Reorganize project structure to match openenv-course multi-mode deployment guidelines

Update stdout format to include 'score=' field and flush=True parameter on print functions to conform with strict openenv validator log spec

Update default MODEL_NAME to gpt-4.1-mini and use ValueError for missing HF_TOKEN

Implement rigorous 3-tier difficulty grader using dataset line labels

Update OpenAI client to use HF_TOKEN as api_key instead of OPENAI_API_KEY

Deploy OpenEnv Code Review to HF Spaces