Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

What is aries-vcx

Aries-vcx https://github.com/hyperledger/aries-vcx/ is Rust implementation of Aries protocols and collection of supporting projects built on top it:

  • Aries-vcx - the first class citizen Rust crate,library to speak Aries protocols, deal with wallets, interact with ledger

  • Mobile applications building blocks:
    • deprecated approach: libvcx-java, libvcx-ios: these are C binding which enable calling aries-vcx from Objective-C and Java
    • endorsed approach: New approach based on UniFFI is in staged of Proof of Concept. This can generate Swift and Kotlin wrappers for aries-vcx. If you are interestd in mobile development, your contributions here would be invaluable! POC: https://github.com/hyperledger/aries-vcx/pull/737
  • NodeJS: we have NodeJS bindings to build NodeJS applications on top of aries-vcx

How to get involved

We welcome new contributors to improve aries-vcx or build on top of it!

Built on top of aries-vcx

Aries-vcx is generic crate which can be used to build Rust applications and libraries. We strive to be opinion-less to enable arbitrary use-case. Aries-vcx can be used to build web, mobile, cli and potentially embedded agents.
Example of crates built on top of aries-vcx:

  • libvcx - wrapper around aries-vcx to enable certain FFI approach
  • uniffi wrapper (POC phase) - wrapper around aries-vcx to generate Kotlin and Swift mobile wrappers, using uniffi library 
  • aries-vcx-agent - simple agent implementation with in-memory storage
  • aries-vcx-backchannel - simple web agent built on top of aries-vcx-agent, used for cross testing aries-vcx with other Aries implementations in Aries Agent Test Harness 

Project ideas

  • Aries-vcx uniffi wrapper - continue implementation of uniffi_aries_vcx
    • This can unlock building native mobile apps with aries-vcx

  • Aries mediator client - implement rust client capable of connecting with mediation agent and use pick-up protocol https://github.com/hyperledger/aries-rfcs/tree/main/features/0685-pickup-v2 to receive messages
    • This can unlock integration of aries-vcx with different aries mediation agent implementations! This is extremely important for mobile use-cases.

  • Verifier agent - build aries agent specialized for proof verification
    • In the real world, there will be presumably handful of issuers but likely many verifiers! Deploying full featured agent capable of both issuance, verification and perhaps other protocols can be complex, a narrowly focused Verifier agent could be built on aries-vcx . This would require minimal configuration, the agent can be largely stateless as it would only need to READ ledger, not write. Aries-vcx is perfectly positioned to be core of high-performance proof verifier. 

  • Load tester - create tool to simulate many agents simultaneously 
    • Creating tool which can represents many agents can be useful for load/stress testing across the aries ecosystem

  • CLI agent - building on top of aries-vcx-agent and some of many console UI interface frameworks available in Rust, you could build tool to manage an Aries agent from CLI interface
  • Aries message mediator - build aries mediator
    • there's nodejs implementation of aries-mediator-service in progress, but hey, why not to build an alternative!

  • Embedded device aries agent - try to build aries agent for embedded uses case
    • this is yet rather unknown journey, who knows what it entails?


  • DID parser - a proper DID type and parser
    • DIDs, while similar to URLs, have slightly different composition.
    • we need to be able to compose and decompose DIDs in a consistent manner


  • DID resolver - a component that takes a DID and returns the data it points to
    • things such as DIDDocs  or services  are stored on the ledger (or some other place) and need to be retrieved by resolving their DID


  • Protocol Registry - a static, compile-time component that holds the protocols and versions an agent supports
    • the first implementation uses a lazily initialized map
    • a better approach would be to have code generation in place controlled by feature flags

Research ideas