I think it just hit me, but I am not sure if this is right. On the

graphql yoga

server we are adding Prisma to the context. Does that just mean that we can access Prisma from within our own resolvers? In other words, does that mean that the whole Prisma schema is not exposed to the client? If it is not auto exposed, is there a way to make it partly auto exposed?

Does that just mean that we can access Prisma from within our own resolvers?

Yea.

In other words, does that mean that the whole Prisma schema is not exposed to the client?

The schema exposed from your application server is defined by yourself. using

import

. If you want to expose certain mutations, queries or subscriptions that also appear in your Prisma API, you need to add them to that schema. Example: https://github.com/graphcool/prisma/tree/master/examples/resolver-forwarding

If it is not auto exposed, is there a way to make it partly auto exposed?

See above.

Oh my gosh, thank you, that clears up soooo much