Gemini CLI

An interactive AI assistant powered by the open-source Gemini PyTorch implementation. Gemini CLI provides a feature-rich command-line interface with multimodal capabilities, tool integration, and advanced conversation management.

✨ Features

🤖 Interactive AI Chat: Engage with the Gemini model through an intuitive CLI
🔧 Tool Integration: Built-in tools for file operations, shell commands, web search, and more
💾 Persistent Memory: Conversation history and context management across sessions
🎨 Customizable Themes: Multiple color themes and UI customization options
🔒 Sandbox Execution: Safe execution environment for code and shell commands
📁 File Discovery: Intelligent file inclusion with pattern matching and gitignore support
🔄 Session Management: Save, resume, and manage conversation sessions
⚡ Performance Tracking: Built-in statistics and performance monitoring
🧩 Extension System: Modular architecture with extension support
🌐 Multi-modal Support: Text, image, audio, and video processing capabilities

📋 Prerequisites

Python: 3.8 or higher
Operating System: Windows, macOS, or Linux
Memory: At least 4GB RAM (8GB+ recommended for larger models)
Storage: 2GB+ free space for model weights and cache

Optional Dependencies

Docker: For sandbox execution (recommended)
Git: For version control integration
CUDA: For GPU acceleration (if available)

🚀 Installation

Method 1: Using pip (Recommended)

# Install from PyPI (coming soon)
pip install gemini-cli

# Or install with all features
pip install gemini-cli[all]

Method 2: From Source

# Clone the repository
git clone https://github.com/kyegomez/Gemini.git
cd Gemini

# Create a virtual environment (recommended)
python -m venv gemini-env
source gemini-env/bin/activate  # On Windows: gemini-env\Scripts\activate

# Install dependencies
pip install -r requirements.txt

# Install the package
pip install -e .

Method 3: Development Installation

# Clone and install for development
git clone https://github.com/kyegomez/Gemini.git
cd Gemini

# Create virtual environment
python -m venv gemini-dev
source gemini-dev/bin/activate

# Install development dependencies
pip install -r requirements-dev.txt

# Install in development mode
pip install -e .[dev]

Quick Setup Script

For a one-line installation:

curl -sSL https://raw.githubusercontent.com/kyegomez/Gemini/main/install.sh | bash

⚙️ Configuration

First Run Setup

# Initialize configuration
gemini-cli --init

# Or start with interactive setup
gemini-cli

Configuration File

Gemini CLI looks for configuration files in the following order:

./gemini/settings.json (project-specific)
~/.gemini/settings.json (user-specific)
/etc/gemini-cli/settings.json (system-wide)

Basic Configuration

Create ~/.gemini/settings.json:

{
  "model": "gemini-torch",
  "maxTokens": 4096,
  "temperature": 0.7,
  "theme": "Default",
  "autoAccept": false,
  "sandbox": {
    "enabled": true,
    "type": "docker"
  },
  "memory": {
    "enabled": true,
    "maxMemoryFiles": 50
  }
}

Environment Variables

# Model configuration
export GEMINI_MODEL="gemini-torch"
export GEMINI_MAX_TOKENS=4096
export GEMINI_TEMPERATURE=0.7

# Feature toggles
export GEMINI_AUTO_ACCEPT=false
export GEMINI_SANDBOX=true
export GEMINI_DEBUG=false

# Paths
export GEMINI_CONFIG_PATH="~/.gemini/settings.json"

🎯 Usage

Basic Commands

# Start interactive mode
gemini-cli

# Single prompt mode
gemini-cli "Hello, how can you help me today?"

# Include files in context
gemini-cli "@README.md Explain this project"

# Execute with specific model
gemini-cli --model gemini-large "Analyze this code: @src/main.py"

Interactive Mode Commands

Once in interactive mode, you can use these commands:

Slash Commands

/help - Show available commands
/quit - Exit the CLI
/clear - Clear the screen
/stats - Show session statistics
/theme [name] - Change color theme
/tools - List available tools
/memory show - Display memory context
/chat save [name] - Save current conversation
/restore [id] - Restore from checkpoint

File Commands

@file.py - Include file content in prompt
@src/*.py - Include multiple files with glob patterns
@docs/ - Include all files in directory

Shell Commands

!ls -la - Execute shell command
! - Toggle shell mode

Advanced Usage

Custom Configuration

# Use custom config file
gemini-cli --config my-config.json

# Override specific settings
gemini-cli --temperature 0.9 --max-tokens 8192

# Enable debug mode
gemini-cli --debug --verbose

Sandbox Mode

# Enable sandbox for safe execution
gemini-cli --sandbox

# Use specific sandbox image
gemini-cli --sandbox --config '{"sandbox": {"image": "custom-image"}}'

Memory Management

# Disable memory loading
gemini-cli --no-memory

# Add persistent memory
gemini-cli
> /memory add "I prefer Python over JavaScript"

🔧 Tools

Gemini CLI includes a rich set of built-in tools:

File Operations

read_file - Read file contents
write_file - Create or modify files
list_directory - Browse directories
glob - Find files with patterns
search_file_content - Search text in files

Code Execution

run_shell_command - Execute shell commands
run_python_code - Execute Python code
run_javascript_code - Execute JavaScript code

Web Operations

web_fetch - Fetch content from URLs
google_web_search - Search the web

Utility Tools

save_memory - Save information to memory
compress_conversation - Compress chat history

Tool Usage Examples

# In interactive mode
> Can you read the contents of package.json?
# Gemini will use the read_file tool automatically

> Please create a Python script that prints "Hello World"
# Gemini will use write_file to create the script

> Search for TODO comments in my Python files
# Gemini will use search_file_content tool

🎨 Customization

Themes

Available themes:

Default - Standard color scheme
Dark - Dark mode optimized
Light - Light mode optimized
Monokai - Developer-friendly
GitHub - GitHub-inspired

# Change theme
gemini-cli --theme Dark

# List available themes
gemini-cli --list-themes

Custom Themes

Create ~/.gemini/themes/mytheme.json:

{
  "name": "MyTheme",
  "description": "My custom theme",
  "colors": {
    "primary": "BLUE",
    "secondary": "CYAN",
    "success": "GREEN",
    "warning": "YELLOW",
    "error": "RED",
    "info": "BLUE",
    "prompt": "GREEN",
    "user_input": "WHITE",
    "ai_response": "CYAN",
    "tool_call": "MAGENTA",
    "system": "DIM",
    "accent": "BRIGHT_BLUE"
  }
}

Extensions

Create custom extensions in ~/.gemini/extensions/:

# ~/.gemini/extensions/my-extension/extension.py
from gemini_cli.core.extensions import BaseExtension

class MyExtension(BaseExtension):
    def __init__(self, name: str, version: str = "1.0.0"):
        super().__init__(name, version)
        self.description = "My custom extension"

    async def initialize(self, cli_context):
        return True

    async def shutdown(self):
        return True

    def get_commands(self):
        return {
            "hello": self._hello_command
        }

    async def _hello_command(self, args):
        return f"Hello from {self.name}!"

📊 Monitoring & Statistics

Session Statistics

# View current session stats
> /stats

# Export session data
gemini-cli --export-stats session.json

Performance Monitoring

# Enable verbose logging
gemini-cli --verbose --log-level DEBUG

# Monitor memory usage
gemini-cli --debug
> /stats memory

Health Checks

# Run system health check
gemini-cli --health-check

# Test tool functionality
gemini-cli --test-tools

🔒 Security & Sandboxing

Sandbox Configuration

{
  "sandbox": {
    "enabled": true,
    "type": "docker",
    "image": "gemini-cli-sandbox",
    "memoryLimit": "512m",
    "cpuLimit": "0.5",
    "allowNetwork": false,
    "readOnlyFilesystem": true
  }
}

Safe Execution

# Enable automatic sandboxing
gemini-cli --sandbox

# Review commands before execution
gemini-cli --no-auto-accept

🛠️ Troubleshooting

Common Issues

Installation Problems

Issue: Numpy installation fails with setup.py error

# Solution: Update pip and use wheel installations
pip install --upgrade pip setuptools wheel
pip install numpy --no-use-pep517

Issue: PyTorch CPU installation problems

# Solution: Install PyTorch separately first
pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cpu
pip install gemini-cli

Runtime Issues

Issue: ModuleNotFoundError for gemini_torch

# Solution: Install in development mode
pip install -e .

Issue: Permission denied for sandbox

# Solution: Add user to docker group
sudo usermod -aG docker $USER
# Then log out and back in

Configuration Issues

Issue: Config file not found

# Solution: Initialize configuration
gemini-cli --init
# Or create manually
mkdir -p ~/.gemini
echo '{}' > ~/.gemini/settings.json

Debug Mode

# Enable detailed logging
gemini-cli --debug --verbose

# Check configuration
gemini-cli --config-info

# Validate installation
gemini-cli --health-check

Getting Help

Documentation: Check the GitHub Wiki
Issues: Report bugs on GitHub Issues
Discussions: Join GitHub Discussions
Discord: Join our Discord community

🧪 Development

Setting Up Development Environment

# Clone repository
git clone https://github.com/kyegomez/Gemini.git
cd Gemini

# Create development environment
python -m venv gemini-dev
source gemini-dev/bin/activate

# Install development dependencies
pip install -r requirements-dev.txt
pip install -e .[dev]

# Install pre-commit hooks
pre-commit install

Running Tests

# Run all tests
pytest

# Run with coverage
pytest --cov=gemini_cli

# Run specific test file
pytest tests/test_cli.py

# Run in parallel
pytest -n auto

Code Style

# Format code
black gemini_cli/
isort gemini_cli/

# Lint code
flake8 gemini_cli/
pylint gemini_cli/

# Type checking
mypy gemini_cli/

Building Documentation

# Install docs dependencies
pip install -e .[docs]

# Build documentation
cd docs
make html

# Serve locally
python -m http.server 8000 -d _build/html

📝 Contributing

We welcome contributions! Please see our Contributing Guide for details.

Quick Contribution Steps

Fork the repository
Create a feature branch: git checkout -b feature/amazing-feature
Make your changes and add tests
Ensure tests pass: pytest
Format code: black . && isort .
Commit changes: git commit -m 'Add amazing feature'
Push to branch: git push origin feature/amazing-feature
Open a Pull Request

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

🙏 Acknowledgments

Built on the Gemini PyTorch implementation
Inspired by the original Gemini by Google DeepMind
Uses Rich for beautiful terminal output
Powered by PyTorch for model inference

📈 Roadmap

⭐ Star this repository if you find it helpful!

For more information, visit our GitHub repository or check out the documentation.

Name		Name	Last commit message	Last commit date
Latest commit History 154 Commits
.github		.github
bin		bin
data		data
gemini_cli		gemini_cli
gemini_torch		gemini_torch
tests		tests
tokenizer		tokenizer
.gitignore		.gitignore
LICENSE		LICENSE
README.md		README.md
agorabanner.png		agorabanner.png
example.py		example.py
gcli		gcli
gemini.png		gemini.png
gemini_block.py		gemini_block.py
install.py		install.py
install_cli.py		install_cli.py
long_gemini_example.py		long_gemini_example.py
pyproject.toml		pyproject.toml
requirements-dev.txt		requirements-dev.txt
requirements.txt		requirements.txt
setup.py		setup.py

License

gali1/Gemini_OpenSource

Folders and files

Latest commit

History

Repository files navigation