Home > eID > Library-Evaluation
Library Evaluation¶
Last modified: 2026-10-28
Description¶
In 2025 the eid-privacy project (WP2) rated open-source libraries for e-ID and anonymous credentials, to pick the basis for its proofs of concept (eid-privacy-WP2-Libraries). Each library got a score from 1 to 5 on six criteria; the mean is its quality score. The scores show the state in mid-2025 and are a quick assessment, not an audit. The project then used docknetwork-crypto for BBS + Bulletproofs and Noir for circuits (GHE26-MS2).
Criteria¶
Each criterion is rated 5 (excellent), 4 (good), 3 (adequate), 2 (poor) or 1 (very poor / missing).
| Criterion | Short | 5 means | 1 means |
|---|---|---|---|
| Software structure / architecture / engineering | Str | Well organised, clear separation of concerns, modular and extensible | No clear architecture, very hard to understand or maintain |
| Unit tests & CI/CD | Test | Coverage > 80 %, integration tests, CI on several platforms (4: 60–80 %, 3: 30–60 %, 2: < 30 %) | No tests, no CI |
| License, ownership and access | Lic | Permissive license (MIT, Apache 2.0, BSD), clear contributor rules, no proprietary parts (4: copyleft) | Closed source, very restrictive or contested ownership |
| Amenity to change (issues, PRs) | Chg | Issues triaged within days, PRs reviewed and merged quickly (4: weeks, 3: months) | Abandoned, maintainers do not answer |
| Documentation | Doc | Tutorials, API reference, examples, architecture diagrams, up to date | Only a basic README |
| Probable support in the next 2–3 years | Sup | Backed by a large organisation, active community, roadmap, funding | Abandoned or on life support |
Scores¶
Zero-knowledge libraries¶
| Library | Language | Purpose | Str | Test | Lic | Chg | Doc | Sup | Mean | Note |
|---|---|---|---|---|---|---|---|---|---|---|
| Arkworks Algebra | Rust | Low-level zkSNARK building blocks | 5 | 5 | 5 | 5 | 5 | 5 | 5.0 | Probably the best engineered and best supported project in this list |
| Crescent | Rust | e-ID ZKP library | 5 | 3 | 5 | 5 | 3 | 5 | 4.3 | Full e-ID solution with a good client and examples |
| Arkworks crypto-primitives | Rust | Crypto primitives | 3 | 5 | 5 | 5 | 2 | 5 | 4.2 | Bad documentation, not clear how to use it |
| Plonky3 | Rust | Polynomial-IOP primitives for STARK-based zkVMs | 3 | 3 | 5 | 5 | 2 | 5 | 3.8 | Great community, but weak documentation and a high entry barrier |
| Spartan | Rust | Microsoft's Spartan proof system | 5 | 3 | 5 | 5 | 3 | 2 | 3.8 | |
| Docknetwork crypto | Rust | Signatures, ZKPs and high-level constructs | 5 | 3 | 3 | 3 | 2 | 5 | 3.5 | One of the most complete e-ID focused crypto libraries we know of |
| Longfellow | C++ | e-ID ZKP library | 5 | 3 | 3 | 3 | 2 | 5 | 3.5 | C++ without foreign-language wrappers, unclear APIs |
| MATTR pairing_crypto | Rust | BBS signatures | 2 | 5 | 5 | 3 | 3 | 2 | 3.3 | Bad documentation, but wrappers for several languages |
| zkcrypto Bulletproofs | Rust | Fast Bulletproofs implementation | 5 | 3 | 5 | 2 | 3 | 2 | 3.3 | Only Bulletproofs; a fork of dalek-cryptography/bulletproofs |
| Plonky2 | Rust | Polygon's ZK proving system | 2 | 2 | 5 | 2 | 5 | 2 | 3.0 | On its way to deprecation (replaced by Plonky3) |
Zero-knowledge circuit frameworks¶
| Library | Language | Purpose | Str | Test | Lic | Chg | Doc | Sup | Mean | Note |
|---|---|---|---|---|---|---|---|---|---|---|
| Circom | Rust / WebAssembly | Most popular circuit framework | 5 | 5 | 5 | 5 | 5 | 5 | 5.0 | |
| libiop | C++ | Library for IOP-based zkSNARKs | 5 | 5 | 5 | 3 | 5 | 3 | 4.3 | Could be a basis for IOP-based constructions |
| Halo2 (Zcash) | Rust | Halo2 proof system and circuit framework | 3 | 5 | 5 | 3 | 3 | 3 | 3.7 | |
| Noir | Rust | Rust-like DSL for ZK circuits | 2 | 2 | 5 | 3 | 2 | 2 | 2.7 | Huge project with a big community |
DIDComm libraries¶
| Library | Language | Purpose | Str | Test | Lic | Chg | Doc | Sup | Mean | Note |
|---|---|---|---|---|---|---|---|---|---|---|
| ACA-Py | Python | SSI agent | 5 | 5 | 5 | 5 | 5 | 5 | 5.0 | |
| DIDComm Mediator Credo | TypeScript | DIDComm mediator for message routing and delivery | 5 | 5 | 5 | 5 | 5 | 5 | 5.0 | |
| Veramo | TypeScript | Framework for verifiable data and DIDComm messaging | 5 | 5 | 5 | 5 | 5 | 5 | 5.0 | |
| didcomm-rust | Rust | DIDComm v2 messaging | 3 | 5 | 5 | 3 | 3 | 3 | 3.7 |
Verifiable credential format libraries¶
| Library | Language | Purpose | Str | Test | Lic | Chg | Doc | Sup | Mean | Note |
|---|---|---|---|---|---|---|---|---|---|---|
| SpruceID ssi | Rust | VC data model, among other things | 5 | 5 | 5 | 5 | 5 | 5 | 5.0 | Maybe the best in this category, needs more research |
| Procivis One Core | Rust | All-in-one application, not specific to VC formats | 3 | 3 | 5 | 2 | 5 | 3 | 3.5 | |
| did-jwt | TypeScript | Simple DID-JWT implementation | 5 | 2 | 3 | 2 | 5 | 2 | 3.2 |
Crypto and communication libraries are mature, but there are few dedicated VC-format libraries: the format is usually a small part of a bigger project (for example, Longfellow parses CBOR itself).
Related¶
Resources¶
- Docs: eid-privacy-WP2-Libraries
- Paper: GHE26-MS2