Soooo...what are you looking for here? You want Feed to have it's columns implicitly populated using SQLAlchemy's reflect functionality?'

In that case, I believe AutoMap is what you want and specifically automap_base.

I've not tried this code, but it might be as simple as:

AutoMapModel = automap_base(db.Model)
class Feed(AutoMapModel):
    __tablename__ = 'autofeeds'

AutoMapModel.prepare(db.engine, reflect=True)

Let me know. If it works or you can give me the version that works, I don't see any reason why we can't add it to the docs.

We generally don't add docs for things that are documented by SQLAlchemy unless they have different behavior. Is there something specific to Flask-SQLAlchemy here? We do provide db.reflect(), which is a shortcut for calling db.Model.metadata.reflect() for each bind, but there's no added behavior.

I think the thing that is causing confusion, and may be worth documenting, is that unlike traditional SA, we "hide" the declarative_base() work in SQLAlchemy.make_declarative_base().

With pure SA, you'd swap your usage of declarative_base() with automap_base(). But since we "hide" that from the user, it's not exactly clear how you'd use FSA with automap.

We already have some documentation on the metaclass: http://flask-sqlalchemy.pocoo.org/2.3/customizing/#model-metaclass. We also document make_declarative_base as part of the public API to override: http://flask-sqlalchemy.pocoo.org/2.3/api/#flask_sqlalchemy.SQLAlchemy.make_declarative_base. Based on the SQLAlchemy docs: https://docs.sqlalchemy.org/en/13/orm/extensions/declarative/table_config.html#using-reflection-with-declarative, I think the only thing wrong with the code in the op is that it needs to set __table__ to a reflected table.