Skip to content

For people with a game in mind

Build, verify,
and evolve your
game with AI.

Your idea deserves more than a promising first build. We're building Starward to help creators, experts, and AI work from clear goals, check what works, and learn from real play.

↳ Join the waitlist for early access. Initial pilots are invitation-only.

The work, in practice / 001
Real Panda Wonderland gameplay: animal tiles in a forest puzzle scene
Panda WonderlandGameplay

01 / Intent

Start with the game
you want to make.
Ideas, rules & human judgment

03 / Evidence

What changed?
What actually works?
Results you can review
Make. Check. Learn.

Real game imagery · Workflow illustration

For creators with a clear idea and limited technical resources.
A practical path toward something people can play.

Your direction / Expert judgment / AI execution

From an idea to the next better version.

A game keeps changing. The workflow we're building connects each change to the rules behind it, evidence you can review, and what happens when people play.

  1. 01

    Idea & rules

    Define the game, its rules, and what a good result should do.

  2. 02

    Build

    Turn an agreed scope into a working change with AI agents.

  3. 03

    Verify

    Check behavior and review the evidence before moving on.

  4. 04

    Play

    Put a reviewed build in the hands of invited players.

  5. 05

    Observe

    Connect player experiences to what actually happened in the game.

  6. 06

    Improve

    Use what you learn to choose and verify the next change.

Our first milestone: a verified build for invited players. People review the work and decide when to release.

Meet Starward Assembler.

Our main platform product brings game creation, AI-assisted changes, and verification into a shared workspace. See an early project-assembly workflow in this recorded prototype demo.

These recordings show prototype workflows. Assembler is in development; access is invitation-only.

Panda Wonderland's existing puzzle game in playA game to learn from

A real project, and an ongoing experiment

Building the foundation. Learning from Panda.

Working solo with AI, Hengstar designed and developed Starward Game Services from the ground up on open-source foundations. He then migrated Panda Wonderland, his existing commercial game, onto it. Building the backend was the central work; Panda gives it a real setting to test assumptions and discover what needs to improve.

I need to define what good looks like, review the result, and be able to follow what happened when something goes wrong.

Hengstar Gong · From the published account of building Starward Game Services

Read the story

Make progress you can understand.

Clear rules before more code

Give people and agents a shared understanding of the game and what each change needs to achieve.

Evidence you can review

Bring changes and verification results together so you can judge what works and what still needs attention.

Feedback from real play

Use evidence from a running game to understand problems and guide the next fix, test, or improvement.

Meet the Assembler workbench

Early access starts here

Have a game in mind?

Join the Starward waitlist for updates and early access. We're starting with invitation-only pilots for creators making a new game. Joining the list does not grant immediate platform access.

Early access
Join the waitlist. Initial pilots are invitation-only.
First milestone
A verified build for invited-player testing.
Support
Initial invited trial is free. Extra human support is scoped and quoted separately.
Your team
Bring your ideas and judgment. A technical partner is optional.

Join the Waitlist

Tell us what you'd like to make.

Start with a build. Grow toward the whole lifecycle.

Our long-term goal is to automate development, verification, release, and operations within goals, budgets, and permissions you approve. Expert agents would help along the way, with exceptions and major decisions returned to people. We're building toward that future one reviewable step at a time.

Hengstar GongFounder, Starward Game Studios

More time shaping the game.

I'm Hengstar, Starward's founder and a former Meta Staff Engineer. Building my own games with AI keeps bringing me back to the same question: how can I spend more time deciding what the game should become, and less time rebuilding context and checking disconnected pieces? Starward is how I'm working through that question, in practice.

More about Starward

Built from making games ourselves.

Panda WonderlandPuzzle · Existing game using SGS
Cyber ImposterSocial deduction · Studio game