Warning
This library is alpha software under active development. Breaking changes are expected, APIs are not yet stable, and no official support is provided. Early adopters are encouraged to open issues for bugs and feedback to help shape the project's direction.
rblib is a high-performance, modular Rust SDK for constructing Ethereum-compatible block builders. Built on top of the Reth execution engine, it provides robust, platform-agnostic primitives and declarative workflows designed for both L1 and L2 applications.
rblib
is designed for sophisticated engineering teams who require granular control and composability in their block-building logic. Our philosophy centers on three key principles:
- Modularity: Core components are decoupled. Use the low-level Payload API for fine-grained control, or compose high-level Pipelines for declarative workflows.
- Performance: By leveraging Reth and a zero-copy design for payload state,
rblib
is built for high-throughput, latency-sensitive environments. - Extensibility: A clean
Platform
trait abstraction allows for seamless extension to support custom EVM-based chains (L2s, app-chains) without forking the core logic.
- Composable Payload API: A flexible, low-level API for constructing and inspecting block payloads via immutable
Checkpoint
transformations. - Declarative Pipelines API: A high-level, composable system for defining block-building workflows (e.g., ordering, revert protection) as reusable
Step
s. - Platform-Agnostic Design: Out-of-the-box support for
Ethereum
andOptimism
, with a clear interface for adding new platforms. - Integrated Testing Framework: A rich test suite with utilities for multi-platform testing, local Reth nodes, and isolated component tests.
Add rblib
to your project's dependencies:
cargo add rblib
rblib
is split into two primary APIs that can be used independently or together.
Located in rblib::payload
(see src/payload
), this API provides the foundational primitives for block construction. It enables exploring many different payload variations efficiently.
A Checkpoint<P>
is an immutable snapshot of a payload's state after a mutation (e.g., applying a transaction or bundle). Checkpoints are cheap to clone and fork, making it trivial to explore alternative block constructions from a common state. Each checkpoint retains the history of its parent, allowing it to act as a DatabaseRef
for the chain state at that specific point. Common algorithms for inspecting checkpoints are available in src/payload/ext/checkpoint.rs
.
Example: Building and Forking a Payload
use rblib::{*, test_utils::*};
let ctx = BlockContext::<Ethereum>::mocked();
// Create a linear history
let checkpoint1 = ctx.apply(tx1)?;
let checkpoint2 = checkpoint1.apply(tx2)?;
// Fork from an earlier state to explore an alternative
let checkpoint3_alt = checkpoint1.apply(tx3)?;
// Compare the outcomes of the two different forks
let gas_main = checkpoint2.cumulative_gas_used();
let gas_alt = checkpoint3_alt.cumulative_gas_used();
let balance_main = checkpoint2.balance_of(coinbase_addr);
let balance_alt = checkpoint3_alt.balance_of(coinbase_addr);
A Span<P>
is a view over a linear sequence of checkpoints, useful for analyzing or manipulating a specific portion of a payload's history.
use rblib::payload::*;
let full_history = checkpoint.history();
let total_gas = full_history.gas_used();
// Analyze a sub-section of the payload
let sub_span = full_history.skip(2).take(4);
let sub_gas = sub_span.gas_used();
Located in rblib::pipelines
(see src/pipelines/
), this API uses the Payload API to create declarative, reusable block-building workflows. It is particularly powerful for L2s, where standardized logic can be composed and customized.
Pipelines are built from Steps (src/pipelines/step.rs
) and control-flow components (src/pipelines/mod.rs
). A rich library of common, reusable steps is provided in src/pipelines/steps/
.
Example: A Minimal Builder Pipeline
This example defines a pipeline that loops over a set of ordering and protection steps before finalizing the block.
use rblib::*;
fn main() {
// Define a reusable pipeline workflow
let pipeline = Pipeline::<Ethereum>::default()
.with_epilogue(BuilderEpilogue)
.with_pipeline(
Loop,
(
AppendOneOrder::default(),
PriorityFeeOrdering,
TotalProfitOrdering,
RevertProtection,
),
);
// Integrate with a Reth node
Cli::parse_args()
.run(|builder, _| async move {
let handle = builder
.with_types::<EthereumNode>()
.with_components(
EthereumNode::components()
.payload(pipeline.into_service()))
.with_add_ons(EthereumAddOns::default())
.launch()
.await?;
handle.wait_for_node_exit().await
})
.unwrap();
}
Chain-specific logic (e.g., transaction types, block validation rules) is abstracted via the Platform
trait (src/platform/mod.rs
). This allows the core building logic to remain generic while providing concrete implementations for Ethereum
and Optimism
.
To support a custom chain, implement the Platform
trait. See examples/custom-platform.rs
for a practical example.
Pipeline::into_service()
is the primary entry point for converting a pipeline into a Reth-compatible PayloadServiceBuilder
. For complete integration examples, see the examples/
directory and the reference builder implementation in bin/flashblocks/src/main.rs
.
The project includes a comprehensive testing infrastructure to ensure reliability.
- Run all tests:
cargo test
- Run a specific test with verbose logging:
TEST_TRACE=on cargo test smoke::all_transactions_included_ethereum
- Multi-Platform Testing: Use
#[rblib_test(Ethereum, Optimism, YourCustomPlatform)]
to run tests across platforms. - Local Test Nodes:
LocalNode<P, C>
(src/test_utils/node.rs
) provides full Reth nodes for integration testing. - Step Testing:
OneStep<P>
(src/test_utils/step.rs
) enables isolated step testing. - Funded Accounts: Use
FundedAccounts::by_address()
or.with_random_funded_signer()
for test transactions. - Mocks: Mocking utilities are available for types like
BlockContext
andPayloadAttributes
.
Contributions are welcome. Please feel free to open an issue to discuss a bug, feature request, or design question. Pull requests should be focused and include relevant tests.
This project is licensed under the MIT License.
Made with ☀️ by the ⚡🤖 collective.