Integration

Gemini / Antigravity Integration: Run Your Validated Playbook Through Google's Coding Agent

Stage 9 of the ShipFit playbook ships your validated decisions to Antigravity (Google's coding agent IDE) so the agent works from your buyer, V1 scope, pricing, and launch plan from the first prompt.

What Antigravity / Gemini does

Antigravity is Google’s coding agent IDE, built around Gemini. It lives somewhere between Cursor (IDE-native) and Claude Code (terminal-first): GUI but more agentic, willing to ship multi-file diffs and run long-form workflows in one shot.

Antigravity reads project-level context on session start and uses it as system prompt for every Gemini conversation. Gemini’s 1M+ token context window means you can load the entire validated playbook without worrying about cost or context budget.

What ShipFit ships to Antigravity

When you reach stage 9 (What to Export?) and pick Gemini / Antigravity, the export takes the validated playbook produced by the previous 8 stages and lands it in your repo as Antigravity configuration:

  1. Buyer persona from stage 2 (Who Pays?).
  2. Above-the-line pains from stage 3 (What Hurts?).
  3. Winning angle from stage 4 (How to Win?), tied to the 7 Powers call.
  4. V1 scope from stage 5 (What’s V1?). Differentiator + Operational; Delights held back for V2.
  5. Pricing model from stage 6 (How to Charge?).
  6. Launch plan from stage 8 (How to Launch?).

ShipFit also installs project-level shortcuts that re-apply the framework gates during development, so the 1M-token context window is paired with actual scope discipline.

Install

After running stage 9, unzip the export into your repo root and commit. The next Antigravity / Gemini session loads the new context.

Why bother

Gemini’s strength is the context window. You can paste the entire validated playbook (and a chunk of your codebase) into a single prompt and ask the model to reason across both. The export makes that automatic: the playbook is always in context, you don’t pay for the paste, and the agent’s responses cascade from your validated decisions instead of guesswork.

The big-context-window advantage compounds across long agent sessions. Where Cursor / Claude Code have to be more strategic about which context to load, Antigravity can keep all of it loaded for the lifetime of the project.

Common mistakes

1. Treating the export as a one-time setup. Re-export after any meaningful pivot.

2. Confusing the 1M context window with scope discipline. A loaded context doesn’t enforce scope on its own; the framework gates that ship with the export do.

Further reading

Frequently asked questions

What is Antigravity?
Antigravity is Google's coding agent IDE, built around Gemini. Agentic workflow similar to Cascade or Claude Code: the agent reads your repo, writes code, runs tests, commits. It's newer than Cursor or Windsurf but the model under the hood is Gemini, which has a generous context window (1M+ tokens) that makes loading the entire validated playbook into context essentially free.
Does this work with regular Gemini (not Antigravity)?
Yes. The same project context format is also accepted by Gemini's coding mode from any Gemini-integrated environment (Google's web Gemini, Gemini in VS Code, etc.). Antigravity is the most agentic surface, but the same export works across the Gemini family.
Do I need a paid Gemini plan?
Antigravity is currently free during preview; expect a paid tier at GA. Gemini (the model) is available on Google AI Studio's free tier with generous limits. The ShipFit export uses no separate API.
How does this compare to the Cursor or Claude Code export?
Same validated playbook, different tool. Pick all the tools you want in stage 9 and ShipFit emits the right configuration for each; they share the same source-of-truth so the tools agree about V1 scope.

Ready to make your next product a success?

9 decisions between your idea and a product worth building.

No credit card required.

Try an example: