vrtx

A modern CLI tool for interacting with multiple LLMs using LangChain.

Features

• Unified interface for multiple LLM providers (Google Gemini, OpenAI, Anthropic Claude)
• Clean architecture following SOLID principles
• Conversation history management
• Shell command debugging
• Easy model switching

Installation

pip install vrtx

Gemini works out of the box. OpenAI and Anthropic are optional extras, so their packages are only downloaded if you ask for them:

pip install "vrtx[openai]"     # OpenAI models
pip install "vrtx[anthropic]"  # Anthropic Claude models

To work on Vertex-CLI itself, install from a clone instead:

git clone https://github.com/prathamhole14/vrtx.git
cd vrtx
pip install -e ".[openai,anthropic]"

Quick Start

# Initialize configuration
vrtx --setup

# Configure a model (provider auto-detected from model name)
vrtx config gemini-3.6-flash YOUR_API_KEY
vrtx config gpt-4 YOUR_OPENAI_KEY
vrtx config claude-opus-5 YOUR_ANTHROPIC_KEY

# Or specify provider explicitly
vrtx config my-model YOUR_KEY --provider google

# Select active model
vrtx select gemini-3.6-flash

# Use it
vrtx "explain quantum computing"

Commands

vrtx "your question"                  # Ask a question
vrtx chat <text>                      # Chat mode
vrtx debug                            # Debug shell commands
vrtx config <model> <api_key>         # Configure model (auto-detects provider)
vrtx config <model> <key> --provider <provider>  # Specify provider explicitly
vrtx list                             # List models
vrtx select <model>                   # Select active model
vrtx remove <model>                   # Remove model

Supported Providers

• Google Gemini: google or gemini — installed by default
• OpenAI: openai — needs pip install 'vrtx[openai]'
• Anthropic: anthropic or claude — needs pip install 'vrtx[anthropic]'

Configuration

Models are configured in ~/.config/ai_model_manager/models_config.json

Example configuration:

{
  "selected_model": "gemini-flash-latest",
  "models": {
    "gemini-flash-latest": {
      "provider": "google",
      "api_key": "your-api-key",
      "temperature": 0.7
    }
  }
}

Documentation

• Architecture - Technical architecture

Requirements

• Python 3.10+
• LangChain and provider packages

Usage

Once installed and configured, you can start chatting or debugging commands:

Chat with the LLM

You can either use the chat subcommand or omit it entirely:

# Explicit subcommand
vrtx chat "Tell me about the solar system"

# Shortcut form (no subcommand)
vrtx "Tell me about the solar system"

Replace the quoted string with any query you'd like.

🔗 Complete CLI Documentation: CLI Commands

Debugging Mode (Beta Feature)

Debugging is currently in beta but can analyze recent shell commands to identify issues.

Debug the Last 3 Commands (default)

vrtx debug

Specify the Number of Commands to Debug

vrtx debug -n 5

Add a Custom Debugging Message

vrtx debug -n 5 -p "Explain why \`git commit\` failed"

Contributing

Contributions are welcome! Follow these steps to contribute:

1. Fork the repository

2. Create a new branch:

   git checkout -b feature/your-feature-name
   

3. Commit your changes:

   git commit -m "Add your feature description"
   

4. Push your branch:

   git push origin feature/your-feature-name
   

5. Open a pull request

🔗 Contributor Guide: How to Contribute

License

This project is licensed under the MIT License. See the LICENSE file for more details.

Support

If you encounter any issues, open an issue on the GitHub repository.