How to Configure Continue Redshift Read Only Access
You set up continue redshift read only access by linking a viewer role to your editor. This setup lets your AI assistant read tables and schemas while blocking all write commands. Many teams start by pasting admin database keys into their local assistant settings file. That shortcut gives a coding assistant the power to drop tables or change production rows. A strict viewer role stops the assistant from running destructive commands on your warehouse. How to set up continue redshift read only a
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 continue redshift read only access by linking a viewer role to your editor. This setup lets your AI assistant read tables and schemas while blocking all write commands. Many teams start by pasting admin database keys into their local assistant settings file. That shortcut gives a coding assistant the power to drop tables or change production rows. A strict viewer role stops the assistant from running destructive commands on your warehouse.
How to set up continue redshift read only access
First, your data team creates a dedicated database user or group inside the data warehouse. They grant this user usage rights on the target schema and select rights on tables. They also set default privileges so future tables in that schema stay readable as well. They leave out insert, update, delete, and drop rights for this specific database user.
Next, the developer opens the local settings file for the coding assistant in their editor. The assistant uses context providers or tool servers to talk to external data sources. You add a database entry to that local file using the viewer user details. You also supply the warehouse host name, port number, and database name in this step.
When a developer asks a question, the assistant runs a select query with this role. The warehouse checks the role permissions before it reads any rows from disk. If the assistant tries to alter a table, the database engine rejects the command. The assistant then uses the returned rows or schema details to write code suggestions.
What native viewer roles solve and what they miss
A native database viewer role solves the problem of accidental data loss or corruption. Your coding assistant can inspect column types, check table joins, and test read queries. Developers get accurate SQL help in their editor without risking production table structures. The warehouse blocks every write statement at the database engine level.
Still, native database permissions only decide who logs in and which tables they can read. They do not inspect what individual select queries do while they run in flight. A read role still lets the assistant pull raw emails, phone numbers, or card numbers. That sensitive data then goes straight to an external model provider inside the prompt context.
Auditor Read-Only Access + Customer Support Access to Production: architecture patterns and best practices
Continue reading? Get the full guide.
Free. No spam. Unsubscribe anytime.
Storing static database passwords on developer laptops also creates a key management problem. If a laptop is lost, teams must rotate those shared keys across every developer machine. Native roles also struggle to tie queries back to a single person when teams share keys. Finally, an unguarded select query on a massive table can still slow down the cluster.
Why query-level controls matter for AI coding tools
AI coding tools run queries faster and more often than human engineers do by hand. When an assistant explores a schema, it may scan several large tables to gather context. Database admins need a way to check each query before the warehouse runs it. They also need to rewrite matched fields like national IDs or card numbers in results.
Relying on static local settings files makes it hard to enforce these rules across a team. Every developer manages their own file, and someone might swap in a personal admin key. Moving credential handling and data masking out of the laptop keeps the warehouse safe.
Check database and API queries on the wire
Identity tools, roles, and network rules decide who opens a session. Setup alone cannot check what each query does or mask sensitive values in a response once a session starts. When workflows touch Postgres, MySQL, SQL Server, MongoDB, ClickHouse, Kubernetes APIs, SSH, or internal HTTP APIs, hoop.dev runs beside those services as an open-source proxy, a sidecar for your Postgres or API. It inspects requests coming in and results going back on the wire.
Your login details and client tools stay as they are while hoop.dev guardrails refuse bad deletes or drops before they run. On responses, hoop.dev data masking rewrites matched sensitive values in memory. Furthermore, hoop.dev holds SOC 2 Type II, is MIT licensed, and writes every query and decision to an audit trail that generates evidence for HIPAA, GDPR, and PCI DSS. See the source on GitHub.
Frequently asked questions
Can a read-only database role stop an AI assistant from reading sensitive columns?
You can revoke table rights or build database views that leave out sensitive columns. Still, managing custom views for every table takes a lot of extra work for data teams. Dynamic data masking rewrites matched fields like emails and phone numbers in query results instead.
Why should teams avoid putting static database keys in local assistant settings?
Local settings files sit in plain text on each developer laptop. If malicious code reads that file, the attacker gets direct access to your data warehouse. Removing static keys from laptops keeps your warehouse credentials safe from local leaks.
Can a read-only query still hurt data warehouse performance?
Yes, a poorly written select query can scan billions of rows and use up cluster memory. AI assistants often generate broad queries when they do not know the table size. Setting statement timeouts and row limits helps protect the cluster from heavy read queries.
Related guides
In this post