Bruin CLI

Appearance

Bruin MCP (Model Context Protocol) ​

Bruin is a CLI tool that allows users to ingest data from many different sources, transform data using SQL and Python, run data quality checks, compare table outputs, and more.

Bruin MCP

Bruin MCP allows you to extend your AI agents to analyze, understand, and build upon your data using Bruin CLI. It allows AI agents to query data, compare tables, ingest data, and build pipelines on them.

Setup ​

TIP

Please make sure you have Bruin CLI installed.

Claude Code ​

To use Bruin MCP in Claude Code, run the following command in your terminal:

bash
claude mcp add bruin -- bruin mcp

Cursor IDE & Cursor CLI ​

To use Bruin MCP in Cursor IDE, go to Cursor Settings > MCP & Integrations > Add Custom MCP, then add the following configuration to the configuration file:

json
{
  "mcpServers": {
    "bruin": {
      "command": "bruin",
      "args": ["mcp"]
    }
  }
}

Codex CLI ​

To use Bruin MCP in Codex CLI, add the following configuration to your ~/.codex/config.toml file:

toml
[mcp_servers.bruin]
command = "bruin"
args = ["mcp"]

Bruin Environment Setup ​

In order to make use of Bruin MCP, you need to have a Bruin project. A Bruin project is a folder that contains a .bruin.yml file.

The simplest way to get started is to create a new Bruin project using bruin init command.

bash
# Initialize a new Bruin project with a DuckDB template
bruin init duckdb

# If you don't give any arguments, it will ask you to select a template
bruin init

This will create a new Bruin project with a DuckDB database as a starting point.

You can

Bruin allows you to use different environments for your projects. You can read more about environments to learn more.

Usage ​

There are a couple of areas Bruin MCP can be helpful for you. Bruin CLI has complete functionality, and MCP is a simple layer to bridge the gap between the CLI and the AI editor.

Analyzing data ​

Bruin CLI already has the ability to run queries across different databases. MCP server allows your AI agents to understand how to run queries on the given databases.

A common example is to ask the AI agent to analyze a data in a certain table, such as "What is the average revenue of the products in the products table?" or "What is the total revenue of the orders in the orders table?"

Data ingestion ​

Bruin already has connectors for tens of sources, Bruin MCP allows your AI agents to understand how to ingest data from the given sources.

You can use Bruin MCP to get data from Shopify, for instance. You can simply say "bring all of my Shopify order data into BigQuery", and just enjoy the rest.

Data comparison ​

A common task when dealing with changes in data pipelines is to compare data between two different environments. You create a new version of the table in your development environment, and you want to compare it with the production table to make sure it's correct.

Thankfully, Bruin CLI already supports:

Using the Bruin MCP server, you can build new data models while ensuring that the data is correct and up-to-date using AI.

And more ​

Once MCP setup is complete, you can ask questions in Cursor IDE or Claude Code like:

  • "How do I create a BigQuery asset in Bruin?"
  • "How is a pipeline.yml file configured in Bruin?"
  • "What data sources does Bruin support for ingestion?"
  • "How do I set up a Snowflake connection in Bruin?"
  • "How can I connect to my PostgreSQL database and run a query with Bruin?"
  • "How do I create a table in my data warehouse using Bruin?"
  • "How can I build a data pipeline in Bruin for ingesting CSV files?"
  • "How do I run data quality checks on my tables in Bruin?"

Or you can give direct commands like:

  • "Create a BigQuery asset in Bruin"
  • "Configure a pipeline.yml file in Bruin"
  • "Show me the data sources Bruin supports for ingestion"
  • "Set up a Snowflake connection in Bruin"
  • "Connect to my PostgreSQL database and run a query with Bruin"
  • "Create a table in my data warehouse using Bruin"
  • "Build a data pipeline in Bruin for ingesting CSV files"
  • "Run data quality checks on my tables in BVruin"

The AI assistant will answer these questions using up-to-date Bruin documentation and provide you with accurate examples. It can also execute Bruin commands directly to help you connect to databases, run queries, perform ingestion tasks, and create complete data pipelines.

Feedback ​

We'd love to hear your feedback on Bruin MCP. Please create an issue so that we can improve Bruin CLI & Bruin MCP.