Merge pull request #1960 from Capsize-Games/develop #148

	name: Eval Tests

	on:
	pull_request:
	branches: [ master, llm-refactor ]
	push:
	branches: [ master, llm-refactor ]
	workflow_dispatch:

	permissions:
	contents: read

	jobs:
	eval-tests:
	runs-on: ubuntu-latest

	strategy:
	matrix:
	python-version: ["3.13"]

	steps:
	- name: Checkout code
	uses: actions/checkout@v4

	- name: Set up Python ${{ matrix.python-version }}
	uses: actions/setup-python@v5
	with:
	python-version: ${{ matrix.python-version }}

	- name: Install system dependencies
	run: \|
	sudo apt-get update
	sudo apt-get install -y \
	libgl1-mesa-glx \
	libglib2.0-0 \
	libsm6 \
	libxext6 \
	libxrender-dev

	- name: Install Python dependencies
	run: \|
	python -m pip install --upgrade pip
	pip install -e .
	pip install pytest pytest-timeout

	- name: Run eval tests
	run: \|
	pytest src/airunner/components/eval/tests/ \
	-v \
	-m eval \
	--timeout=300 \
	--tb=short
	env:
	AIRUNNER_HEADLESS: "1"
	AIRUNNER_HTTP_PORT: "8188"
	AIRUNNER_HTTP_HOST: "127.0.0.1"

	- name: Upload test results
	if: always()
	uses: actions/upload-artifact@v4
	with:
	name: eval-test-results-${{ matrix.python-version }}
	path: \|
	pytest-report.xml
	pytest-coverage.txt

Provide feedback