In WP1 we created a list of existing algorithms and tools for anonymous proofs in electronic identities. The findings of WP1 were used to create an overview of existing libraries which we can use to build our research on. This post describes how we evaluated the relevant libraries we found, and for which ones we did a deeper evaluation.

The Challenge

The E-ID open-source ecosystem faces significant challenges: competing standards that scatter the community efforts, rapid evolution of the field, and a high complexity of the algorithms. This makes it difficult for new contributors to join existing projects, leaves many libraries abandoned or poorly maintained, and creates uncertainty for potential adopters!

Our Evaluation Approach

We focused our search on the implementation of the cryptographic primitives identified in WP1, as this is the area we want to contribute the most.

To systematically evaluate the available libraries, we established a set of criteria based on the evaluation framework we developed for comparing cryptographic libraries. Each library was assessed across six key dimensions:

1. Software Structure: Code quality, architectural decisions, and engineering practices
2. Testing & CI/CD: Test coverage and continuous integration/deployment setup
3. License, Ownership and Access: Licensing clarity and accessibility, with preference for permissive open-source licenses
4. Amenity to Change (Issues, PRs): How actively maintained the project is and responsiveness to issues and pull requests
5. Documentation: Quality and completeness of documentation, including tutorials and API references
6. Probable Support on the Next 2-3 Years: Likelihood of continued maintenance and support

For a complete overview of our evaluation methodology and all libraries we considered, see our WP2-Libraries repository.

Why we Chose Docknetwork

While some libraries scored higher overall in our evaluation (notably Microsoft’s Crescent, which received the highest quality score), we ultimately chose Dock Network’s crypto library (github.com/docknetwork/crypto) because it best matched our functional requirements. The highest-ranked libraries, while excellent in their own right, did not provide the specific combination of cryptographic primitives we needed for our research on E-ID systems. Dock’s library offered the right balance of functionality, focus, and architectural quality for our use case.

Docknectwork/crypto

The library was created and open-sourced by dock.io, a digital identity startup, and is currently maintained by their lead cryptographer Lovesh Harchandani.

Supported Primitives

The codebase is well-engineered, with each primitive living in its own Rust crate and code references to academic papers. It supports both backend servers and WebAssembly for browsers, making it suitable for a wide range of deployment scenarios.

Category	Primitives
Signatures	BBS+, BBS, PS, group signatures
Credentials	Coconut, KVAC, Delegatable Credentials
Proofs	Schnorr PoK, Sigma protocols, LegoGroth16, Bulletproofs++, Verifiable encryption
Accumulators	VB dynamic accumulators
Misc	Secret sharing, DKG

Library Architecture Reference

The Rust crates follow a layered structure. Lower levels provide primitives. Higher levels build protocols. The top-level proof_system crate ties everything together for composite proofs.

graph TD
    proof_system["Proof System"]
    
    %% the arrows -.-> are used to keep the nodes inside each group horizontally.
    subgraph L3["Proof Of Knowledge (PoK)"]
        bulletproofs_plus_plus["bulletproofs_plus_plus"] -.-> legogroth16["legogroth16"] -.-> verifiable_encryption["verifiable_encryption"]
        smc_range_proof["smc_range_proof"] -.-> vb_accumulator["vb_accumulator"] -.-> equality_across_groups["equality_across_groups"]
    end
    
    subgraph L2["Anonymous Credentials"]
        bbs_plus["bbs_plus"] -.-> coconut["coconut"] -.-> syra["syra"] -.-> delegatable_credentials["delegatable_credentials"]
    end
    
    subgraph L1["Cryptography"]
        oblivious_transfer["oblivious_transfer"] -.-> secret_sharing_and_dkg["secret_sharing_and_dkg"] -.-> kvac["kvac"] -.-> short_group_sig["short_group_sig"]
    end
    
    subgraph L0["PoK Foundation"]
        direction LR
        merlin["merlin"] -.-> schnorr_pok["schnorr_pok"]
    end
    
    proof_system --> L3
    proof_system --> L2

    L3 --> L1
    L3 --> L0

    L2 --> L1
    L2 --> L0 
        
    %% Styling for better visual hierarchy
    classDef level0 fill:#e1f5ff,stroke:#01579b,stroke-width:2px
    classDef level1 fill:#fff4e1,stroke:#e65100,stroke-width:2px
    classDef level2 fill:#e8f5e9,stroke:#2e7d32,stroke-width:2px
    classDef level3 fill:#fce4ec,stroke:#c2185b,stroke-width:2px
    classDef topLevel fill:#fff9c4,stroke:#f57f17,stroke-width:3px
    
    class merlin,schnorr_pok level0
    class oblivious_transfer,secret_sharing_and_dkg,kvac,short_group_sig level1
    class bbs_plus,coconut,syra,delegatable_credentials level2
    class bulletproofs_plus_plus,legogroth16,verifiable_encryption,smc_range_proof,vb_accumulator,equality_across_groups level3
    class proof_system topLevel
    
    %% Hide horizontal arrows (links 0-10)
    linkStyle 0,1,2,3,4,5,6,7,8,9,10 stroke:transparent,stroke-width:0px

Project Structure

The library spans three repositories:

1. Crypto (Rust): Core cryptographic logic
   → github.com/docknetwork/crypto

2. Crypto-Wasm (Rust/TS): WebAssembly wrapper for browser use
   → github.com/docknetwork/crypto-wasm

3. Crypto-Wasm-TS (TypeScript): High-level APIs for web developers
   → github.com/docknetwork/crypto-wasm-ts

Our Contributions

Like most other libraries, this one lacks contributors. However, it’s in a very good state, almost feature-complete, and production-ready.

This is why we decided to contribute as much as possible. Over the past year, we’ve made several Pull Requests to add missing documentation and examples [1] [2] [3] [4] [5],

and updated the TypeScript project to the latest Rust counterpart [6] [7] [8] [9]. We were also happy to do some housekeeping and update the dependencies [10] [11] [12].

Challenges

During this endeavor, we’ve encountered a couple of challenges when contributing to this library:

• Complicated workflow: Any change in the API will require changes to 3 repositories to propagate from Rust to the user-facing TypeScript library.
• Small team: Currently, the library only has one maintainer. A great maintainer, but taking the bus factor into account, it’s not ideal.
• Sparse documentation: Tutorials and guides are lacking. In such a highly technical field, documentation and tutorials need to be kept up-to-date.
• Evolving standards: Needs to track IETF BBS standardization

Conclusion

Dock’s crypto library is the most complete option we’ve seen for privacy-preserving credentials. While it doesn’t cover all we need for our research on e-ID, it serves as an excellent comparison point and a solid starting point for our work. It’s well-structured and covers all we needed to build an E-ID system. The main challenge is its small community. Broader adoption would help ensure long-term maintenance.

We’ve contributed documentation, updates, and integration examples back to the project.

While finalizing our proof-of-concept with Dock Network’s libraries, a new domain-specific language (DSL) caught our attention. Noir: a DSL developed by the Aztec Foundation, specifically targeting the development of privacy-preserving ZK applications. Noir’s main audience is Blockchain developers, but it can also be used in e-ID and other privacy-preserving applications which are not based on a blockchain.

We’re currently wrapping up a second proof-of-concept using Noir, and the early results are very promising. The developer experience feels fundamentally different—more intuitive and enabling significantly faster iteration.

You can find our findings, comparisons, and insights in our post.