Following initial consultation with the Kadi development team, a prototype workflow was implemented using the Python tool kadi_apy. The workflow generates a structured JSON representation of all collections and records within a root collection, thereby providing a quick overview of the data stored in Kadi. The metrics derived from these, such as the number of collections, records, and associated files, enable simple statistics and targeted data access.

A second prototype was developed during an internal hackathon: It launches local simulation runs, automatically retrieves the necessary input data from Kadi, and then transfers the generated results back into the system, again via kadi_apy. This makes the results immediately available for further analysis.

The tests show that Kadi can be used in the future not only as a data repository but also as an integral part of automated research workflows.