Repository navigation

add automatic issue classification #1

	name: Issue classifier validation

	on:
	pull_request:
	paths:
	- .github/issue-classifier/**
	- .github/workflows/issue-classifier.md
	- .github/workflows/issue-classifier.lock.yml
	- .github/workflows/issue-classifier-validation.yml
	- .github/workflows/shared/issue-classification.md
	workflow_dispatch:
	inputs:
	evaluate:
	description: Run the 40-case historical evaluation (uses OpenAI API credits)
	required: true
	default: false
	type: boolean

	permissions: {}

	jobs:
	test:
	name: Test classifier
	runs-on: ubuntu-latest
	permissions:
	contents: read
	steps:
	- name: Check out repository
	uses: actions/checkout@d23441a48e516b6c34aea4fa41551a30e30af803 # v6
	- name: Set up Python
	uses: actions/setup-python@5fda3b95a4ea91299a34e894583c3862153e4b97 # v7
	with:
	python-version: "3.12"
	- name: Run tests
	run: python -m unittest discover -s .github/issue-classifier -p "test_*.py"

	evaluate:
	name: Evaluate classifier
	if: github.event_name == 'workflow_dispatch' && inputs.evaluate
	needs: test
	runs-on: ubuntu-latest
	timeout-minutes: 45
	permissions:
	contents: read
	issues: read
	steps:
	- name: Check out repository
	uses: actions/checkout@d23441a48e516b6c34aea4fa41551a30e30af803 # v6
	- name: Set up Python
	uses: actions/setup-python@5fda3b95a4ea91299a34e894583c3862153e4b97 # v7
	with:
	python-version: "3.12"
	- name: Evaluate historical issues
	run: python .github/issue-classifier/classifier.py evaluate
	env:
	GITHUB_TOKEN: ${{ github.token }}
	OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}