01 — Overview
Plenty of revenue still runs through a MySQL database that grew up with the company. Chamfer lets the people who depend on it work with it directly, without a database login and without a ticket to engineering. Describe the job, such as reviewing new sellers, fixing a subscription or reassigning a delivery, and the Bench builds the screens against your real tables. Most teams start with the table that generates the most support tickets.
You can split reads and writes across two connections, so lists and charts query a replica while edits go to the primary. Every write is parameterized and appears in the Logbook with the user and the values changed. Write a Spec like “never show the password_hash column” and the rule applies to every app that touches the table. Staging and production connections stay separate, and apps move between them only after approval.
02 — What you can do
Replica-aware reads
Send lists and charts to a read replica and keep the primary for the writes that need it.
Bulk actions with limits
Update hundreds of rows at once, with a cap on batch size and approval above it.
Column rules
Hide or mask columns such as emails or tokens in every app with one plain-English rule.
03 — FAQ
Before you connect MySQL
Does Chamfer work with MariaDB and Aurora MySQL?
Can we keep heavy queries off the primary?
What stops someone running a bad bulk update?
Do our users need their own MySQL accounts?
04 — Pairs well with
