Quantum control systems with limited state access are crucial for near-term quantum devices, which often lack full wave-function information. Researchers have made progress in addressing this challenge by developing a method for disentangling qubit pairs based on reduced state observations. This approach involves a controller that receives two-qubit reduced density matrices and selects the qubit pair to disentangle at each step, using a modular hybrid policy that combines quantum and classical components1. The goal is to optimize disentanglement scheduling, which is essential for maintaining control over quantum systems. By introducing this framework, scientists can better understand how to manage quantum systems with limited information, paving the way for more efficient quantum computing and cryptography applications. This development matters to practitioners because it has significant implications for the security and reliability of quantum computing systems, potentially rewriting assumptions about computation and cryptography.