oai - Ask question from bash shell and get tailored documentation response

oai is a command line tool that uses OpenAI's language model to provide a documentation-like experience to users. With oai, users can get answers to their questions and receive step-by-step guidance to complete tasks.

New in 0.7.0, run LLM locally on your computer, by using Ollama. Ollama install: https://github.com/ollama/ollama/blob/main/README.md Ollama will pull the requested LLM, like llama3:8b.

( Anyone who might be able to help me get this set up with a snap or flatpack? I've been working on trying to set up these workflows )

Quick Start

1. Clone repository `git clone https://github.com/draupner1/oai.git'
2. Change directory cd oai
3. Install dependencies pip install -r requirements.txt
4. Run python oai and enter OpenAI key as prompted to complete setup
5. Prompt with python oai or python oai "how do I..."
6. Optional: Install link in PATH, or in ~/.local/bin/
   Python Requirements

• Python 3
• OpenAI API Key
• Rich library (install via pip install rich)

Usage

To use oai, simply run the script and provide a prompt that describes the task you want to complete or the question you want to ask.

For example:

python oai.py "How to install and run a web server on Ubuntu?"

If you have added the script to your path, you can run it from anywhere:

oai "How to install and run a web server on Ubuntu?"

New locally run LLMs with Ollama.

To use Ollama as local LLM provider, Add in a in your oai command. Ex: oai oll "Tell me a joke"

Ollama integration only supports prompt:answer format today. See oai -h for details.

New openAI lib v 1.2.4

On new installs after Nov6-23 we get the new python lib v 1.2.4

This breaks some calls and response types.

OAI v.0.6.2 is adapted to the new lib!

Also, providing new shortcuts for -s size when generating Images. Use '1:1', '16:9', '9:16' as shorthand for '1024x1024' and...

Also-2, when generating images, printing the returned eventually altered prompt.

Image Generation

Available from v.0.6.0 As the API provides, creation -c is DALL-E3, Variants and Edit are DALL-E2.

-c, --create Create a new Image, DALL-E3 -w, --variant Create a variant of an Image-file. Provide "", DALL-E2 -e, --edit Edit part of an Image-file. Provide ",, Prompt-string",DALL-E2 -d DEFAULT, --default DEFAULT How many Images to create. 1-10, default=2 -s SIZE, --size SIZE Image size. default=1024x1024

Will respond with clickable link (that expires) to be opened in Browser.

Plug-ins

Available from v.0.5.0

You can now define your own Functions. (See OpenAI-chatAPI for details)

Implement them in a Class file, and drop it in the functions directory where OAI is installed.

Your functions/FuncTemplate.txt, is a class template to get you started.

Options

• If you wish, you can override automated settings by use of .env file
• You can set the OpenAI model to use by setting the OPENAI_MODEL environment variable in your .env file.
• You can disable the notice that is displayed at the end of the response by setting the OPENAI_DISABLE_NOTICE environment variable in your .env file.

Contributing

If you'd like to contribute to oai, feel free to create a pull request or open an issue. All contributions are welcome!