Scanning transmission electron microscopy (STEM) is a powerful tool for exploring structure, composition, and chemical properties on nanometer and atomic level and hence is a powerful tool for materials discovery. However, the STEM paradigm is based on use of classical scanning and point and grid based spectroscopies We report the development of the tool for direct control of operational STEM that enables materials discovery using imaging and spectroscopic modes, following either predefined policies of myopic discovery workflows. The latter are illustrated by human-in-the-loop automated experiments \textbf{(hAE)} with bayesian optimisation (BO), that allows dynamic control of reward functions while running the microscope. This is a new paradigm of STEM where we dynamically tune the behaviour of ML agent rather then directly controlling the instrument. While several studies have highlighted the potential of human-in-the-loop automated experiments (hAE), there remains a need for a software interface to facilitate these experiments effectively. In this work, we developed a web tool that integrates Bayesian optimization with human intervention capabilities, allowing users to modify scalarizers (also referred to as reward or target properties) and adjust experimental policies dynamically to favour exploration or exploitation. The tool also provides flexibility to manipulate descriptor sizes, thereby capturing the complex structure-property relationships at nanoscale. We have also provided a simulator version on pre-acquired data as an educational interface for experimentalist new to automated experiments. The code is provided as the supplementary material. The simulator can be accessed live at https://02b6-192-249-0-29.ngrok-free.app/