What is BlockX?

BlockX is an EVM-native offchain compute layer. You use Python functions to process EVM state and events, read contract state, and obtain contract-call results. A task groups function invocations and selects how to return or persist their outputs.

EVM-native computation

Building block	How you use it
Chain and block	Identify the chain and historical state your function reads
State and events	Select trigger data and dependencies for your computation
Contract calls	Obtain values from contracts through the runtime’s EVM node access
Derived results	Return values or write records into a target BlockDB table

Pass the chain and block context explicitly where a function needs them. Selecting a trigger block does not automatically bind every node or table read inside the function to that block.

Choose a workflow

Workflow	Input	Result
Process a new block	Rows from a source table at that block	Derived records written to a target table
Backfill history	Source rows from a bundle partition	Results submitted through the bundle write path
Run once	Explicit function arguments	Returned outputs or table writes
Invoke synchronously	One function and its arguments	One response, optionally with debug output

The caller decides when to submit tasks and which historical ranges to process. BlockX executes the submitted work.

Responsibilities

Participant	Responsibility
BlockX	Expand input into calls, execute functions, run the Writer Plugin, and report task results
BlockDB	Serve table data and handle the storage-side completion and visibility of writes
EVM node access	Read chain state and execute contract queries requested by a function
Caller	Choose input readiness, submit work, and confirm the conditions for downstream consumption

Execution vocabulary

A Task is one submitted unit of work.
A Call is one invocation of a user function. That function may make EVM contract calls.
A Call Builder expands task input into Calls.
A Writer Plugin handles the resulting outputs.
A Python Executor executes user function code; EVM requests go through the runtime to node capabilities.

Summary

Start from EVM chain, block, state, and event semantics.
Use the same vocabulary for new-block computation and historical backfill.
Choose the output path and confirm its completion conditions.

Continue reading:

Get started

Core concepts

Guides

Reference

Architecture

Components

Development guide

EVM-native computation

Choose a workflow

Responsibilities

Execution vocabulary

Summary

​EVM-native computation

​Choose a workflow

​Responsibilities

​Execution vocabulary

​Summary

EVM-native computation

Choose a workflow

Responsibilities

Execution vocabulary

Summary