✨ Features

• Natural Language to TUI: Describe what you want, get a working terminal app instantly
• Multiple AI Providers: Anthropic Claude, OpenAI GPT, Google Gemini, and local Ollama
• App Library: Save, refine, and reuse your generated apps
• Interactive Refinement: Iterate on your apps with follow-up prompts
• Zero Config: Works out of the box if you have API keys in your environment

🎯 Perfect For

• Quick Prototypes: "Make me a file browser with search"
• System Monitoring: "Show me which ports are active"
• Data Visualization: "Create a dashboard for my server stats"
• Task Management: "Build a todo list with categories"
• Database Tools: "Diff these two SQL tables"
• Container Management: "Help me manage my Docker containers"

📦 Installation

Install Globally (Recommended)

npm install -g term-v0

Use with npx (No Installation)

npx term-v0 "create a system monitor"

Install from Source

# Clone the repository
cd term-v0
npm install
npm link  # Make it available globally

🔑 Setup API Keys

term-v0 requires an LLM provider. Choose one of the following:

Option 1: Anthropic Claude (Recommended)

Models: claude-sonnet-4-20250514, claude-3-5-sonnet-20241022, claude-3-5-haiku-20241022

# Get your API key from https://console.anthropic.com
export ANTHROPIC_API_KEY="sk-ant-..."

# Optional: Choose a specific model
export ANTHROPIC_MODEL="claude-3-5-sonnet-20241022"

Option 2: OpenAI GPT

Models: gpt-4o, gpt-4o-mini, gpt-4-turbo, o1, o1-mini, o3-mini

# Get your API key from https://platform.openai.com
export OPENAI_API_KEY="sk-..."

# Optional: Choose a specific model
export OPENAI_MODEL="gpt-4o"

# Optional: Use a custom API endpoint
export OPENAI_BASE_URL="https://..."

Option 3: Google Gemini

Models: gemini-2.0-flash-exp, gemini-1.5-pro, gemini-1.5-flash

# Get your API key from https://aistudio.google.com
export GEMINI_API_KEY="your-api-key"

# Optional: Choose a specific model
export GEMINI_MODEL="gemini-1.5-pro"

Option 4: Ollama (Local & Free)

Run models locally without API costs

# Install Ollama from https://ollama.ai
# Then pull a model:
ollama pull llama3.2

# Set the model
export OLLAMA_MODEL="llama3.2"

Setting Environment Variables Permanently

macOS/Linux:

# Add to ~/.bashrc, ~/.zshrc, or ~/.profile
echo 'export ANTHROPIC_API_KEY="your-key"' >> ~/.zshrc
source ~/.zshrc

Windows:

# Run in PowerShell as Administrator
[System.Environment]::SetEnvironmentVariable('ANTHROPIC_API_KEY','your-key','User')

🚀 Usage

Basic Usage

# Generate a new app
term-v0 "create a file explorer with preview"

# The app will be generated, saved, and launched automatically

Configure Your Provider

# Set up or change your LLM provider
term-v0 --configure

# List available providers
term-v0 --list-providers

# Override for a single run
TERM_V0_PROVIDER=google TERM_V0_MODEL=gemini-1.5-pro term-v0 "build a dashboard"

# Override via flags
term-v0 --provider openai --model gpt-4o "build a todo app"

Manage Your Apps

# List all your generated apps
term-v0 --list

# Run a previously generated app
term-v0 --run-app my-file-explorer

# Refine an existing app
term-v0 --refine my-file-explorer

Advanced Options

# Use a specific UI library (blessed or ink)
term-v0 --library blessed "create a dashboard"

# Enable iterative refinement
term-v0 --refine "make the UI more colorful"

# Control refinement iterations
term-v0 --refine-iters 3 "optimize the layout"

# Set max error fix attempts
term-v0 --fix-iters 5 "build a complex app"

This provides a visual interface for:

• Managing your app library
• Running and testing apps
• Refining existing apps
• Configuring providers

📁 Where Apps Are Saved

Your generated apps are saved in:

• macOS/Linux: ~/.term-v0/apps/
• Windows: %USERPROFILE%\.term-v0\apps\

Each app includes:

• app.js - The generated application code
• design.txt - The design document
• metadata.json - App information and prompt history

⚠️ Security Notice

term-v0 generates and executes code based on your prompts. While we implement safety measures:

• Review generated code before running in production
• Use in isolated environments when testing
• Be specific with your prompts to get expected behavior
• Don't run on systems with sensitive data without review

🤝 Contributing

Contributions are welcome! Please feel free to submit issues and pull requests.

📄 License

MIT License - see LICENSE file for details