A list of improvements: more dynamic resources than with pysmartdatamodels/model-assets

[agaldemas]

We need a bit of automation for some important things...

Here are some proposals for improvements

• Replace the file pysmartdatamodels/model-assets/official_list_data_models.json with something dynamic !
  => we should have a resource built dynamically from the repos, like a vector database, which graph the repo structure
• The process of building/updating this smardatamodels resource graph could be started through a hook to github in an n8n workflow with an agent !
• we should do the same for /pysmartdatamodels/model-assets/datamodels_metadata.json
• also with pysmartdatamodels/model-assets/smartdatamodels.jsonwhich is huge !
• still don't know how exactly to have something working, but with n8n, it should be rather easy !...

[agaldemas]

use github api to extract information from smartdatamodels github repos, and build the resources in json:

• official_list_data_models.json
• datamodels_metadata.json
• smartdatamodels.json

then insert this content into a graph database and provide an access to in pysmartdatamodels to implement functions
and in addition transform this resources into embeddings in a vector database
it might be we find a vector database with graph features !

[agaldemas]

We should for example use a neo4j database, which offers graph & vector usage features, and a connector in n8n such as : https://github.com/Kurea/n8n-nodes-neo4j !
we can do a first simple workflow, but also an agent based workflow in n8n !

but still we have to see how we can put at work the n8n workflows, somewhere on some server !!!!....
and if we use some backend llm who will provide the tokens ;O) !?!?
this world of agentic is so in motion, evolving day by day....