01 — Overview
If your team already exposes data through GraphQL, Chamfer can build on it without a new backend. Connect the endpoint and the Bench reads the schema to learn your types, queries, mutations and enums. Describe the tool in a sentence and it writes operations that request only the fields each screen uses. Point it at staging and production endpoints separately, and promote the app between them with a Sign-off.
Enums become dropdowns, required arguments become required form fields, and nested objects become linked detail panels. You can read and edit every operation in the Code tab. Mutations follow the same rules as any other write in Chamfer: app roles decide who can run them, sensitive ones can wait for a Sign-off, and the Logbook stores the variables sent with each call. Schema changes show up the next time the app opens in the editor, with proposed fixes for the affected screens.
02 — What you can do
Schema introspection
The Bench reads types, enums and arguments from your endpoint or an SDL file.
Only the fields you show
Generated queries request the fields each screen displays, and nothing extra.
Guarded mutations
Limit mutations by role, and hold the ones that touch money or access for approval.
03 — FAQ
Before you connect GraphQL
What if introspection is off in production?
Does Chamfer support subscriptions?
How are auth headers handled?
Can we reuse queries we already have?
04 — Pairs well with