Cirq

This gives Claude the ability to write, simulate, and optimize quantum circuits using Google's Cirq framework. You'd reach for it when targeting Google Quantum AI hardware like Sycamore or Weber, when you need fine-grained noise modeling, or when running characterization experiments like randomized benchmarking. It covers the full stack from basic circuit construction through parameterized gates, noisy simulation, hardware compilation, and variational algorithms. The skill includes integration patterns for Google Quantum Engine, IonQ, Azure Quantum, and other providers. If you're working with IBM hardware you want qiskit instead, and if you need automatic differentiation for quantum ML, pennylane is the better fit.