Trying to create abstraction layer to support multiple third party as document management system

I am trying to create content management library for sharepoint and in future we could deal with any CMS like google drive etc.

Right now I am using sharepoint as Document Management system hence i am trying to write an abtraction layer in such a way that no matter what cms we use our client code should not be changed.It should always work with that abstraction.

So this is the structure which i have in mind :

I will have 1 ContentManagement.Library.Common which will have abstraction like this :

public interface IContentRepositoryConnection : IDisposable
    {
        string Username { get; set; }
        string Password { get; set; }
        string ClientId { get; set; }    
        string ClientSecret { get; set; }
        void Create(); // handshake with sharepoint/google drive server using username/password or ClientId/Secret
    }

public abstract class ContentRepositoryConnection : IContentRepositoryConnection,IDisposable
    {
        public abstract string Username { get; set; }
        public abstract string Password { get; set; }
        public abstract string ClientId { get; set; }
        public abstract string ClientSecret { get; set; }

        public void Create()
        {
            throw new NotImplementedException();
        }

        public void Dispose()
        {
            throw new NotImplementedException();
        }
    }

Here is my seperate Sharepoint.library which will deal with sharepoint specific apis and sharepoint server :

public class SharepointConnection : ContentRepositoryConnection
    {
        //Where should i place SiteUrl Property which deal with sharepoint?
         //override relevant methods 
    }

Client (Console/Winform) : Will have reference of only ContentManagement.Library.Common

Client will always work with ContentRepositoryConnection class so that in future if client can easily swap between sharepoint/google drive.

Now in order to create ClientContext(just like EF Dbcontext), there are couple of parameters which sharepoints needs like SiteUrl,Username,password which might differ from Google Drive so how do i create this abstraction?

Now suppose tomorrow if i change my abstraction layer in a way that it can work with google drive and not impacting sharepoint client is possible?

This is how Client Context(Which is everything is Sharepoint) is created :

var ctx = new ClientContext(siteUrl,username,password);

I know every ORM like EF,dapper always works with abstraction(DbConnection class) instead of Concrete class(SqlConnection,OracleConnection etc..) so I am also trying to achieve same sort of abstraction in a nutshell.

Can anybody please guide me with some design changes or structure or can suggest some changes in above structure if there is any?