Antigravity Starter is a pristine, AI-first Next.js 16 sandbox that gives your AI editors perfect vision.
Stop wasting tokens, suffering from AI hallucinations, and struggling with messy folder structures. Coding with AI should feel like magic—and now it does.
📦 Free template by andiupn (kuncimu.com) · Licensed under MIT License
☕ If useful, buy me a coffee · 🚀 Need status-first multi-project monorepos? Try the PRO version
AI Agents (Cursor, Gemini, Claude) are phenomenally capable. But when dropped into standard, cluttered directories, they get lost. They read irrelevant folders, blow through your token budget, hallucinate, and place files in the wrong directories because there are no clear boundaries.
We designed every folder, config, and .gitignore rule to act as roadmaps for your AI. The workspace is extremely clean, allowing Cursor or Gemini to understand your entire Next.js 16 + React 19 + Tailwind v4 architecture in less than 3 seconds. Zero wasted tokens, maximum accuracy.
You don't talk to a blank AI box. You get a team of 2 pre-configured, highly specialized, and autonomous AI agents:
@code-reviewer— Your rigorous, strict QA team that reviews code quality, lints, and structure before commits.@research-assistant— A high-speed, cost-effective research agent tailored to explore new API docs and libraries safely.
Execute complex development workflows with simple commands. Equipped with 8 reusable automation workflows (/skills) that turn manual scaffolding, environment checking, and test writing into a 1-second task.
| What You Can Do | 🆓 LITE (Free) | 💎 PRO (Paid) |
|---|---|---|
| Specialized Agents | 2 | 5 (+ planner, architecture, security) |
| Workflow Skills | 8 | 23 (+ Drizzle sync, Docker, rules-audit, etc.) |
| Workspace Structure | Simple (src/, docs/) |
Status-first Monorepo (active/, staging/, etc.) |
| Ops Ready Config | ❌ | ✅ (Docker Compose, Caddyfile) |
| Project Presets | 1 (Next.js) | 4 (Next App, SPA, Laravel, Flutter) |
👉 View Full Feature Comparison & Upgrade Guide
your-workspace/
.gemini/ # AI agent configs, automation skills, and wiki
docs/ # Pristine research and implementation plans
src/ # Main directory for your application source code
.env.example # Template for environment variables
.gitignore # Standard Git ignore file
.mcp.json # Model Context Protocol (MCP) configuration
GEMINI.md # AI coding rules & workspace guidelines
LICENSE # Project license (MIT License)
cp .env.example .env(Insert your GITHUB_TOKEN to enable repository integrations).
cp .gemini/settings.local.example.json .gemini/settings.local.json(Uncomment and set permissions according to your OS).
Open this workspace in your favorite AI-friendly editor, and experience the speed.
If this template has accelerated your coding workflow, consider supporting:
- Ko-fi: ko-fi.com/andiupn
- Patreon: patreon.com/AndiUpn
- Trakteer: trakteer.id/andi_upn/gift
- Saweria: saweria.co/andiupn
This project is licensed under the MIT License. See the LICENSE file for more information.