AI Creators
Google Cloud Tech
End-to-end agentic development with Google Gemini, Antigravity, and GitLab
Sep 2, 20268 min watch
Quick Overview
This technical demonstration, presented by GitLab Senior Ecosystem Solutions Architect Regnard Raquedan, showcases agentic software development using Google Gemini, Antigravity, and GitLab. The video walks through redesigning and deploying an existing Flask application to illustrate how AI tooling integrates with modern source control and cloud deployment platforms.
Key Points
- 1.GitLab integrates directly with Google Cloud IAM and Google Artifact Registry to manage permissions and container images.
- 2.Developers can use Gemini inside the Antigravity IDE to create visual UI mockups and translate design images into working application code.
- 3.Model Context Protocol (MCP) server integration allows the Antigravity IDE to create and update GitLab issues and track planning workflows smoothly.
- 4.GitLab Duo Agent Platform enables automated code review using specialized agents such as Develop with Gemini before merging code.
- 5.GitLab CI/CD pipelines run automated security tests, upload build artifacts to Google Artifact Registry, and deploy directly to Cloud Run.
Summary
Regnard Raquedan demonstrates an end-to-end agentic software development workflow combining Google Gemini, the Antigravity IDE, and GitLab. The starting point is a Python Flask application hosted in GitLab that retrieves financial data to display a plain S&P 500 index report. Raquedan verifies that project integrations for Google Artifact Management and Google Cloud IAM are already configured.
Within the Antigravity development environment, Raquedan prompts Gemini to generate a proposed user interface mockup as a PNG image representing a dark-themed, glassmorphic redesign. The generated image file is saved into the project codebase. Next, Antigravity communicates with GitLab via a Model Context Protocol server, allowing Gemini to automatically open a planning issue in GitLab for the UI refresh.
In the GitLab web interface, Raquedan converts the newly created issue into a merge request, generating a dedicated feature branch named 11-ui-refresh-planning. Returning to Antigravity, he switches the local environment to this new branch. He then prompts Gemini in the IDE chat to analyze the mockup image and write the necessary template and styling changes. Gemini modifies index.html and app.py, verifies the app locally by launching a test server, and proposes a commit message, which Raquedan accepts and pushes back to the remote repository.
Back in GitLab, the merge request displays the code changes. Raquedan assigns the Develop with Gemini agent from the GitLab Duo Agent Platform as a reviewer. The agent analyzes the pull request in real time and completes the review. After verifying the results, Raquedan merges the changes into the main branch.
Merging triggers an automated GitLab CI/CD pipeline comprising build, test, upload, and deploy stages. The test phase executes dependency scanning and secret detection, after which the container image is pushed to Google Artifact Registry and deployed directly to Cloud Run. The demonstration concludes with a live walkthrough of the newly deployed, modern interactive dashboard running in the browser.
Setting Up Integrations and Visualizing the UI
The demonstration begins in GitLab with a Python Flask application displaying a basic S&P 500 index report. After ensuring active integrations with Google Artifact Management and Google Cloud IAM, the Antigravity IDE is used with Gemini to produce a modern glassmorphism UI mockup image saved directly into the codebase.
Planning and Generating Code via Agentic IDE Workflows
Using an integrated Model Context Protocol server, Gemini creates a planning issue in GitLab directly from Antigravity. A corresponding merge request and feature branch are created in GitLab, after which Gemini inspects the mockup image in Antigravity, writes the updated HTML and Flask code, tests the app locally, and commits the changes.
Automated Review and Cloud Run Deployment
In GitLab, the merge request is reviewed automatically by assigning the Develop with Gemini agent from the GitLab Duo Agent Platform. Once the review passes and the branch is merged into main, the GitLab CI/CD pipeline builds the container, executes security scans, pushes to Google Artifact Registry, and deploys the modernized application to Cloud Run.
The Bottom Line
The video demonstrates how developers can combine Gemini, Antigravity, and GitLab to move from visual concept to live deployment through automated agentic workflows. It establishes a practical pattern for using AI models both inside the IDE for code generation and within the repository platform for issue tracking, code review, and automated deployment to Cloud Run. The demonstration leaves advanced configuration details and multi-environment scaling for developers to explore through the Google Cloud Marketplace.
FAQ
What is end-to-end agentic development with Google Gemini and GitLab?
It is a development workflow where AI agents powered by Google Gemini assist across the full software lifecycle, including UI design mockup generation, planning issue creation, code generation in the Antigravity IDE, automated merge request review, and CI/CD deployment to Cloud Run.
How does Antigravity connect directly to GitLab to create issues?
Antigravity connects to GitLab using a Model Context Protocol (MCP) server, allowing Gemini to receive instructions in the IDE and automatically create or update GitLab issues.
What role does the Develop with Gemini agent play in GitLab?
The Develop with Gemini agent operates within the GitLab Duo Agent Platform to automatically review code changes submitted in merge requests before they are merged into the main branch.
Which Google Cloud services are integrated into the GitLab CI/CD pipeline?
The pipeline integrates with Google Artifact Registry to store application container images and Google Cloud Run to host and serve the deployed web application.
What testing steps are executed in the GitLab CI/CD pipeline before deployment?
The pipeline runs dependency scanning using Gemnasium and secret detection during the test stage before pushing the build to Google Artifact Registry and deploying to Cloud Run.
Worth watching for
DevOps engineers, cloud architects, and software developers interested in integrating Google Cloud, Gemini AI agents, and GitLab into automated development pipelines.
- google-cloud
- gitlab
- gemini
- antigravity
- cloud-run
- ci-cd