How to Configure JetBrains AI Bigquery Prevent Delete

Published
Length
3 min read

Luis Felipe Dias

All posts

To set up jetbrains ai bigquery prevent delete rules, restrict your cloud role permissions. Remove table delete and row update rights from the identity used by your IDE. Engineering teams often link coding assistants to cloud data warehouses using local developer keys. When those keys hold broad editor roles, an assistant can run destructive queries by mistake. How to set up jetbrains ai bigquery prevent delete Your IDE assistant uses the same database connection that you set up for manual wo

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.

To set up jetbrains ai bigquery prevent delete rules, restrict your cloud role permissions. Remove table delete and row update rights from the identity used by your IDE. Engineering teams often link coding assistants to cloud data warehouses using local developer keys. When those keys hold broad editor roles, an assistant can run destructive queries by mistake.

How to set up jetbrains ai bigquery prevent delete

Your IDE assistant uses the same database connection that you set up for manual work. It reads your database schema to write SQL queries and runs them with your user identity. To stop delete actions at the cloud level, you must adjust your Google Cloud IAM roles. Standard predefined roles often group read, write, and delete actions together in one broad package.

You can fix this by creating a custom IAM role in your cloud project. Start by copying the permissions from a basic data viewer or data editor role. Next, remove the specific permissions that allow table deletion or data modification. Assign this new custom role to the user account or service account in your IDE.

When the assistant tries to drop a table, the cloud API checks your custom role. The API blocks the request before it touches your stored data. You can also turn on table deletion protection inside the cloud console for critical datasets. This adds a second check that stops anyone from dropping a whole table by accident.

Why static cloud roles leave a gap for coding assistants

Custom IAM roles control who can log in and which API calls they can make. This works well for accounts that only ever need to read data. Still, developers often need to write data, update rows, and manage staging tables during daily work. In BigQuery, a single permission often controls all data manipulation statements, including inserts, updates, and deletes.

AI Data Exfiltration Prevention + BigQuery IAM: architecture patterns and best practices

Continue reading? Get the full guide.

Free. No spam. Unsubscribe anytime.

If you block that permission in IAM, the developer cannot insert test rows or update records. If you allow that permission, the AI assistant can also run a row delete statement. Static cloud roles cannot read the text of a SQL query to tell these actions apart. They only check the general API call type at the project or dataset level.

Also, native cloud permissions do not change the raw data returned to the IDE. When the assistant runs a select query, the database sends back full rows of customer data. Sensitive fields like emails, phone numbers, card numbers, and national IDs flow straight into the prompt context. Teams need a way to inspect individual SQL statements and rewrite sensitive fields in query results.

Govern statements and responses across databases and APIs

Good config and scoped roles are needed setup. They do not check single queries in flight or mask sensitive fields in returned rows. Across Postgres, MySQL, SQL Server, MongoDB, ClickHouse, Kubernetes APIs, SSH, and internal HTTP endpoints, hoop.dev runs beside each supported service as an open-source proxy, a sidecar for your database or API.

First, hoop.dev guardrails block banned queries before they run. Then hoop.dev data masking rewrites matched sensitive values in responses in memory. The audit trail records every statement and verdict per resource. Furthermore, hoop.dev holds SOC 2 Type II and is open source on GitHub.

Frequently asked questions

Can I block row deletes in BigQuery while allowing inserts through IAM?

No, standard cloud permissions group row inserts, updates, and deletes under the same data update right. You need statement-level inspection to block delete queries while still allowing insert statements.

Does the IDE assistant use a separate database account by default?

No, the assistant shares the active database session configured inside your IDE database tool window. You must create a separate connection profile if you want to isolate its cloud role.

How can I protect sensitive columns when the assistant runs a select query?

You can apply policy tags in your cloud catalog to restrict specific columns by role. Alternatively, you can apply dynamic data masking to rewrite emails and card numbers in query results.

The query runs. The emails stay hidden.