vorp 0.1.8

A terminal-based AI companion.

🧠 Vorp

🎬 Demo

Vorp is a terminal-based AI pair programmer and companion. It seamlessly indexes your codebase, allowing you to ask context-aware questions, retrieve relevant code snippets, and even modify files or execute commands without leaving your command line environment.

Note: This project is under active development.

🚀 Key Features

• Flexible Deployment: Run vorp in two modes:
  • Local Mode: Use your own API keys for direct access to LLM providers (Groq, Google Gemini).
  • Cloud Mode: Route requests through a secure proxy backend (either hosted by you or a public instance) for a frictionless experience without personal API keys.
• RAG (Chat with Codebase): Index any project folder to enable context-aware queries using local embeddings.
• Autonomous Tools: The agent can read, write, and manage files, list directories, and execute shell commands.
• Session Persistence: Chat history is saved locally, allowing you to resume sessions later.
• Context Management: Manually inject specific files into the context window for targeted assistance.
• Safety First: Critical operations like file deletion or shell execution require explicit user confirmation.
• Cross-Platform: Designed to work consistently on Windows (PowerShell/CMD), macOS, and Linux.

🛠️ Installation

Via PyPI (Recommended)

You can install Vorp directly from PyPI:

pip install vorp

From Source

1. Clone the repository:

   git clone https://github.com/SiddharthBayapureddy/vorp.git
   cd vorp
   

2. Create a virtual environment:

   • Windows: python -m venv venv and .\venv\Scripts\activate
   • macOS/Linux: python3 -m venv venv and source venv/bin/activate

3. Install in editable mode:

   pip install -e .
   

4. Configure API Keys (Local Mode): Create a .env file in the root directory:

   GROQ_API_KEY=your_groq_api_key_here
   GEMINI_API_KEY=your_gemini_api_key_here
   

💻 Usage

Start the application:

vorp

Interactive Commands

Command	Description
/index <path>	Scans and indexes the specified directory for RAG.
/rag	Toggles RAG mode on or off.
/add <file>	Loads the content of a specific file into the active chat context.
/context	Displays currently loaded files and the active RAG project.
/key	Interactive setup for your API keys.
/clear	Clears the terminal screen.
/exit-v	Exits the application and saves the current chat history.
/exit	Exits the application and deletes the current chat history.

CLI Arguments

Flag	Description
--model <id>	Starts the session with a specific model ID.
--list	Lists all supported models and their IDs.

🏗️ Architecture & Capabilities

Core System

Vorp is built using Typer for the CLI structure and Rich for beautiful terminal rendering (Markdown, Spinners, Tables). It uses LiteLLM to provide a unified interface to multiple LLM providers.

Autonomous Capabilities (Tools)

In Local Mode, Vorp provides the LLM with a set of tools to interact with your system:

• read_file: Allows the AI to examine your code.
• write_file: Enables the AI to create or update files (overwrites entire content).
• delete_file: Permanently removes files (requires confirmation).
• list_files: Lets the AI explore your directory structure.
• run_shell_command: Allows the AI to run tests, install packages, or use git (requires confirmation).

RAG (Retrieval-Augmented Generation)

The RAG system ensures the AI has a deep understanding of your specific codebase:

1. Ingestion: Files are split using a Sliding Window (1000 chars, 200 overlap).
2. Embeddings: Uses all-MiniLM-L6-v2 locally via Sentence-Transformers.
3. Storage: Vectors are stored in ChromaDB at ~/.vorp_rag_db.
4. Retrieval: Uses Cosine Similarity filtered by project_id to fetch the top 5 most relevant code snippets.

Cloud Proxy Mode

For a zero-config experience, Vorp can route requests through a FastAPI-based proxy. This proxy securely handles API keys and streams responses back to the CLI using Server-Sent Events (SSE).

⚙️ Configuration

Advanced configuration is managed via src/vorp/constants.json. You can customize:

• Models: Add or remove supported model IDs.
• Ignore Patterns: Define which files/folders RAG should skip.
• System Prompt: Modify the core instructions given to the AI.

🔮 Roadmap

• Multi-file Editing: Improving the logic for refactoring across multiple files.
• Better Diff Support: Visualizing changes before they are applied.
• Plugin System: Allow users to define custom tools.

🤝 Contributing

Contributions are welcome! Please open an issue or submit a pull request.