Skip to content

Github Repo Description

Join the chat at License: AGPL v3

COVA Github We wanted to write this to reflect on our overall development and technical thought process and how it has evolved over time. Hopefully, this will also serve as a decent guide to how one should navigate through our (mostly) open sourced code repos on github and contribute in the spirit of making verifiable computing, a new paradigm available to everyone.

Overview of Documentations

When we are working with technologies as novel as verifiable computing, TEE, distributed trust, and scalable blockchain, everything moves fast. As a consequence of that even over a year timeframe, we ended up with three different technical documentation. While they are not contradictory (in fact, they show a very interesting temporal snapshot of how we have matured into designing a usable protocol), they have been somewhat confusing to the community at times regarding which one is the “real” documentation and state of the world that they should follow. Here is an overview of our 3 publicly available technical documentations:

Current Development

Currently, we have 37 Github repos (public and private) and about half of them are under active development. However, the most valuable (and release ready) public Github repos to Look for are:

  1. Cova-Core: Contains core COVA protocol (gossip network), CovaVM, Centrifuge, and dockerfiles to run routing and compute nodes
  2. Python-Library: Application abstraction layer for Data Owners and Data Users to build their dApps
  3. Cova-Blockchain: Persistence layer (currently implemented with go-ethereum fork and tendermint in TEE nodes)
  4. Cova-Enclave: Contains Makefile and bash recipes along with Dockerfiles to setup Graphene, SGX, and CovaClave to run Cova-Core securely

Pre-release and under active development:

  1. Js-Library: A JS port of the python library
  2. dApps: Various example dApps made using COVA

Some older repos (primarily used for the private testnet and were released to private reviewers) we are making available too:

  1. Secure-Model: Initial example of COVA data control policy: enables data users to write machine learning scripts without violating data privacy
  2. Secure-Server: Spawn AWS servers from inside TEE nodes, keeping credentials secret to server owners