mito_ai

AI chat for JupyterLab. This codebase contains two main components:

1. A Jupyter server extension that handles the backend logic for the chat.
2. Several JupyterLab extensions that handle the frontend logic for interacting with the AI, including the chat sidebar and the error message rendermime.

Requirements

• JupyterLab >= 4.0.0

Install

To install the extension, execute:

pip install mito-ai

Configuration

This extension has two AI providers; OpenAI and Mito (calling OpenAI). Mito is the fallback but the number of request is limited for free tier. To use OpenAI directly, you will to create an API key on https://platform.openai.com/docs/overview. Then set the environment variable OPENAI_API_KEY with that key.

The OpenAI model can be configured with 1 parameters:

• OpenAIProvider.model: Name of the AI model; default gpt-4o-mini.

You can set those parameters through command line when starting JupyterLab; e.g.

jupyter lab --OpenAIProvider.max_completion_tokens 20 --OpenAIProvider.temperature 1.5

If a value is incorrect, an error message will be displayed in the terminal logs.

Uninstall

To remove the extension, execute:

pip uninstall mito-ai

Contributing

Development install

To ensure consistent package management, please use jlpm instead of npm for this project.

Note: You will need NodeJS to build the extension package.

The jlpm command is JupyterLab's pinned version of yarn that is installed with JupyterLab.

# Clone the repo to your local environment
# Change directory to the mito-ai directory

# Required to deal with Yarn 3 workspace rules
touch yarn.lock

# Install package in development mode
pip install -e ".[test, deploy]"

# Install the node modules
jlpm install

# Build the extension
jlpm build

# Link your development version of the extension with JupyterLab
jupyter labextension develop . --overwrite

# Start the jupyter server extension for development
jupyter server extension enable --py mito_ai

# Watch the source directory in one terminal, automatically rebuilding when needed
# In case of Error: If this command fails because the lib directory was not created (the error will say something like
# unable to find main entry point) then run `jlpm run clean:lib` first to get rid of the old buildcache 
# that might be preventing a new lib directory from getting created. 
jlpm watch

Then, in a new terminal, run:

# Run JupyterLab in another terminal
jupyter lab --autoreload

With the watch command running, every saved change will immediately be built locally and available in your running JupyterLab. With the --autoreload flag, you don't need to refresh JupyterLab to load the change in your browser. It will launch a new window each time you save a change to the backend.

By default, the jlpm build command generates the source maps for this extension to make it easier to debug using the browser dev tools. To also generate source maps for the JupyterLab core extensions, you can run the following command:

jupyter lab build --minimize=False

Development uninstall

pip uninstall mito-ai

In development mode, you will also need to remove the symlink created by jupyter labextension develop command. To find its location, you can run jupyter labextension list to figure out where the labextensions folder is located. Then you can remove the symlink named mito-ai within that folder.

Testing the extension

Frontend tests

• Frontend tests for mito-ai are written using Playwright and Gelata in the mito/tests directory.
• Backend tests for mito-ai are written using pytest in the mito/tests directory.

To run the pytests, just run pytest in the mito-ai directory.