neat

neat is a ReAct-like agent accessing all sorts of data & APIs

Getting started

This project uses Poetry for dependency management. If you haven't installed Poetry yet, you can do so by following the instructions on the official website.

Sidenote: poetry >>> conda & pip.

Once you have Poetry installed, you have 2 options.

Include it in your dependencies

[tool.poetry.dependencies]
neat-ai-assistant = {git = "https://github.com/NickyHavoc/neat-demo.git", branch="main"}

Needless to say, you need poetry set up in your environment to do this.

Optionally, you can include some extra dependencies:

neat-ai-assistant = {git = "https://github.com/NickyHavoc/neat-demo.git", branch="main", extras = ["tool-extension"]}

Clone the repo

git clone https://github.com/NickyHavoc/neat-demo.git
cd neat-demo
poetry install

If you want to use some extra tools that require special dependencies you can install them like so:

poetry install -E tool-extension

Create your .env file

In order to set credentials to the api that should never be pushed to a repository, copy the file ".env_example", rename the copy to ".env" and paste your personal API tokens. At least, you must include your "OPENAI_API_KEY" here.

Launching the server

Backend

This guide assumes that you have already installed FastAPI and an ASGI server like Uvicorn, which is required to run a FastAPI application. If you haven't installed these yet, you can do so using poetry:

poetry add fastapi uvicorn

Navigate to Your Project Directory: Open a terminal or command prompt and navigate to the directory where your main.py file is located.

uvicorn main:app --reload

Once the server is running, you should see output in the terminal indicating that it's live. By default, the server will be available at http://127.0.0.1:8000. You can use a tool like curl to interact with the API.

API Documentation: FastAPI automatically generates interactive API documentation for your server. You can visit it at http://127.0.0.1:8000/docs. This page will allow you to see all your endpoints and test them directly from the browser.

Frontend

Simply navigate to the react-ui directory and start npm, like so:

cd react-ui
npm start

To learn more, check out the frontend-readme.

Run tests locally

To run tests use either of the following commands

pytest # this by default runs all tests in tests/
pytest -v # run all tests verbose (see which test is currently running)
pytest -s # run all tests and print whatever the code would print
pytest -x # run all tests and stop after first failure
pytest -vsx # all of the above

License

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