Basically MOSP is a platform for creating, editing and sharing JSON objects.
It is written in Python (Flask) and very easy to deploy (you can test on Heroku).
I have already presented MOSP here. I wanted to make a new presentation since some good progress have been made recently.
On the backend side (Python code) a lot of thins have changed. I improved the core architecture and the deployment process with Flask. Which is now super easy. The project is also now using poetry.
One of the main cool new feature is the generations of relations graphs thanks to the hosted JSON schemas. Two particularly interesting graphs are available at:
- https://objects.monarc.lu/schemas/#references_between_schemas 
These graphs are generated dynamically, thanks to JSON refs. No need to update some relations in the database. Adding a new JSON schema with some refs will automatically update the schema of relations.
The first schema,  is generated with Networkx. The display is made with d3 but you can for example use Gephi. Here is the digraph object: https://objects.monarc.lu/stats/schemas/relations.json
The next work will be to rewrite the current API with Flask-RESTX.
The advantage of MOSP is that you can deploy a dedicated instance for your specific needs. Just by defining a set of JSON schemas. For example this instance is dedicated to information security:
You can use any available JSON schemas in order to create new JSON objects via a web form. This web form is dynamically generated and based on the selected schema (and support self referring objects). It is possible to interact with MOSP programmatically thanks to its API. If you do not want to use the graphical editor.