pip install pipx
pipx ensurepath
pipx install gemcli

Close and reopen terminal, then gemcli will work.

Manual PATH Setup: Add Python’s Scripts directory (usually C:\Users\YourName\AppData\Local\Programs\Python\Python3X\Scripts\) to your system PATH via System Properties > Environment Variables, then restart terminal

Authentication Issues

• Browser Lock: Ensure your browser is closed if the application cannot access the cookie database
• Login Status: Verify your session is active by visiting gemini.google.com
• Permissions: On Windows, try running your terminal as Administrator

File Operations

• Use absolute or relative paths from your current directory
• Type / to trigger autocomplete in Agent/Semi-Agent modes
• Review git status before making extensive changes

---

License & Disclaimer

License: Distributed under the MIT License. See LICENSE file for details.

Disclaimer: This is an unofficial tool for educational and personal use. It utilizes a web-based API wrapper and may be subject to changes based on Google platform updates. Use responsibly and in accordance with Google’s Terms of Service.

---

Made by 89P13

---