Bruin AI Step 3 of 4

Build a Self-Healing Pipeline Agent

Local or Cloud

Set up one path first. Both read the same asset files and the same policy, but the access controls are different.

Pick a local coding agent whenPick a Bruin Cloud agent when
You want to work with an agent in your IDE, terminal, or coding agent app, such as Cursor.The team wants one shared agent in Bruin Cloud, Slack, or Teams.
You want local Bruin MCP and the CLI.You want a browser agent attached to a Cloud project.
Access is scoped per command with --environment, against the environments already in .bruin.yml.Access is scoped by the connection set attached to the agent.
Your Git setup can create branches and open pull requests.Read-only diagnosis is enough, or a separate Git integration has write access.

A connection set is a named group of database credentials attached to a Cloud agent, separate from the connections your pipelines run on. That separation is the whole point: the agent can have less access than the pipeline it is fixing.

Sign up to our newsletter

Practical updates on open-source data pipelines, AI analysts, governance, and what we are shipping at Bruin.

The signup form is hosted by Brevo. Accept cookies to load it.