Skip to content

Latest commit

 

History

History
84 lines (59 loc) · 3.67 KB

README.md

File metadata and controls

84 lines (59 loc) · 3.67 KB
Raw

Proof of Twitter: SP1

Prove ownership of a X (Twitter) account using an email from Twitter. Built on the SP1 zkVM from Succinct. alt text

How it works

You can use a Password Reset email from Twitter to generate a ZK proof that you own the Twitter acocunt @username. Verifying other types emails are possible with SP1 - the password reset email was chosen because of it's simplicity and lack of user generated content.

Running Locally

Prerequisites

  • node.js
  • yarn
  • rust
    • after installing, switch to the nightly channel: rustup default nightly
  • sp1
  • a c linker for compiling rust. not sure? run sudo apt-get update && sudo apt install build-essential
  • libssl-dev: sudo apt-get install libssl-dev
  • pkg-config: sudo apt-get install pkg-config

Getting Started

  1. Clone and open repo
  2. Run ./start.sh
  3. Open the frontend by going to http://localhost:5173/ in your browser

If you're running this on a remote server, you may have to forward ports 5173 (frontend) and 8000 (backend) when ssh'ing in.

  • ssh -L 5173:localhost:5173 -L 8000:localhost:8000 username@host_ip_address

Reccomended Specs

64GB+ of ram

Also tested on a 36GB M3 Max Macbook Pro

Supported Platforms

Linux/MacOS

Future Prospects

This project is a scrappy proof-of-concept to demonstrate what is possible with ZKPs. Under the hood, there are a few optimizations that could make proof generation even faster:

  • RSA Precompile: Currently, email signatures are verified manually with rust's rsa libray. Succinct has a precompile system of popular functions such as sha256, ed25519, and more. Precompiles speed up excecution by orders of magnitude. A RSA precompile would make proofs of email signatures far more efficient.
  • Regex pre-indexing: The Regex queries search the entire email for labels such as the twitter handle and email sender (x.com). The indices of these labels could be pre-computed and passed into the program, saving many compute cycles that were just spent on searching.

Examples

To try examples, paste contents of .eml file into app

Base Case
  • A valid password reset email from Twitter
  • Result: Generates valid proof that passes verification
  • Base Case.eml
Non-Twitter
  • A valid email not from Twitter
  • Fails at email sender check inside zkproof
  • Result: Generates proof that fails verification
  • Non Twitter.eml
Invalid Signature
  • Same email as Base Case except first character of signature is changed from D to E
  • Fails while verifying RSA signature inside zkproof
  • Result: Generates proof that fails verification
  • Invalid Signature.eml
Not PW Reset Email
  • A valid email from twitter that's not about password resetting
  • Fails at email type check inside zkproof
  • Result: Generates proof that fails verification
  • Not PW Reset Email.eml