ai_readme_generator

AI Readme Generator reads a Github repository and suggest a README.md file based on that code using Langchain and OpenAI's GPT4.

Main functionalities of this project are listed below:

The script clones a given repo from its url or local path, and fetches file content and metadata.
The content and metadata fetched are then passed to GPT-4 model to generate the README suggestion.
It can filter out files based on their extensions.
The script can also work in debug mode. In debug mode, the script provides more detailed output logs.

Usage

Please note, in order to use this script, you will need the OPENAI_API_KEY.

To use the script from command line, please use the following template and fill in the relevant details:

pipenv run python -m ai_readme_generator.main -h

usage: main.py [-h] [-u REPO_URL] [-b BRANCH] [-e FILE_EXT_FILTER] [-m {gpt-3.5-turbo,gpt-3.5-turbo-16k,gpt-4,gpt-4-32k}]
               [-t TEMPERATURE] [-D DEBUG]

AI README Generator

options:
  -h, --help            show this help message and exit
  -u REPO_URL, --repo_url REPO_URL
                        Git repository URL or local path
  -b BRANCH, --branch BRANCH
                        Repository branch name. Default: main
  -e FILE_EXT_FILTER, --file_ext_filter FILE_EXT_FILTER
                        Filter the files based on their extensions. For instance: 'py,md', 'js,jsx' or 'php'
  -m {gpt-3.5-turbo,gpt-3.5-turbo-16k,gpt-4,gpt-4-32k}, --model {gpt-3.5-turbo,gpt-3.5-turbo-16k,gpt-4,gpt-4-32k}
                        GPT model. Default: gpt-3.5-turbo-16k
  -t TEMPERATURE, --temperature TEMPERATURE
                        Temperature for the GPT model. Default: 0.7
  -D DEBUG, --debug DEBUG
                        Show debug information

REPO_URL: repository URL for which you want to generate README. It is a mandatory parameter. This can be also a local path, so repo won't be cloned in the /tmp directory.
BRANCH: It is the branch of the repo from which you want to generate README. This is an optional parameter, the main branch will be used by default if no branch is specified.
FILE_EXT_FILTER: It is to filter the files based on their extensions. For instance: py,md, js,jsx or php. This is an optional parameter, all files will be used if no filter is given.

For instance,

pipenv run python -m ai_readme_generator.main -u https://github.com/bard/ai-assistant -b main -e py,md

Dependencies

Following Python libraries are required:

git (Repo)
langchain.document_loaders (GitLoader)
openai
os, sys
dotenv (load_dotenv)
pprint

Installation Steps

Clone the repo:

git clone https://github.com/tomkat-cr/ai_readme_generator

Change the directory:

cd ai_readme_generator

Start the virtual environment:

python3 -m venv venv
source venv/bin/activate

Install requirements:

pip install -r requirements.txt

Create the .env file:

cp .env-example .env
vi .env

Configure parameters:

OPENAI_API_KEY=sk-XXXXXX
DEBUG=0

Run it (it'll ask for git repo URL/local path and branch):

pipenv run python -m ai_readme_generator.main

or run it with repo URL/path and branch

pipenv run python -m ai_readme_generator.main -u https://github.com/tomkat-cr/ai_readme_generator -b main

Note

Please be aware that script will not attempt to delete or modify any files or directories outside of its designated working area.

Enjoy exploring the project!

Contributions

Feel free to submit pull requests create issues or spread the word.

Built with

Technologies used in the project:

python3
openai
langchain
gitpython
pipenv
make

License

This project is licensed under the GNU 2.0

yllvar/ai_readme_generator