Chapyter is a JupyterLab extension that bridges GPT-4 with your coding environment, enabling natural language programming by translating descriptions into executable Python code. It's designed for developers and researchers seeking to boost productivity and explore new ideas within their familiar Jupyter workflow.

How It Works

Chapyter leverages an IPython magic command (%%chat) to interpret natural language prompts within a Jupyter cell. It then calls OpenAI's GPT models to generate Python code, which is automatically executed. The extension supports using coding history and execution output for context-aware code generation, facilitating tasks like data visualization. It also offers in-situ debugging and customization of prompts and AI configurations.

Quick Start & Requirements

• Install via pip: pip install chapyter
• Requires JupyterLab ≥ 4.0.
• Requires OpenAI API key and organization set as environment variables.
• Official examples are available at examples/01-quick-start.ipynb.

Highlighted Details

• Seamlessly integrates GPT-4 into JupyterLab for natural language programming.
• Supports code generation from natural language with automatic execution.
• Allows using coding history and execution output for context-aware code generation.
• Enables in-situ debugging and code editing within the Jupyter environment.
• Offers transparency and customization of prompts and AI configurations.
• Emphasizes privacy by adhering to OpenAI API data usage policies.

Maintenance & Community

• Development notes indicate a two-component architecture: IPython magic command and frontend execution listener.
• Requires NodeJS for building the extension package.
• Development setup involves cloning the repo, installing in development mode (pip install -e "."), and linking the extension (jupyter labextension develop . --overwrite).
• A jlpm watch command is available for continuous rebuilding during development.

Licensing & Compatibility

• The README does not explicitly state a license.

Limitations & Caveats

• Installation may upgrade JupyterLab to ≥ 4.0, potentially breaking existing environments.
• The README does not specify a license, which could impact commercial use or closed-source integration.