7 tips for using Antigravity 2.0 on enterprise codebases, coding phase
Hey everyone, welcome back. Martin Omander here. Most AI tutorials show someone building a toy app from scratch in five minutes. That’s neat, but it’s not reality when you're working on a massive, living enterprise codebase. Following up on our last planning video, this one is all about how to safely let Antigravity 2.0 loose to write, test, and refactor code inside complex, existing systems. To keep myself in the right managerial mindset, I treat the agent like a digital intern (represented by my physical stand-in, Ducky). Here are the 7 practical tactics that moved me from "prompting" to actually managing an agent effectively: 1. Encoding practices into skills: Baking your team's exact standards (like TDD) directly into the agent's DNA. 2. Committing the paper trail: Saving agent-generated implementation plans straight to your Git branch for better team context. 3. The test-and-tea workflow: Setting up an automated self-correction loop where the agent runs your tests and fixes errors completely on its own. 4. Giving the agent eyes: Using a headless local browser to catch broken layouts and misaligned UI before a human ever looks at it. 5. Offload the clock: Utilizing async background tasks so heavy integration suites don’t freeze your workspace. 6. Visualizing the architecture: Using the built-in Mermaid diagram viewer to instantly spot tight coupling or layering violations. 7. Pre-review and refactor: Running automated verification checks to catch code smells before submitting a PR, taking the load off human reviewers. With an agent handling the heavy lifting, I spend nearly 50% of my development time just ruthlessly refactoring to make the code as clean as humanly possible.If you’re running Antigravity 2.0 or similar agents on an established codebase, how are you structuring your workflows? Drop your favorite tricks or bottlenecks in the comments! I read every single one. Watch more Serverless Expeditions → https://goo.gle/ServerlessExpeditions 🔔 Subscribe to Google Cloud Tech → https://goo.gle/GoogleCloudTech #ServerlessExpeditions #GoogleCloud Speakers: Martin Omander Products Mentioned: Gemini, Antigravity

Inside Google Antigravity 2.0: The complete developer guide | The Agent Factory

Antigravity CLI Crash Course: Migrate from Gemini CLI

Saas style_ Google Drive

I stopped using /grill-me for coding. Here’s what I use instead:

Stop Prompting Claude. Use Karpathy's Method Instead.

AI-First Playbook: Do a Team's Work With AI (2026) | Peter Yang

Why The Best Engineers Are Solving Code Review Bottlenecks

Google Antigravity 2.0 Tutorial for Beginners: Learn 90% Of Antigravity in Under 25 Minutes

Finally. Agent Loops Clearly Explained.

/handoff is my new favourite skill

Antigravity 2.0 vs IDE vs CLI vs SDK: Real Scenarios

5 tips for using Antigravity 2.0 on enterprise codebases, planning phase

Which Antigravity surface should I use? 2.0 vs CLI vs IDE vs SDK

Full Walkthrough: Workflow for AI Coding — Matt Pocock

Headroom: A Context Optimization Layer for LLM Applications - Tejas Chopra, Netflix, Inc.

"Software Fundamentals Matter More Than Ever" — Matt Pocock

Agile Scrum is dead - Hypervelocity Engineering with AI is here

Graphify + Antigravity Is Actually INSANE (Talk to Any Codebase)

How To De-Slop A Codebase Ruined By AI (with one skill)

