Classifier for open issues on a GitHub repository that uses LlamaExtract to label issues as good first issue (suitable for first time contributors).
Export the necessary environment variables:
export LLAMA_CLOUD_API_KEY="..."
export GITHUB_TOKEN="..."
export REPOSITORY_OWNER="run-llama" # or whatever repository owner
export REPOSITORY_NAME="llama_index" # or whatever repository nameOr store them in a .env file.
Install the NPM package gloablly:
npm install -g @llamaindex/issue-classifier
# verify installation
issue-classifier helpOr run with npx (or other Node-compatible executors like bunx):
npx @llamaindex/issue-classifier helpClone the GitHub repository:
git clone https://github.com/run-llama/issue-classifier
cd issue-classifierTo install dependencies:
bun installTo run:
bun run src/index.ts <LOG_LEVEL>If not provided, LOG_LEVEL defaults to info.
Use the GitHub Action to run this script in your GitHub workflows:
- name: Classify Issues as Good First Issues
uses: run-llama/issue-classifier@v0.1.0
with:
llama-cloud-api-key: ${{ secrets.LLAMA_CLOUD_API_KEY }}
github-token: ${{ secrets.GITHUB_TOKEN }}
repository-owner: "run-llama"
repository-name: "llama_index"Since this action fetches issues from the previous week, it is recommended to run it as a chron job.