TODO Assistant 📝🤖

⚡ Handle Your Notion TODO board with LLMs ⚡

Overview 👀

Notion TODO Board Assistant is an AI-powered productivity tool, which leverages the capabilities of LangChain and LangGraph. The application helps to manage your tasks on a Notion board by actively listening to your commands and executing requested tasks. It interfaces with the Notion API to interact with real Notion database.

A great way to familiarize with the LangChain library trying to make the solution as production ready as possible with a high quality codebase. 🚀

Features ✨

You can perform an array of actions through commands with TO-DO Assistant:

1. Add a new task.
2. Update existing tasks.
3. Delete tasks.
4. Retrieve specific tasks.
5. Generate tasks based on project description.

And much more!

Tech Stack 🛠

• LangChain: For LLMs flows.
• LangGraph: For complex agents definition.
• notion-client: For interfacing with Notion database.
• Streamlit: For interactive web application example.

Installation 💻

Installing and setting up the project is a breeze, all thanks to Poetry:

1. Use poetry to set up the environment and install dependencies:

   poetry install

2. Set the env variables. Create the .env file based on .env.example and fill it with appropriate values

Running The Project 🚀

Using Run Script

You can run the assistant in commandline using the provided run.py script:

python run.py

Using Streamlit App

The Streamlit application provides a more interactive experience. Run it using:

streamlit run streamlit_app.py

Example Commands 🎤

Here are a few examples of what you can do:

• "Create a task titled 'Buy Groceries'"
• "Mark task 'Buy Groceries' as Done"
• "Delete the task 'Buy Groceries'"
• "I want to learn python, define and add a few example tasks to the board."
• "What is the status of 'Buy Groceries'"

Roadmap

• Streamlit application
• Saving conversation history
• Error handling;
• Handle tasks properties like priority or work estimation;
• Question answering based on the whole database not only the specific retrieved task;