SHIELD; A Swiss Army Knife for ZKPs Development

Background:

At Xord, the R&D team has always expanded their knowledge set with diverse concepts from privacy to scalable solutions.

In the Q2 of 2022, the team decided to enroll in the cohort of Harmony’s zkDAO. The goal was to learn about the circom development.

The team graduated from the cohort after submitting two successful projects.

1. ZKPT (Zero-knowledge pool together)
2. SPIC (Semaphore protected incentivized community)

In this cohort, we learned about different zkps projects and library toolsets available for different use cases.

While building these projects, the team realized that the developer experience of circom is raw and can be made better. That’s when the idea of SHIELD came forward.

Vision:

Circom development requires a set of pre-requisite skills. From both technical and mathematical perspectives. But what if we make these pre-requisite skills acquisition easy, improve the developer experience, and make it attractive to increase the mass adoption of zkps?

These developer experience products would be focusing on the following area:

1. Testing suite and efficient debuggers
2. Optimized boilerplates
3. Modular libraries for compilation
4. Supporting more language compatibility, including circom, cairo, snarkyjs, and zokrates

This tool would not focus on the libraries and the core building blocks of zkps such as arkworks. Instead, it would provide extensibility and modularity to the DSL of the zkps, which are either used for building circuits or building applications over a zk-rollup-based blockchain or l2.

Execution:

At the start of July, the execution of SHIELD began, and we didn’t impulsively start the development and planning out MVP. We followed the GO LEAN framework.

Running the GO LEAN and processing all the steps from problem fit to market fit, we laid down very core features that we included in our MVP.

The very first feature we release was the code “boilerplate for circom” and “shield init”.

This URL will provide more information.

For our MVP, we are targeting one feature, “to display the stack trace of circom compilation and error logging”.

Future:

Execution of SHIELD has taught us how to apply product development framework and methods to a developer tool. The developer bias we have states that all the processes of taking the survey, asking questions, and documenting things are a waste of time. But in reality, it is the effort you put in today so that you can work clearly down the road.

Having a clear vision brings a good product forward. As mentioned above, we are not limiting ourselves to circom. We will be expanding our toolset to other DSL.

We will soon be submitting our grant applications to different platforms.

For updates, make sure to follow SHIELD on Twitter.