Volgoda Treasure Hunt

"Vulgard Treasure Hunt" utilizes zero-knowledge proof and multi-party computation in a partially hidden, trustless gaming environment. Players aim to find treasures within 50 steps using provided hints.

A fresh game round involves an externally generated map matrix, which is then fed into the circuit to generate a proof. This proof is placed into the game contract for that round. To prevent local simulation exploitation by malicious players, every step is verified via our snark server and the result returned.

MUD provides an exceptional infrastructure for Ethereum dApps. Yet during our game development, we found handling incomplete information, an essential aspect not currently catered to by MUD.

The challenges are:

1. It's tough to unify the incomplete information framework managed by zk: a process involving abstracting concealed data into an argument, then into a circuit.

2. zk's verification occurs on the local client (browser), consuming considerable local memory.

Our ZK module intends to:

1. Provide a comprehensive zk framework addressing most on-chain games' circuit abstraction needs.

2. Introduce a distinct computing layer to offload zk computation from players, return results via this layer, and potentially amalgamate zk proofs using recursive zk algorithms.

This project employs Circom and Snarkjs for circuit formulation and zk-proof compilation. Additionally, we plan to leverage the MUD architecture for creating the comprehensive game client. We will also deploy a dedicated ZK Snark server to manage all proof verification tasks.

In developing fully on-chain games, we found that maintaining the randomness of the initial map and generating map proofs (to ensure the function of incomplete information) were crucial aspects of the process. This approach allows us to balance the game's fairness and player experience effectively.