Solves the import guessing game for Python AI and ML libraries. Instead of digging through docs to figure out whether it's `from sklearn.ensemble import RandomForestClassifier` or `from sklearn.model_selection import train_test_split`, you get correct import statements, working quickstart snippets, and warnings about common pitfalls. Covers the ecosystem of libraries you'd actually use for machine learning work: scikit-learn, PyTorch, TensorFlow, transformers, and friends. Runs over SSE at mcp.checklist.day. The source repo is oddly named ray-tracer, but the service itself focuses on Python package infrastructure. Reach for this when you're prototyping with unfamiliar libraries and don't want to context switch to Stack Overflow every three minutes.