qq 0.2.1

A CLI tool for getting quick command-line suggestions using any LLM potentially available

Quick Question (qq)

A powerful, cross-platform CLI tool that generates and executes terminal commands using 100+ LLM providers through LiteLLM integration. It intelligently prioritizes local models for privacy and falls back to cloud providers when configured.

🚀 Key Features

Universal LLM Support (100+ Providers via LiteLLM)

• Local Providers (Privacy-first, no API keys):

  • Ollama (port 11434) - Run open-source models locally
  • LM Studio (port 1234) - GUI-based local model management

• Major Cloud Providers:

  • OpenAI (GPT-4o, GPT-5, ChatGPT models)
  • Anthropic (Claude 3.5 Sonnet/Haiku/Opus)
  • Google (Gemini, PaLM)
  • Amazon Bedrock
  • Azure OpenAI
  • Groq (Fast inference)
  • Grok (xAI)

• Specialized Providers (via LiteLLM):

  • Cohere, Replicate, Hugging Face
  • Together AI, Anyscale, Perplexity
  • DeepInfra, AI21, Voyage AI
  • And 80+ more providers!

Intelligent Features

• ⚡ Smart Provider Selection: Automatically detects and uses available providers
• 🎯 Model Optimization: Selects best models based on availability and performance
• 📝 Command History: Track and replay previous commands
• 🎨 Rich Interactive UI: Beautiful terminal interface with Textual TUI
• 📋 Clipboard Integration: Copy or type commands directly
• 🔧 Developer Mode: Extensible framework for custom actions
• 🚄 Simple Mode: Streamlined one-shot command generation
• 💾 Smart Caching: 1-hour TTL for providers and models

📦 Installation

From PyPI (Stable)

pip install qq

From Test PyPI (Latest Features)

pip install --index-url https://test.pypi.org/simple/ --extra-index-url https://pypi.org/simple/ qq2

From Source (Development)

git clone https://github.com/yourusername/quickquestion.git
cd quickquestion
pip install -e .

🎯 Quick Start

Basic Usage

# Get a command suggestion
qq "find all large files over 100MB"

# Simple mode - instant command (no UI)
qq --simple "kill process on port 8080"

# Type command directly to terminal
qq --simple-type "list docker containers"

Configuration

# Interactive settings (Rich UI)
qq --settings

# Advanced configuration (Textual TUI)
qq --config

# View command history
qq --history

# Developer mode
qq --dev

⚙️ Configuration Options

Interactive Settings (qq --settings)

Navigate with arrow keys through:

1. Default Provider - Choose from available providers
2. Default Model - Select model for chosen provider
3. Command Action - Run or Copy commands
4. Simple Mode - Enable/disable streamlined mode
5. Simple Mode Action - Copy or Type behavior

Advanced Config (qq --config)

Beautiful Textual TUI with tabs:

• Quick Setup - Same as --settings but in modern UI
• Providers - Browse and configure 100+ providers
• Settings - General application settings
• About - Version and documentation

Settings are persisted in ~/.qq_settings.json

🔌 Provider Setup

Local Providers (No API Key Required)

Ollama

# Install Ollama
curl -fsSL https://ollama.com/install.sh | sh

# Pull a model
ollama pull llama2

# qq will auto-detect Ollama on port 11434
qq "your question"

LM Studio

1. Download from lmstudio.ai
2. Load any GGUF model
3. Start local server (port 1234)
4. qq auto-detects LM Studio

Cloud Providers

OpenAI

export OPENAI_API_KEY="sk-..."
qq "your question"

Anthropic

export ANTHROPIC_API_KEY="sk-ant-..."
qq "your question"

Other Providers

qq supports 100+ providers through LiteLLM. Set the appropriate environment variable:

export GROQ_API_KEY="..."
export XAI_API_KEY="..."  # For Grok
export GEMINI_API_KEY="..."
export COHERE_API_KEY="..."
# etc.

🎨 Usage Examples

Command Generation

# File operations
qq "find files modified today"
qq "compress all images in current directory"

# System management
qq "show memory usage by process"
qq "find what's using port 3000"

# Git operations
qq "undo last commit keeping changes"
qq "show commits by author in last week"

# Docker/Kubernetes
qq "remove all stopped containers"
qq "get pod logs from last hour"

Simple Mode (No UI)

# Copy to clipboard
qq --simple-copy "create python virtual environment"
# ✓ Copied: python -m venv venv

# Type to terminal
qq --simple-type "activate virtual environment"
# source venv/bin/activate [appears in terminal]

Developer Mode

qq --dev
# Access specialized developer actions and workflows

🛠️ Advanced Features

Custom Developer Actions

Create ~/QuickQuestion/CustomDevActions/my_action.py:

from quickquestion.dev_actions.base import DevAction

class MyAction(DevAction):
    @property
    def name(self) -> str:
        return "My Custom Action"
    
    @property
    def description(self) -> str:
        return "Does something special"
    
    def execute(self) -> bool:
        self.console.print("[green]Executing...[/green]")
        # Your logic here
        return True

Performance Optimizations

• Async Provider Detection: Parallel checking for fastest startup
• Smart Caching: 1-hour TTL for providers, 30-second for other data
• Lazy Loading: Deferred initialization in simple mode
• Model Prioritization: Automatic selection of optimal models

Debugging

# Enable debug output
qq --debug "your question"

# Clear provider cache
qq --clear-cache

📁 File Locations

• ~/.qq_settings.json - User preferences
• ~/.qq_history.json - Command history (last 100)
• ~/.qq_cache.json - Provider and model cache
• ~/QuickQuestion/CustomDevActions/ - Custom actions

🔧 Troubleshooting

Provider Not Detected

# Clear cache and re-detect
qq --clear-cache
qq --settings  # Reconfigure

API Key Issues

# Verify environment variable
echo $OPENAI_API_KEY

# Set in shell profile
echo 'export OPENAI_API_KEY="sk-..."' >> ~/.bashrc

SSL Certificate Errors

# macOS-specific fix
export CERT_PATH=$(python -m certifi)
export SSL_CERT_FILE="$CERT_PATH"

🚢 CI/CD & Deployment

GitHub Actions / Gitea Actions

The project includes automated workflows for:

• Testing on push/PR
• Publishing to PyPI on version tags
• Separate Test PyPI (qq2) and Production PyPI (qq) releases

Manual Deployment

# Build
python -m build

# Test locally
pip install dist/qq-*.whl

# Upload to PyPI
twine upload dist/*

📊 Architecture

quickquestion/
├── qq.py              # Main entry point and CLI
├── llm_lite_provider.py  # LiteLLM integration (100+ providers)
├── settings_manager.py   # Configuration management
├── ui_library.py      # Rich terminal UI components  
├── cache.py           # TTL-based caching system
├── provider_registry.py  # Provider catalog and metadata
├── config_app.py      # Textual TUI for configuration
└── dev_actions/       # Developer mode actions

🌟 What's New in v0.2.0

• LiteLLM Integration: Support for 100+ LLM providers
• Provider Registry: Organized catalog of all providers
• Textual TUI: Modern configuration interface (--config)
• GPT-5 Support: Compatible with latest OpenAI models
• Enhanced Caching: Improved performance and reliability
• CI/CD Pipeline: Automated testing and deployment
• Bug Fixes: Provider persistence, model selection, and more

🗺️ Roadmap

• Web UI for configuration
• Plugin system for extensions
• Multi-command workflows
• Command explanation mode
• Integration with shell history
• Homebrew formula
• Docker image
• VSCode extension

📄 License

Proprietary - All rights reserved. See LICENSE file.

💬 Support & Contact

• Bug Reports: GitHub Issues
• Feature Requests: [e-mail]
• General Support: [e-mail]
• Author: Cristian Vyhmeister ([e-mail])

🙏 Acknowledgments

• Built with LiteLLM for universal LLM support
• UI powered by Rich and Textual
• Thanks to all contributors and users!

---

Quick Question - Your AI-powered command line companion 🚀

southbrucke.com | Documentation | PyPI