Skip to main content
AI Learning Resources

Quick Overview

This video is a product demonstration presented by Karl Evans, Senior Manager of Sales Engineering at Harness, showing integrations between Harness and Google Cloud. Evans demonstrates how developers can use the Antigravity IDE and Gemini models to create, manage, and debug CI/CD pipelines without leaving their coding environment.

Key Points

  • 1.Harness integrates with Google Cloud Antigravity IDE and Gemini models to enable pipeline creation and troubleshooting directly inside developer environments.
  • 2.Developers can generate comprehensive CI/CD pipelines, including Docker builds, security scanning, approval gates, and Kubernetes canary deployments, using natural language prompts.
  • 3.Pipeline configurations authored through the IDE are translated into declarative YAML and visualized within the Harness web console.
  • 4.Gemini can analyze past deployment failures to identify root causes, such as missing pipeline parameters, and suggest fixes without leaving the editor.
  • 5.The integration provides high-level diagnostics across multiple recent deployments to surface common failure trends across microservices and cluster environments.

Summary

Karl Evans, Senior Manager of Sales Engineering for Alliances at Harness, presents how the Harness agentic DevSecOps platform integrates with Google Cloud tooling, including Antigravity IDE and Gemini models. He explains that while a common development pattern is to write code in the IDE and trigger deployments through pull requests, developers still need deployment pipelines in place. To eliminate context switching, Harness enables developers to generate and maintain these pipelines directly from their editor environment.

Inside the Antigravity IDE, Evans uses the Canary assistant powered by Gemini 3.7 Flash to request a new pipeline in the KarlEvans project of the Sandbox organization, naming it Gemini Deployment. The prompt instructs the agent to create a pipeline that builds a Docker image, runs a software composition analysis scan, uploads to an artifact repository, includes an approval process, and performs a canary deployment to a Kubernetes cluster. The assistant processes the prompt and outputs a structured implementation plan with architecture diagrams, stage configurations, and the full YAML specification required by Harness. Evans approves the plan and switches to the Harness web console to verify that the build, scan, approval, and canary stages appear as expected.

Evans then turns to troubleshooting functionality embedded within the IDE. Entering a prompt asking why the last deployment in the Reference Architecture project failed causes the assistant to inspect recent pipeline executions and diagnose the root cause as missing parameters. The assistant also displays associated services and provides the specific configurations needed to repair the failure.

To view broader system health, Evans prompts the agent to analyze the last twenty deployments across the project to detect recurring trends. The agent produces a categorized report showing targeted services, failure frequencies, and underlying issues such as missing branch inputs, deployment timeouts, or infrastructure drift. It couples these findings with actionable recommendations, showing that developers can diagnose systemic pipeline failures and apply adjustments directly within the editor environment.

Authoring Deployment Pipelines from the IDE

Karl Evans demonstrates using natural language prompts within the Antigravity IDE to create deployment pipelines. Instead of navigating separate consoles, a developer asks the Gemini-powered Canary agent to construct a pipeline containing Docker image building, software composition analysis scans, artifact uploads, manual approvals, and a canary deployment to a Kubernetes cluster. The assistant returns a detailed implementation plan and converts it into the underlying Harness YAML specification.

Inspecting Pipeline Stages in Harness

Once generated in the IDE, the deployment pipeline is immediately reflected within the Harness user interface. Evans displays the graphical view of the pipeline, which includes the build and scan stage, the approval gate, and the progressive canary deployment steps leading into Kubernetes. Teams can govern these definitions using templates and policies designed for enterprise consistency.

Diagnosing Failure Trends and Root Causes

The assistant also handles operational troubleshooting inside the IDE. When prompted about the most recent deployment failure in the reference architecture project, the agent inspects execution details and pinpoints missing trigger parameters. Furthermore, asking for an evaluation of the last twenty failed runs generates an aggregated summary listing affected services, failure categories, frequency patterns, and direct recommendations for pipeline corrections.

The Bottom Line

The presentation demonstrates how combining Harness with Google Cloud Antigravity and Gemini brings continuous delivery authoring and diagnostic reporting directly into the development editor. It shows that both pipeline generation and failure investigations can be completed through conversational prompts without switching across distinct operational tools. The video concludes with instructions to install the Harness extension in Antigravity, while leaving specific enterprise governance policy enforcement and production rollout mechanisms to further exploration.

FAQ

What is Harness and how does it integrate with Google Cloud tools?

Harness is an agentic DevSecOps platform that integrates with Google Cloud Antigravity IDE and Gemini models to allow developers to create, update, and troubleshoot continuous integration and deployment pipelines directly within their editing workspace.

How can developers generate a new deployment pipeline using the Google Cloud Antigravity IDE?

Developers can prompt the integrated Canary assistant powered by Gemini with their pipeline requirements, such as image builds, security scans, approval steps, and canary deployments to Kubernetes. The assistant generates an implementation plan, creates the backing YAML specification, and establishes the pipeline in Harness.

How does the Gemini integration troubleshoot failed deployment pipelines directly inside the editor?

When asked about a failed deployment run, the assistant queries the relevant pipeline execution logs, identifies the specific reason for failure, such as missing parameters on a trigger, and outlines the corrective actions required to fix it.

What multi-run diagnostics can the assistant perform across recent deployment pipeline executions?

The assistant can analyze a set of recent deployment runs, such as the last twenty failures, to extract common failure patterns, affected microservices, failure frequencies, and suggested remediations across Kubernetes and cloud environments.

Where can developers install the Harness integration for Google Cloud Antigravity?

Developers can get started by navigating to the extensions marketplace inside the Antigravity IDE and searching for Harness.

Worth watching for

DevOps engineers, cloud architects, and software developers who want to manage and troubleshoot CI/CD pipelines directly inside their IDE using AI assistance.

  • harness
  • google-cloud
  • gemini
  • devsecops
  • kubernetes
  • ci-cd
Watch on YouTubeMore videos from Google Cloud Tech