GitHub Discussion Auto-Labeler

Automatically labels GitHub discussions using AI to analyze content and apply appropriate tags.

Features

• Automatically applies labels to new GitHub discussions
• Uses Azure OpenAI for content analysis
• Supports both GitHub token and GitHub App authentication
• Can be run locally or as a GitHub Action
• Includes a utility to create repository labels

Setup

Prerequisites

• Python 3.10+
• GitHub token or GitHub App credentials
• Azure OpenAI service (for AI-based labeling)

Installation

1. Clone this repository
2. Install dependencies:

   pip install -r requirements.txt

3. Copy .env.sample to .env and fill in your credentials

Environment Variables

Variable	Description
TOKEN	GitHub personal access token
APP_ID	GitHub App ID
APP_PRIVATE_KEY	GitHub App private key (content as string)
APP_INSTALLATION_ID	GitHub App installation ID
DEFAULT_REPO	Repository to monitor in format owner/repo
AZURE_OPENAI_ENDPOINT	Azure OpenAI API endpoint
AZURE_OPENAI_KEY	Azure OpenAI API key
AZURE_OPENAI_API_VERSION	Azure OpenAI API version
RUN_INTERVAL_MINUTES	How often to check for new discussions (in minutes)

Usage

Running Locally

To run the script locally and continuously monitor for new discussions:

python basic.py

Creating Repository Labels

To create the labels defined in tags.json in your repository:

python new-labels.py

GitHub Action

This repo includes a GitHub Action workflow that automatically runs when new discussions are created.

To set up the GitHub Action:

1. Configure repository secrets with the environment variables listed above
2. The Action will run automatically when new discussions are created

Customizing Labels

Edit the tags.json file to customize the available labels:

{
  "tags": [
    {"name": "label-name", "description": "Label description", "color": "hex-color"}
  ]
}

License

MIT