Skip to content

A clone of the game "Where in the World is Carmen Sandiego?" largely re-created with generative AI.

License

Notifications You must be signed in to change notification settings

fmaclen/julia-sanfrancisco

Repository files navigation

Julia Sanfrancisco

Chase Julia and her accomplices across different countries, unraveling clues, solving puzzles, and expanding your geography knowledge in this exciting detective game.

🕹️ Play it now — 💻 Devlog on Discord & Twitter

Collage of artwork from the game


About the project

Julia Sanfrancisco is a clone of the 80s MS-DOS game Where in the World is Carmen Sandiego?. This project is meant as a proof-of-concept for leveraging generative AI to create the game and it's assets.

Short-term roadmap

  • Implement the game mechanics from the original game
  • Generate updated clues for the cities featured in the original game
  • Generate new artwork for the cities featured in the original game
  • Implement a new mobile-based UI with improved UX

Follow the current progress here.


Development

Current stack

  • Written in Typescript using SvelteKit.
  • There is no backend, the game runs entirely on the browser.
  • Your name, score and current play session is stored locally on your device.
  • Clues have been generated with OpenAI's GPT-4.
  • Artwork was generated with Stable Diffusion using deliberate_v2.

Contribute

The game is currently open for contributions. If you want to help, please join the Discord server and/or check the issues for tasks that need to be done.

Environment setup

To run the development environment locally:

  • Clone the repository.
  • Copy .env.example and rename it to .env.
  • Then run:
npm install

npm run start

To create a production version of the game:

npm run build

References