Query it in one request
Your tables appear under one root field named after your database, so names never clash with anyone else's:
HIVE=https://hornethive.io
DB=<database id> # from the dashboard
KEY=<api key> # read or write, one database or all
curl $HIVE/api/graphql/$DB -H "Authorization: Bearer $KEY" \
-H "Content-Type: application/json" \
-d '{"query":"{ u4f1c2d3e_shop { books(limit: 10) { id title } } }"}'How it is kept separate
Hasura is not exposed to the internet. Every GraphQL request goes through HornetHive, which checks the API key, then runs the query as your database's read role, or its write role for a write key. Those roles have permissions on your tables only, so a query cannot reach data that is not yours, whatever it asks for.
After you change the schema
GraphQL picks up new tables when you sync: one click on the database's GraphQL tab, one API call, or the sync_graphql tool over MCP. REST needs no sync at all.
On every plan
GraphQL is included from the free plan up, counted against the same monthly API calls as REST. Use it from your frontend framework of choice, Apollo, urql, Relay, or plain fetch.
Updated