blobstream-ops

This repo contains a set of Blobstream related tooling.

Install

1. Install Go 1.22
2. Clone this repo
3. Install the Blobstream-ops CLI

make install

Usage

# Print help
blobstream-ops --help

BlobstreamX contract verification

One of the tools that the blobstream-ops CLI provides is the verify subcommand. It allows verifying that a BlobstreamX deployment is valid. It works by taking a trusted Celestia RPC endpoints. Then, it goes over the data root tuple roots stored in the contract, and regenerates the data commitment using the provided trusted RPC endpoint. Finally, it compares them and errors if any of the data root tuple roots is not valid.

Since Blobstream makes a 2/3rd honest validator set assumption, if more than 2/3rd of the validator set is malicious, they can commit to an invalid data root tuple root. This tool allows verifying that that didn't happen using a trusted RPC endpoint.

Verification command usage

After installing the blobstream-ops CLI, you can verify the contract using the following subcommand:

blobstream-ops verify contract <args>

The arguments can be provided either through the CLI, or more easily using environment variables.

The .env.verify.example contains an example set of how to provide the environment variables required for the verification.

First, populate the empty fields with the correct values. Then, run:

set -a
source .env

given that .env is the file created by populating the .env.verify.example file.

Then, run the CLI using:

blobstream-ops verify contract

And you should see the contract verification underway.

Blobstream proofs replay

The replay command allows replaying proofs from an existing BlobstreamX deployment to a new one, in a different chain, without having to regenerate them. This reduces the cost of maintaining the BlobstreamX deployment while keeping the same security properties.

Pros

• Cheaper deployment
• Porting the BlobstreamX deployment to any new chain without having to run an operator

Cons

• The ranges of blocks proven by proof will be the same between the existing and the new deployment
• Requires the contract to be initialized to a trusted header that is the start_block or any event of the existing deployment
• Adds dependency on the existing deployment. If it goes down, the new deployment also goes down.

Requirements

To use the replay command, the whole BlobstreamX stack needs to be already deployed on the new chain. Refer to docs.

Also, make sure the trusted block used to initialise the BlobstreamX contract corresponds to a start_block in the existing BlobstreamX deployment. Otherwise, the proofs will not be able to be relayed.

Replay command usage

After installing the blobstream-ops CLI, you can use the replay tool using the following subcommand:

blobstream-ops replay --help

Similar to the above verify command, the arguments can be provided either using the CLI or through environment variables.

To this matter, the .env.replay.example provides an example set of environment variables that need to be set in order for the replay command to run.

First, populate the empty fields with the correct values. Then, run:

set -a
source .env

given that .env is the file created by populating the .env.replay.example file.

Then, run the CLI using:

blobstream-ops replay

And you should see the proofs being queried from the existing deployment and replayed in the new one.

Contributing

Tools

1. Install golangci-lint
2. Install markdownlint

Helpful Commands

# Build a new blobstream-ops binary and output to build/blobstream-ops
make build

# Run tests
make test

# Format code with linters (this assumes golangci-lint and markdownlint are installed)
make fmt

Useful links

The Blobstream documentation is in docs.

The smart contract implementation is in blobstream-contracts.

The BlobstreamX implementation is in BlobstreamX repo.

Blobstream ADRs are in the docs.

Blobstream design explained in this blog.