Skip to main content
Build and deploy your first Buster agent in under 10 minutes. Prerequisites:
  • A GitHub account with admin access to a repository
  • A data warehouse (Snowflake, BigQuery, PostgreSQL, etc.)
  • Warehouse credentials

​
1. Connect your stack

Go to buster.so and create an account. When prompted, install the Buster GitHub App on your repository. Choose “All repositories” or select specific repos. Connect the GitHub App You will then be prompted to connect your data warehouse. See the Data Warehouse integrations section for connection details specific to your warehouse.
What happens during onboarding: When you connect your repository, Buster automatically analyzes your project and creates an AGENTS.md file. This file captures your technology stack, conventions, and patterns—giving all your agents deep context about your specific project.

​
2. Install the CLI

Install the Buster CLI on your local machine: 
curl -fsSL https://install.buster.so | sh
Verify the installation: 
buster --version

​
3. Authenticate

Get your API key from the Buster platform under Settings → API Keys, then authenticate: 
buster auth --api-key YOUR_API_KEY

​
4. Create project configuration

In the root of your repository, create a buster.yml file:
buster.yml
projects:
  - name: my-project
    data_source: my_datasource  # Must match the name in Buster platform
    schema: public

agents:
  - .buster/agents/pr-greeter.yml
The data_source value must exactly match the data source name you configured in the Buster platform.

​
5. Create an agent

Create a directory for your agents and add your first agent file: 
mkdir -p .buster/agents
Create .buster/agents/pr-greeter.yml:
pr-greeter.yml
name: pr-greeter
description: Comments on pull requests with a friendly summary

prompt: |
  You are a helpful PR assistant. When a pull request is opened or updated:
  
  1. Read the PR title, description, and list of changed files
  2. Count the total files changed
  3. Add a friendly comment to the PR with:
     - A greeting
     - The number of files changed
     - The types of files modified (e.g., "SQL models", "YAML configs")
  
  Keep your comment concise and friendly.

triggers:
  - event: pull_request
    types: ['opened', 'synchronize']

​
6. Deploy

Deploy your configuration and agent to the Buster platform: 
buster deploy
You should see output confirming your agent was deployed successfully.
Run buster deploy --dry-run first to validate your configuration without actually deploying.

​
7. Test your agent

Create a test branch and open a pull request. Within a minute, you’ll see:
  • A GitHub Check Run from Buster showing the agent executed
  • A comment from Buster with a friendly summary of your changes
You can view the full execution log in the Runs tab of the Buster web app.

​
Next steps

buster.yml Reference

Complete reference for buster.yml configuration.

Creating Agents

Deep dive into agent configuration, prompts, and triggers.

Auto-Update Documentation

Keep your dbt documentation up to date automatically.

Pull Request Review

Review SQL changes for data impact and breaking changes.