The correct way to provide BigQuery Client Tools for a user for a table supported by sheets, but the API still does not work

Question

In our organization, we have a standard user who created all the projects, tables and some Google tables that they indicate.

We created another user in the organization, granted them Bigquery Data Viewer and Bigquery Data User permissions, and allowed them access to the Google Sheet. They have access to standard tables, but not those supported by Google Sheets; they get an error Encountered an error while globbing file pattern.

Granting permission to the new Bigquery Editor user, they can create their own tables supported by sheets, but they get this dialog box:

After granting permissions, we can remove the Bigquery Editor permission, and the user continues to access all existing Bigquery tables, including those supported by the sheets.

In the new Google account https://security.google.com/settings/security/permissions we see that they have added Bigquery Client Tools:

Question 1: What is the correct way to grant this permission to new users? We found two workarounds:

• Take the URL from the pop-up window and edit it for new users without having to give them editor permission.
• Ask users to save the query to Google Sheets, which pops up.

But we did not find a software way to give users this permission.

Further...

However, the API still seems to have problems. We ran:

 gcloud auth login --enable-gdrive-access # approve permissions in the web browser for the new user bq query --apilog=apil.log "SELECT * FROM [warehouse:catalog.table]" 

and our log shows either "An internal error occurred and the request could not be completed." , or "Encountered an error while globbing file pattern." ; eg:

 INFO:root:{ [...] "jobReference": { "projectId": [...] "jobId": "bqjob_r2410ded84270edc0_00000158fdea47bb_1" }, "status": { "state": "DONE", "errorResult": { "reason": "internalError", "message": "An internal error occurred and the request could not be completed." }, "errors": [ { "reason": "internalError", "message": "An internal error occurred and the request could not be completed." } ] }, [...] } 

Question 2: What else do we need to do to provide data access through the API / cli?