The answer is yes. Searching Discord bites, so impossible to find the discussions for you since the <#816286905163448340> channel has a ton of matches for some reason - and there was a reorg of channels a few months ago that might have left the most recent answers in <#615892613081333794> .

The full answer goes something like this….

• develop locally
• create a new branch staging in your git repo
• push your dev changes to the new staging branch
• create a new NHost project and point it to your staging branch, not your main branch
• fill your staging NHost project with test data, test, validate
• when ready issue a PR from staging to your main/master branch which should still be tied to the production NHost project.

super-helpful, thank you. I guess the thing that worried me more than anything where migrations, and whether having different environments would somehow lose context on what DB changes need to be applied to prod

For reference: https://docs.nhost.io/platform/multiple-environments

@mr_clinton I think I can safely say that it works as expected - changes are made in both places at the right time: when you develop locally using the CLI you get the changes there, then when you push your changes to staging you get them and can see + test them there, then finally pushing to production they apply there as well and you can smoke test for the changes and voila.

Pretty great!

@elitan Cool, hadn't seen the new(ish) page.

btw, is there a way to add seed data for auth.users? I created /seeds/auth/file.sql but doesn't appear to work

This format works for me:

DO $$ 
DECLARE
    userid auth.users.id%TYPE = '12345678-1234-1234-1234-123456789012'; 
    otherid other.id%TYPE;
BEGIN
    INSERT INTO auth.users (id, display_name, email, email_verified, locale, metadata)
     VALUES (userid, 'Fake Tester', 'test@nowhere.com', TRUE, 'en', '{"lastName":"Tester","firstName":"Fake"}')
      RETURNING id INTO userid;
    INSERT INTO other (col1, col2, col3) 
      VALUES ('Val1', 'Val2', 200)
      RETURNING id INTO otherid;
END $$

In folder+file nhost/seeds/default/001-initial.sql. Runs on local dev when you first do an nhost up or perhaps even nhost dev -d. I don't believe it runs in the cloud.

If you want to apply it against your cloud instance(s), use a Postgres client (CLI or pgAdmin etc) to connect to the DB and run the script directly.

https://docs.nhost.io/database#seed-data

This is great, thanks for all your help so far PK. These would be good examples to add to the seed-data section in the docs.