WhaleChess

WhaleChess is a fully on-chain implementation of chess that also includes custom pieces and play-to-earn mechanics. Players accumulate new pieces as they play, which are generated by having an LLM service interact with our contracts. Players can then build new squads from the pieces in their collection and battle them for glory.

Future plans include matchmaking, concealed information via zk, a marketplace for pieces, reward-earning mechanisms, and improvements to our piece generation system including creating agents responsible for deploying entirely new kinds of behaviors and mechanics for the custom pieces.

WhaleChess is built by jmill & Lucy.

We used Solidity & the MUD framework to build the game contracts.

The app frontend is made with React, and utilizes MUD's react library + zustand for state management + viem for interactions with the contracts. The app uses Dynamic for Authentication and social logins. We also use Blockscout for exploring transactions the app creates.

There is an API server built with express that currently holds the logic for generating the custom pieces and calling the transactions which create them. This uses OpenAI API and Phala network LLM inference.

Figma, eden.art, OpenAI, and various other tools were used for image generation and design.

The contracts are deployed to a number of EVM testnets including Flow, Base, Mantle, Inco, Linea, Polygon, Garnet, and Scroll.

The live web app is deployed with Render and configured to use the contracts on the Garnet testnet.