Google Antigravity Git & GitHub Workflows: AI-Powered Version Control Mastery

Git operations in Google Antigravity are supercharged with Gemini 3. From intelligent commit messages to automated PR reviews, this guide covers everything about AI-powered version control.

Git Integration Overview

Google Antigravity includes:

• Built-in Git source control
• GitHub MCP server integration
• AI-generated commit messages
• Conflict resolution assistance
• PR review automation

Setting Up GitHub MCP

Configure GitHub MCP Server

// .antigravity/mcp.json
{
  "mcpServers": {
    "github": {
      "command": "npx",
      "args": ["@modelcontextprotocol/server-github"],
      "env": {
        "GITHUB_PERSONAL_ACCESS_TOKEN": "${GITHUB_TOKEN}"
      }
    }
  }
}

Create Fine-Grained PAT

For security, use a fine-grained token with minimal permissions:

1. Go to GitHub Settings > Developer Settings
2. Select "Fine-grained tokens"
3. Create token with these permissions:
   • Repository: Read & Write
   • Pull Requests: Read & Write
   • Issues: Read & Write

AI-Powered Commit Messages

Automatic Commit Message Generation

You: "Generate a commit message for my changes"

Gemini 3 analyzes your diff and generates:

feat(auth): implement OAuth2 login with Google provider

- Add GoogleAuthProvider component with sign-in button
- Create useGoogleAuth hook for authentication state
- Implement token refresh mechanism
- Add proper error handling for auth failures
- Update AuthContext to support multiple providers

Closes #42

Commit Message Conventions

Configure your preferred style in GEMINI.md:

# Git Conventions

## Commit Messages
- Format: Conventional Commits
- Types: feat, fix, docs, style, refactor, test, chore
- Include scope in parentheses
- Reference issues with Closes #XX
- Keep subject under 72 characters