How to Configure Make Bigquery Prevent Delete

Published
Length
3 min read

Luis Felipe Dias

All posts

You set up a make bigquery prevent delete rule by building a custom role. That role strips out table drop and row removal rights for the account. Most teams link Make to BigQuery with a standard editor account key. Broad editor roles let an automated scenario drop tables or wipe rows by mistake. One bad module setting can erase critical records in seconds. How to set up make bigquery prevent delete To lock down your data, you must change how the cloud account works. Make runs tasks in BigQuer

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 make bigquery prevent delete rule by building a custom role. That role strips out table drop and row removal rights for the account. Most teams link Make to BigQuery with a standard editor account key. Broad editor roles let an automated scenario drop tables or wipe rows by mistake. One bad module setting can erase critical records in seconds.

How to set up make bigquery prevent delete

To lock down your data, you must change how the cloud account works. Make runs tasks in BigQuery using a linked Google account or service key. Every action inside a Make scenario inherits the permissions of that single linked identity. If that identity has full write power, every module in Make has it too.

Start by creating a dedicated service account for your Make automation workflows. Do not reuse an admin account or a broad project editor role. Instead, build a custom role inside your cloud identity settings. Give this custom role permissions to read datasets, run jobs, and insert new rows. Leave out the permissions that allow deleting tables, datasets, or existing records.

You can also split your data into separate datasets based on risk. Keep your raw incoming data in an append-only dataset for Make to use. Place your core business tables in a protected dataset where Make has zero access. Scheduled cloud queries can then move and clean the data on a safe timer. This structure keeps your primary records safe even if a Make scenario goes wrong.

What custom cloud roles solve and what they miss

Custom roles and dataset boundaries solve the biggest risk of accidental table drops. When an API module tries to delete a protected table, the cloud blocks it. This approach stops full table wipes caused by a bad module setup in Make. It also limits the damage if someone steals your Make account key.

Still, static cloud permissions leave a huge gap during daily operations. Cloud identity rules check who logs in, not what a SQL query does. Make includes a module that sends raw SQL queries straight to BigQuery. To insert or update rows with SQL, the account needs general data edit rights. In many cloud setups, general edit rights also allow destructive SQL commands.

BigQuery IAM + End-to-End Encryption: architecture patterns and best practices

Continue reading? Get the full guide.

Free. No spam. Unsubscribe anytime.

A flawed filter in a Make scenario can trigger an unsafe update query. It might overwrite thousands of active customer emails or phone numbers with blank values. Even worse, a custom SQL module might run a command that empties a table. Static role settings cannot read the SQL text while the query runs. They also cannot mask sensitive fields like card numbers or national IDs in results. When a Make scenario reads a table, it pulls all raw fields into memory. Every connected app in that workflow can then see those private details.

Why query-level controls matter for automated workflows

Teams often struggle to balance fast automation with strict database safety. If you lock down the cloud role too much, legitimate Make scenarios fail. Builders then ask for broader permissions just to get their workflows running again. Soon, the automation account is back to having full editor power across your tables.

True safety requires checking the actual commands before the database runs them. You need a way to inspect every query coming from your automation tools. That check must block destructive commands while letting safe inserts and reads pass through. It should also rewrite matched fields like emails or phone numbers in query results. This keeps private data out of third-party automation logs without breaking the workflow.

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 a Make scenario delete BigQuery rows by accident?

Yes, a scenario can erase rows if the linked account has edit permissions. A bad variable in a custom SQL module can wipe out matching records fast.

Does a read-only role work for Make BigQuery modules?

A read-only role works if your scenario only pulls data for reports. If Make needs to add new records, you must grant data creation rights.

How do we protect sensitive fields when Make reads a table?

Native roles grant access to whole tables or specific columns, not dynamic text patterns. Data masking rewrites matched fields like card numbers or national IDs in the results.

The query runs. The emails stay hidden.