Implementation:Mage ai Mage ai GitHub Actions GCP Cloud Run Deploy

Knowledge Sources	Mage AI Mage Deployment Docs
Domains	Deployment, CI_CD, GCP
Last Updated	2026-02-09 00:00 GMT

Overview

GitHub Actions workflow template for building a Docker image and deploying Mage to Google Cloud Run via Google Artifact Registry.

Description

This GitHub Actions workflow template named "Deploy to GCP Cloud Run" triggers on pushes to the master branch and automates the full build-and-deploy pipeline to Google Cloud Run. The workflow runs a single build-and-deploy job on ubuntu-latest with five steps (plus an optional sixth for DBT docs). It first checks out the repository using actions/checkout@master, then authenticates to Google Cloud using google-github-actions/auth@v0.6.0 with a service account credentials JSON stored in the GCP_CREDENTIALS GitHub secret, requesting an access_token format for Docker authentication. It logs in to the Google Artifact Registry (GAR) Docker registry at ${GAR_LOCATION}-docker.pkg.dev using docker/login-action@v1 with OAuth2 access token authentication. It then builds the Docker image and tags it with the full GAR path including project ID, repository, image name, and the git commit SHA. After pushing the image, it deploys to Cloud Run using google-github-actions/deploy-cloudrun@v0, targeting the specified service name and region. An optional commented-out step is included for deploying a separate DBT docs service using the same image with a -docs suffix appended to the service name. Five environment variables must be configured: GAR_LOCATION, GOOGLE_CLOUD_RUN_SERVICE_NAME, IMAGE, PROJECT_ID, and REPOSITORY.

Usage

Copy this template into your repository's .github/workflows/ directory. Replace the placeholder environment variable values with your actual GCP resource identifiers. Add your GCP service account credentials JSON as a GitHub repository secret named GCP_CREDENTIALS. Optionally uncomment the DBT docs deployment step if you use a separate Cloud Run service for DBT documentation.

Code Reference

Source Location

• Repository: mage-ai
• File: templates/github_actions/build_and_deploy_to_gcp_cloud_run.yml
• Lines: 1-60

Signature

name: Deploy to GCP Cloud Run

on:
  push:
    branches:
      - master

env:
  GAR_LOCATION: GAR_LOCATION
  GOOGLE_CLOUD_RUN_SERVICE_NAME: GOOGLE_CLOUD_RUN_SERVICE_NAME
  IMAGE: IMAGE
  PROJECT_ID: PROJECT_ID
  REPOSITORY: REPOSITORY

jobs:
  build-and-deploy:
    name: Build, Publish, and Deploy
    runs-on: ubuntu-latest
    steps:
    - name: Checkout
    - name: Authenticate to Google Cloud
    - name: Docker login
    - name: Docker build
    - name: Docker push
    - name: Deploy to Cloud Run
    # Optional: Deploy DBT docs service

Import

# Template file - copy to your project's GitHub Actions workflow directory
# cp templates/github_actions/build_and_deploy_to_gcp_cloud_run.yml \
#    .github/workflows/deploy_to_cloud_run.yml

I/O Contract

Inputs

Name	Type	Required	Description
GAR_LOCATION	workflow env var	Yes	Google Artifact Registry region (e.g., us-central1); also used as the Cloud Run deployment region
GOOGLE_CLOUD_RUN_SERVICE_NAME	workflow env var	Yes	Name of the Cloud Run service to deploy to
IMAGE	workflow env var	Yes	Docker image name within the GAR repository
PROJECT_ID	workflow env var	Yes	Google Cloud project ID
REPOSITORY	workflow env var	Yes	Google Artifact Registry repository name
GCP_CREDENTIALS	GitHub secret	Yes	GCP service account credentials JSON for authentication and deployment

Outputs

Name	Type	Description
Docker image	GAR image	Docker image tagged with the git commit SHA pushed to Google Artifact Registry at ${GAR_LOCATION}-docker.pkg.dev/${PROJECT_ID}/${REPOSITORY}/${IMAGE}:${GITHUB_SHA}
Cloud Run deployment	GCP Cloud Run service	Cloud Run service updated with the new container image revision
DBT docs service (optional)	GCP Cloud Run service	Optional separate Cloud Run service for DBT documentation (commented out by default)

Workflow Steps

1. Checkout -- Clones the repository using actions/checkout@master
2. Authenticate to Google Cloud -- Authenticates using google-github-actions/auth@v0.6.0 with the GCP_CREDENTIALS secret in access_token format
3. Docker login -- Logs in to the GAR Docker registry at ${GAR_LOCATION}-docker.pkg.dev using the OAuth2 access token from the auth step
4. Docker build -- Builds the Docker image and tags it as ${GAR_LOCATION}-docker.pkg.dev/${PROJECT_ID}/${REPOSITORY}/${IMAGE}:${GITHUB_SHA}
5. Docker push -- Pushes the tagged image to Google Artifact Registry
6. Deploy to Cloud Run -- Deploys the image to the specified Cloud Run service in the ${GAR_LOCATION} region using google-github-actions/deploy-cloudrun@v0
7. (Optional) Deploy DBT docs -- Commented-out step to deploy the same image to a separate Cloud Run service named ${GOOGLE_CLOUD_RUN_SERVICE_NAME}-docs

Docker Image Path

The full Docker image path follows the Google Artifact Registry convention:

${GAR_LOCATION}-docker.pkg.dev/${PROJECT_ID}/${REPOSITORY}/${IMAGE}:${GITHUB_SHA}

For example: us-central1-docker.pkg.dev/my-project/mage-repo/mage-ai:abc123def456

Usage Examples

# .github/workflows/deploy_to_cloud_run.yml
name: Deploy to GCP Cloud Run

on:
  push:
    branches:
      - master

env:
  GAR_LOCATION: us-central1
  GOOGLE_CLOUD_RUN_SERVICE_NAME: mage-ai-service
  IMAGE: mage-ai
  PROJECT_ID: my-gcp-project-id
  REPOSITORY: mage-ai-repo

jobs:
  build-and-deploy:
    name: Build, Publish, and Deploy
    runs-on: ubuntu-latest
    steps:
    - name: Checkout
      uses: actions/checkout@master
    # ... remaining steps as in template

# To enable DBT docs deployment, uncomment the optional step:
    - id: deploy-docs
      uses: google-github-actions/deploy-cloudrun@v0
      with:
        service: ${{ env.GOOGLE_CLOUD_RUN_SERVICE_NAME }}-docs
        image: ${{ env.GAR_LOCATION }}-docker.pkg.dev/${{ env.PROJECT_ID }}/${{ env.REPOSITORY }}/${{ env.IMAGE }}:${{ github.sha }}
        region: ${{ env.GAR_LOCATION }}

Related Pages

Implements Principle

• Principle:Mage_ai_Mage_ai_Source_Connector_Scaffolding

Requires Environment

• Environment:Mage_ai_Mage_ai_Python_3_9_Runtime