How to Configure JetBrains AI Bigquery Read Only Access

Published
Length
3 min read

Luis Felipe Dias

All posts

You set up jetbrains ai bigquery read only access by pairing a viewer role with a service account. Next, you link that key to the database tool inside your editor. Many teams start by reusing an admin key just to make the tool work fast. That shortcut lets an assistant modify tables, drop views, or read private rows by mistake. Scoping the cloud role down to read tasks is the right first step for safety. How jetbrains ai bigquery read only access works The coding assistant inside your editor

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 jetbrains ai bigquery read only access by pairing a viewer role with a service account. Next, you link that key to the database tool inside your editor. Many teams start by reusing an admin key just to make the tool work fast. That shortcut lets an assistant modify tables, drop views, or read private rows by mistake. Scoping the cloud role down to read tasks is the right first step for safety.

How jetbrains ai bigquery read only access works

The coding assistant inside your editor does not talk to your cloud data warehouse on its own. Instead, it relies on the built-in database tool to read schemas and run queries. You first create a cloud identity that only has permission to view data and run jobs. You then attach this identity to a project or a single dataset in your cloud console.

Once the cloud role is ready, you open your editor and add a new data source. You pick the warehouse driver and load the credentials for your scoped cloud identity. The editor tests the link and pulls down the list of tables, columns, and views. When you ask the assistant to write or check a query, it reads this local schema context. If you let the assistant run a query, the warehouse checks the cloud role first. Any command that tries to write, update, or delete rows fails right away at the database level.

What native read roles solve and where they fall short

Native cloud roles are great at stopping changes to your stored data and table structures. A viewer role makes sure a bad prompt cannot drop a table or overwrite records. It also lets you limit which datasets show up inside the editor for a given developer. For basic schema checks and test data, this native setup covers the main write risks.

Still, cloud permissions only decide who can log in and which tables they can read. They do not inspect what individual queries do in flight or what rows come back. A read role still lets a developer or assistant pull millions of rows in one run. If a table holds emails, phone numbers, card numbers, or national IDs, the query returns them in full. Those raw values can then flow into prompt context and go out to external model servers. Also, shared service account keys sit on laptops and are hard to rotate when people change roles.

Auditor Read-Only Access + AI Model Access Control: architecture patterns and best practices

Continue reading? Get the full guide.

Free. No spam. Unsubscribe anytime.

Managing query risks and data privacy for AI editors

To keep production data safe, teams need controls that work at the query and column level. Static roles in a cloud console cannot tell if a query comes from a human or an assistant. They also cannot rewrite sensitive fields in query results on the fly without extra database views. Building and maintaining separate masked views for every table takes too much time for data teams.

A better pattern checks each request right before it hits the warehouse and before results return. Data masking rewrites matched fields like emails, phone numbers, and card numbers in the query output. This step keeps private values out of the editor while still letting queries run normal checks. Teams also benefit when developers log in with their own identity instead of sharing static key files. That way, audit logs show the exact person behind every query that the editor runs.

Add runtime guardrails and masking beside your databases and APIs

Login rules, role grants, and CI checks govern who connects to your stack. Yet any allowed script, engineer, or AI agent still runs commands with the full scope of its role. For workloads on Postgres, MySQL, SQL Server, MongoDB, ClickHouse, Kubernetes, SSH, or internal HTTP services, hoop.dev sits beside the service as an open-source proxy, a sidecar for your database or API.

Before a query runs, hoop.dev guardrails check the statement against your rules. An AI Analyzer you configure can score intent or hold risky calls for human review. Next, hoop.dev data masking rewrites sensitive fields in responses, and each verdict is logged per listener. Explore the MIT-licensed project on GitHub.

Frequently asked questions

Can JetBrains AI run BigQuery queries without write permissions?

Yes, the tool only needs permission to read dataset metadata and start read query jobs. It can inspect schemas, suggest SQL, and read results without any write roles.

Does a BigQuery viewer role hide sensitive columns from the AI assistant?

No, a standard viewer role lets the user read every column and row in the dataset by default. Unless you add column tags or masking rules, raw emails and card numbers return in full.

Why should teams avoid using shared service account keys in IDEs?

Shared key files sit on local disks and never expire unless someone deletes them by hand. They also hide which team member ran a specific query in your cloud audit logs.

The query runs. The emails stay hidden.