MariaDB Company
You set up a mariadb company database by creating separate user accounts and roles. Then you map each team member to the lowest role they need for work. Most teams start by sharing one root account across their early apps and scripts. Shared accounts cause trouble when someone runs a bad update on a live table. Broad roles also let staff read full tables that hold emails and phone numbers. How to set up a mariadb company permission model A clean setup starts by turning off shared logins for h
Your coding agents run with your access.
Claude Code reaches prod with a developer's credentials, and approval prompts get clicked through. hoop.dev checks every query, masks sensitive data, and holds risky changes for a human.
You set up a mariadb company database by creating separate user accounts and roles. Then you map each team member to the lowest role they need for work. Most teams start by sharing one root account across their early apps and scripts. Shared accounts cause trouble when someone runs a bad update on a live table. Broad roles also let staff read full tables that hold emails and phone numbers.
How to set up a mariadb company permission model
A clean setup starts by turning off shared logins for humans and apps. You build roles that match daily tasks, such as read-only or schema change work. Database admins create these roles inside the engine and attach table rules to them. Then they grant those roles to specific user accounts tied to real people. When an engineer leaves the team, you drop their single user account right away. You do not need to rotate keys for every app when one person leaves.
Next, you restrict where each account can log in from on your network. The database checks both the user name and the host address during login. An app account should only log in from the app server address. Human accounts should only log in from approved office or cloud networks. You can also connect the database to an external directory for central login checks. This stops admins from managing passwords by hand across ten or twenty database servers. Every server checks the central directory before it lets a user open a session.
What native database roles solve and what they miss
Native database roles do a good job of stopping unauthorized people at the door. They make sure a billing app cannot read tables owned by another service. They also keep support staff from dropping tables or changing schemas by mistake. Audit plugins inside the database can write basic logs of who logged in.
Still, native roles only check who logs in and which tables they can touch. Once a user has read rights on a table, they get every column back by default. A single query can pull thousands of card numbers or national IDs at once. Native roles cannot rewrite matched fields in results to hide emails or phone numbers. They also cannot pause a risky update statement and ask a peer for approval. If an engineer has write rights for a fix, those rights stay on. One typo in a filter clause can still wipe out rows in production. Teams need statement-level checks and field masking to close this gap during live sessions.
Company: architecture patterns and best practices
Continue reading? Get the full guide.
Free. No spam. Unsubscribe anytime.
Running live checks on queries and results
To protect live data, teams add controls that inspect every query in flight. Instead of granting standing write rights, you approve single sessions or specific statements. When someone runs a query, a rule checks the statement before the database executes it. If the result holds emails or card numbers, masking rewrites matched fields in results. Every query and result check goes into a central log for audits.
Check MySQL queries and results on the wire
MySQL grants control who logs in. They do not check single queries or mask sensitive columns once a session starts. With hoop.dev, you run an open-source proxy, a sidecar for your MySQL, inline in the MySQL wire protocol. On every query an AI agent or human runs against production MySQL, RDS, or Aurora, hoop.dev decides to pass, mask, block, or hold it for review.
Multi-statement batches and prepared queries are split and checked one by one by hoop.dev guardrails, which refuse banned queries with a native MySQL error or hold risky ones for approval. On the response path, hoop.dev data masking rewrites string values in returned rows in memory, and every decision is logged. See the MIT-licensed project on GitHub.
Frequently asked questions
What is the difference between the open source project and the enterprise vendor?
The foundation runs the community database project and keeps the core server open source. The commercial mariadb company sells enterprise support, cloud hosting, and extra server plugins.
How should teams separate app accounts from human logins?
Give each app its own user account locked to the app server network address. Give humans individual accounts tied to your central directory with read-only roles by default.
Can native database roles hide sensitive fields in query results?
Standard roles only block or allow access to whole tables or fixed columns. To rewrite emails or card numbers on the fly, you need dynamic data masking.
Related guides
In this post