Quantum Machine Learning (QML) models have the potential to outperform classical machine learning by increasing entanglement when their training data provide specific structural features [MBLV23; SCH+22]. However, researchers have been missing automated tooling to verify these structural features in scalable settings. This thesis extends the plugin-based platform Quantum Humanities Analysis Tool (QHAna) with a suite of seven analysis plugins that inspect quantum states supplied as vectors or OpenQASM circuits. Six classical plugins compute the Schmidt rank, detect various forms of orthogonality, and different kinds of linear dependence. A seventh, quantum plugin implements a swap-test circuit to decide orthogonality on quantum hardware or simulators. All plugins expose Representational State Transfer (REST) Application Programming Interfaces (APIs) and micro-frontends, enabling their integration into experimental setups. Evaluation using test cases confirms correct classification and shows that the analysis can be performed using either classical or quantum methods. By automating state analysis, the thesis lays the groundwork for empirically exploring how entangled datasets influence the performance of Quantum Neural Networks (QNNs) and facilitates future research on data-efficient QML.