How to Configure Windsurf Bigquery Permissions

Published
Length
3 min read

Luis Felipe Dias

All posts

You set up windsurf bigquery permissions by assigning least-privilege roles to the identity your agent uses. This identity comes from your local shell credentials or a service account in your server settings. Teams often reuse their personal developer accounts to get the agent running fast. Still, broad roles let an AI agent read private tables or drop datasets by mistake. How windsurf bigquery permissions work in practice On June 2, 2026, Cognition renamed Windsurf to Devin Desktop. The edit

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 windsurf bigquery permissions by assigning least-privilege roles to the identity your agent uses. This identity comes from your local shell credentials or a service account in your server settings. Teams often reuse their personal developer accounts to get the agent running fast. Still, broad roles let an AI agent read private tables or drop datasets by mistake.

How windsurf bigquery permissions work in practice

On June 2, 2026, Cognition renamed Windsurf to Devin Desktop. The editor keeps the same features, and the old web docs now redirect to the new site. Its built-in agent can run terminal commands on its own after you grant permission. It opens a dedicated shell on your machine, using zsh on macOS or bash on Linux.

Because this shell loads your normal profile, the agent inherits your aliases and exported keys. If you log in to Google Cloud locally, the agent shares those exact rights. The legacy Cascade agent also reaches external systems through Model Context Protocol servers. Whether the agent runs a shell command or calls an MCP tool, Google Cloud checks one thing. It checks the IAM roles attached to the active user account or service account key.

To keep data safe, you should create a dedicated service account for the agent. Grant that account read-only roles on specific datasets instead of the whole project. Do not give the agent admin roles or permission to run big jobs without limits. This stops the agent from modifying schemas or reading restricted tables outside its scope.

What native IAM roles solve and what they miss

Native Google Cloud IAM roles decide which principals can reach a project or dataset. They are great for blocking access to entire datasets that an agent does not need. They also let you separate read actions from write actions at the table level.

Still, static roles only check who logs in when a session starts. Once the agent has read rights on a table, IAM cannot inspect the SQL query in flight. An agent with valid read permissions can still pull millions of rows at once. It can also query columns that hold emails, phone numbers, card numbers, or national IDs. Without query-level checks, that raw data flows straight into the model context window.

BigQuery IAM + AI Agent Permissions: architecture patterns and best practices

Continue reading? Get the full guide.

Free. No spam. Unsubscribe anytime.

Why local shell state creates hidden risks

Because the agent loads your shell profile, it picks up your active cloud session by default. Developers often hold broad editor or owner roles for debugging tasks in staging or production. When the AI agent runs a command in that terminal, it acts with that same power. Even if you tell the agent to only read data, a bad prompt can trigger a destructive command.

Storing a separate service account key on disk is not much safer. Any process running in that user shell can read the key file and use it elsewhere. Teams need a way to govern individual statements and mask sensitive fields in query results. They also need to keep raw cloud keys off the developer laptop.

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 sits inline in the wire protocol beside those services as an open-source proxy, a sidecar for your Postgres or API. On every query an AI agent or engineer runs against production, hoop.dev decides to pass, mask, block, or hold it for human review.

Your login details and client tools stay as they are while hoop.dev guardrails refuse bad deletes or drops before they run, and hoop.dev data masking rewrites matched sensitive values in responses 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

Did Windsurf change its name to Devin Desktop?

Yes, Cognition announced on June 2, 2026, that Windsurf is becoming Devin Desktop. It keeps the same IDE features, and the old docs links now redirect to the new site.

How does the agent authenticate to BigQuery from my machine?

The agent runs commands in a local shell that loads your profile and exported keys. It can also use credentials passed to an MCP server that queries BigQuery on its behalf.

Can BigQuery IAM stop the agent from reading sensitive columns?

Standard dataset roles apply to whole tables, though you can set up policy tags for columns. Still, IAM alone cannot rewrite query results to mask emails or card numbers on the fly.

The query runs. The emails stay hidden.