evmquery

n8n community node for evmquery: query EVM smart contracts via CEL expressions.

Package Information

Released: 5 mo ago
Downloads: 29 weekly / 75 monthly
Latest Version: 0.2.0
Author: abi group GmbH

Available Nodes

evmquery icon
evmquery
Read on-chain data from any supported EVM chain using CEL expressions
evmquery Trigger icon
evmquery Trigger
Trigger a workflow when a CEL expression's on-chain value changes on any supported EVM chain

Documentation

n8n-nodes-evmquery

n8n community node for evmquery: query EVM smart contracts using CEL expressions, both in regular n8n workflows and as an AI Agent tool.

What is evmquery?

evmquery is a managed HTTP API that turns CEL (Common Expression Language) queries into typed reads against EVM chains. One expression can call multiple contracts, combine their results, and return a structured value: no RPC plumbing, no ABI juggling, no fan-out across view calls.

Install (self-hosted n8n)

Settings → Community Nodes → install n8n-nodes-evmquery.

Credentials

  1. Create an API key at evmquery.com.
  2. In n8n, create an evmquery API credential and paste the key. The credential test hits GET /usage to verify the key.
  3. Optional: override Base URL if you're pointing at a private deployment.

Operations (v1)

Resource Operation Purpose
Query Execute Evaluate a CEL expression and return a typed result plus $meta sidecar
Query Validate Type-check + credit estimate, no on-chain calls
Query Describe Schema Return the callable methods on contracts and CEL helpers
Chain List List supported EVM networks (also powers the Chain dropdown)
Usage Get Credit balance, tier, billing period

Output formats (Execute)

  • Simple (default): object results are spread at the top level for direct field access; scalar results are wrapped under value. A $meta key is always attached with { blockNumber, credits, rounds, onChainCalls, latencyMs, type }.
  • Raw: the full API envelope, useful when you need result.type alongside the value or want to inspect performance / meta fields yourself.

Cookbook

v1 ships without UI presets; the recipes below show common queries you can paste into the Expression field. v2 will add an in-node preset dropdown powered by the same library.

ERC-20 balance

# Contracts
Token = 0xA0b86991c6218b36c1d19D4a2e9Eb0cE3606eB48   # USDC

# Context Variables
holder : sol_address = 0x…

# Expression
Token.balanceOf(holder)

Native (gas) balance

# Context Variables
account : sol_address = 0x…

# Expression
chain.balance(account)

NFT owner

# Contracts
Collection = 0xb47e3cd837dDF8e4c57F05d70Ab865de6e193BBB   # CryptoPunks

# Context Variables
tokenId : sol_int = 1234

# Expression
Collection.ownerOf(tokenId)

AI Agent usage

The node is marked usableAsTool: true: attach it to a Tools Agent or AI Agent node and the LLM can call any of the five operations. Typical pattern:

  1. Call Describe Schema with the contracts the agent wants to use.
  2. Call Validate with a drafted CEL expression to confirm it type-checks.
  3. Call Execute to get the typed result.

The node accepts both the normal UI-filled Contracts / Context Variables collections and $fromAI-produced JSON objects with the same shape: one code path, zero bespoke tool-mode plumbing for authors.

Credits & pricing

evmquery meters reads in credits. Every Execute response includes $meta.credits (Simple mode) or credits.consumed (Raw mode). Validate returns estimatedCredits without touching the chain; use it to pre-flight expensive expressions.

Links

License

Apache-2.0; see LICENSE.md.

Discussion