zkMinaOnramp - zkMinaFiatBridge

This repository contains a proof-of-concept of a fiat(USD)-MINA bridge that aims to be off-chain identity-hiding (the on-ramper) and decentralized.

It was created during the ZK Hack Krakow (2024) hackathon for the "proto-kit track" (https://protokit.dev/) https://zk-hack-krakow.devfolio.co/

The project is not fully operational as of hackathon submission time - we ran out of time before applying all the polish and required functionalities.

IMPORTANT for reviewers The project submission is divided into two branches:

proofs2 - that contains more mocked-out proofs that work with the UI
proofs-signature - that contains better proofs, but the UI is not compatible yet.

The goal idea overview

The goal idea is to provide a decentralized (or decentralizable) platform to create p2p orders that trustlessly allow exchange between FIAT and MINA without revealing token buyer off-chain identity. To be expanded.

Figure 1: Representation of User Roles

Figure 2: User Types

The intended proof flow

Or: How to verify the off-chain payments

Make a payment using the gateway with an agreed upon sender_id, receiver_id, and amount
Receive DKIM signed message from the payment domain
Build the proof by selecting and locking the order and providing the .eml file

The circuit for the proof is built with zk-email-o1js , o1js-rsa, zk-regex. The proof verifies against the public data of the order details that in essence is a hash of all the necessary information.

What was accomplished

We were able to prepare a custom Protokit branch that works with the new o1js version and successfully build a full-stack app on top of it
The design of the system allows for necessary interactions.
The almost finished implementation of the system logic.
Next.js web app being the user entry point - it already provides a big chunk of the
The sequencer correctly implements the system logic (minus the zkEmail type proof verification) - it now assumes that data signed by the on-ramper is truthful.
The suite of tests that test out the logic of the system.

What was not accomplished

The essential part of verifying the e-mail receipt - the tools we meant to use are in alpha-level functionality and we ran out of time before preparing that part.
The proto-kit sequencer is now the only entity holding the off-chain state of the system which means that is not yet trustless.
The system does not yet use L1 bridge - instead of bridging with L1 the tokens are minted and burned.
UI is not finished
- it doesn't provide enough information, smooth UX, and necessary feedback.
- it doesn't work with the newest circuits we prepared
The code quality suffered heavily because of the pace of the prototyping

Summary

We have a valid proof-of-concept with essential, but well-defined gaps, that can easily (but with a bigger time budget) be filled with attainable pieces. We showed that MINA, o1js, and proto-kit provide an innovative platform for rapid prototyping of full-stack ZK-based apps and that the ecosystem is still in its early stage.

The potential road ahead

How to run the project

The monorepo contains 2 packages and 1 app:

packages/chain contains the implementation of the proto-kit app-chain including runtime modules (something akin to "smart contracts")
packages/proofs contains the implementation of the zk program for proving and verifier the authenticity of the payment receipt
apps/web contains a demo UI that connects to a wallet and the locally hosted app-chain sequencer

Prerequisites:

Node.js v18
pnpm
nvm

Setup

# ensures you have the right node.js version
nvm use
pnpm run setup_protokit # we need to prepare custom framework version
pnpm install
pnpm run build # this may hang during the web app compilation you may stop it

The tests

The tests for the runtime modules are defined in the chain package

cd packages/chain
pnpm run test

The tests may be seen as a partial demo of the functionality.

Running the sequencer & UI

# starts both UI and sequencer locally
pnpm dev
# then visit localhost:3000

adamczykm / zk-mina-onramp