LLM creates an evaluation rubric for the case
No data yet. Run the pipeline:
uv run python scripts/run_pipeline.py --config experiments/_templates/minimal.yaml