Products

Data Access Portal API SQD Network Supported Chains

Development SDK SQD Cloud SQD 360

Solutions

By Use Case DeFi & Trading Analytics AI Agents Wallets & Payments Compliance All Solutions

By Industry Stablecoins RWA All Customers

Developers

Get Started Documentation Subsquid Subsquid Labs Dev Chat

Resources

Learn Guides Compare Blog Videos Changelog

Company About Ecosystem Brand Assets

Customers Pricing

Talk to us

Home / Solutions / AI Agents

AI Agents

Blockchain data for AI agents, over MCP

A raw event log is hex-encoded topics and data. The SQD Portal MCP server hands an agent pre-decoded events, transactions, traces, and fills as typed JSON across 225+ networks through one endpoint, so the model reads onchain state instead of decoding it.

Talk to us Set up the MCP server

225+ Networks

~28 MCP tools

Read-only No writes or signing

Typed JSON output

Trusted by leading teams

The Problem

Your agent shouldn't waste tokens parsing hex

A raw Ethereum event log is a blob of hex-encoded topics and data fields. An LLM needs to know the ABI, decode the parameters, and parse binary data into typed values, all before it can start reasoning about what actually happened.

That is context-window budget burned on data wrangling for every single event. Multiply by hundreds of events per block across multiple chains, and a large share of the agent's context goes to decoding instead of reasoning.

SQD delivers pre-decoded, typed JSON with human-readable field names. Your agent gets structured data it can reason about immediately.

What your agent receives USDC · Base

{

"event": "Transfer",

"token": "USDC",

"contract": "0x833589fcd6edb6e08f4c7c32d4f71b54bda02913",

"from": "0x61040e143a77f165ba44543af4a079f2c809d14b",

"to": "0x3fe04a59ebd38cf06080a6f60a98d124eb59392a",

"value": "27269390699",

"blockNumber": 47672965,

"timestamp": "2026-06-22T13:34:37Z"

}

A real row from the Portal MCP server. value is in base units: with 6 decimals that is 27,269.39 USDC. Verify it at 0x4933…86b1. No ABI lookup, no hex conversion.

How It Works

From a plain-English question to decoded onchain data

Your agent speaks to the MCP server. The MCP server calls Portal. Portal serves the data from a network of 2,000+ worker nodes. Your agent gets typed JSON it can reason about.

01AgentNatural language

"Get all USDC transfers over $1M on Base in the last hour"

02MCP ServerTool routing

portal_stream → base-mainnet → ERC-20 Transfer filter

03Portal APIStreaming query

POST /datasets/base-mainnet/stream → 2,000+ worker nodes

04ResponseDecoded JSON

decoded Transfer rows · from, to, value · typed JSON

What Your Agent Can Do

Query events & transactionsEVM, Solana, Hyperliquid

Stream real-time dataNDJSON over HTTP

Discover datasets225+ networks

Decode automaticallyTyped JSON output

View on GitHub →

Works With

Claude

Codex

Cursor

Windsurf

VS Code

Any MCP-compatible client · Plain HTTP POST · No SDK required

Get Started

One line to connect your agent to 225+ networks

The endpoint is https://portal.sqd.dev/mcp. In Claude Code, one command adds it:

claude mcp add --transport http sqd-portal https://portal.sqd.dev/mcp

Cursor and VS Code take the same URL in their MCP config. One thing to tell the agent up front: heavy queries come back as a bounded preview page with a cursor, so scope by block range and page through _pagination.next_cursor. The full setup, with three worked queries, is in the guide.

Setup guide → GitHub →

Demo

SQD MCP Server in action

Watch how the MCP server lets your agent query onchain data through natural language, no SDK setup required.

LLM Resources

Documentation designed for machines

LLM-optimized docs, OpenAPI schemas, and skill definitions, everything your agent needs to understand SQD without reading HTML.

llms.txt Lightweight documentation index for LLM context llms-full.txt Complete documentation as plain text skill.md AI agent skill definitions EVM OpenAPI Portal API schema for EVM chains Tron OpenAPI Portal API schema for Tron Solana OpenAPI Portal API schema for Solana Bitcoin OpenAPI Portal API schema for Bitcoin Hyperliquid OpenAPI Portal API schema for Fills Substrate OpenAPI Portal API schema for Polkadot

Playground

Try the Portal API

Live queries against real blockchain data. Pick a network, choose a query, hit run.

The public endpoint is free for development. See plans for dedicated portals and higher limits.

Network+200 more

Query

1curl --compressed -X POST \
'https://portal.sqd.dev/datasets/ethereum-mainnet/stream' \
-H 'Content-Type: application/json' \
-d '{
    "type": "evm",
    "fromBlock": 21000000,
    "toBlock": 21000050,
    "fields": {
      "block": {"number":true,"timestamp":true},
      "transaction": {
        "hash": true,
        "from": true,
        "to": true,
        "value": true,
        "sighash": true
      }
    },
    "transactions": [
      {
        "to": [
          "0x3fC91A3afd70395Cd496C647d5a6CC9D4B2b7FAD"
        ]
      }
    ]
  }'

Ethereumethereum-mainnet/stream

Transactions hitting the Uniswap router

These are example queries, explore the full Playground →

Capabilities

Built for AI Agents at scale

Pre-decoded, typed data

Events arrive with decoded parameter names and typed values. No ABI lookup, no hex-to-int conversion. Your agent gets {"from": "0x...", "to": "0x...", "value": "27269390699"}, not raw log topics.

Streaming context window

Feed the agent new blocks, events, and state changes as a continuous NDJSON stream over one HTTP connection. Its context stays current without polling or cron jobs.

One query shape, every chain

The same query shape works across Ethereum, Arbitrum, Base, Polygon, and Solana. An agent reuses one integration and one tool instead of five RPC clients with five different schemas.

MCP or plain HTTP

Add the MCP server to Claude Code, Cursor, or VS Code, or call the same Stream API over a plain HTTP POST from any agent framework. No SDK lock-in, no client library required.

Why SQD

What you would build yourself

An agent's real cost is not fetching onchain data but turning it into something the model can reason about, and that work compounds with every event and every chain.

Building it yourself With SQD

Reading a log
Yourself The agent needs the ABI, decodes each topic and data field, and parses binary into typed values before it can reason, spending context-window budget on every event.

With SQD Events arrive as typed JSON with decoded parameter names, so the model reads the value instead of the hex.
Multi-chain access
Yourself Five RPC clients with five different schemas, a separate integration per chain the agent reads.

With SQD One query shape and one tool across Ethereum, Arbitrum, Base, Polygon, and Solana.
Continuous updates
Yourself Poll RPC on a cron and reconnect per chain to feed the agent new activity.

With SQD New blocks and events arrive as one continuous stream over a single HTTP connection, with no polling or cron jobs.
Integration path
Yourself A client library or SDK wired into each agent framework before it can read onchain data.

With SQD The MCP server drops into Claude Code, Cursor, or VS Code, or any framework calls the same API over a plain HTTP POST.

See this on your own contracts

Tell us the chains and contracts you care about, and we will show you the query that returns them.

Talk to us Try the public API

Use cases

What teams build with SQD

Autonomous trading agents with real-time market state
Onchain research assistants that cite verifiable rows
Smart wallet copilots that explain transaction history
Governance monitoring agents tracking proposal lifecycles
Risk assessment bots with cross-protocol exposure analysis
Cross-chain bridge monitors detecting anomalous flows

Background reading

Connect Claude or Cursor to onchain data: the Portal MCP server

Wire the SQD Portal MCP server into Claude Code, Cursor, or VS Code, then run three real onchain queries and read the rows back.

Hyperliquid data for agents: perp analytics through one tool family

An agent answers whole-market Hyperliquid perp questions through the same analytics, OHLC, and query tools it uses for every other chain.

AI agents and onchain data: what an agent needs to act on blockchain state

What an agent actually needs from onchain data, the Model Context Protocol (MCP) pattern, and how on-demand and pre-indexed pipelines differ in agent workflows.

Frequently asked questions

How do AI agents access blockchain data?

SQD exposes the Portal API through the Model Context Protocol (MCP), so agents (Claude, Cursor, custom LLMs) can query 225+ chains in natural language. Agents call typed Portal tools that return decoded transactions, logs, traces, state diffs, and token transfers. No RPC mocking, no per-chain integration code; one MCP server covers every supported chain. The server is read-only: it returns onchain data for analysis and does not reconstruct account balances or move funds.

What is SQD's MCP server?

The SQD MCP server is an open-source Model Context Protocol implementation that wraps the SQD Portal API. Drop it into Claude Desktop, Cursor, VS Code, or any MCP-compatible host, and the agent can answer blockchain questions like "How many USDC transfers over $1M happened on Base in the last 24 hours?" with live data. See the MCP server documentation for setup details.

What agent frameworks are supported?

Anything that speaks Model Context Protocol. Common MCP-compatible hosts include Claude Code, Claude Desktop, Cursor, VS Code with MCP support, and Windsurf. The Portal API can also be called directly over HTTP from any agent framework (LangChain, LlamaIndex, custom Python or TypeScript agents) without using MCP; both paths use the same underlying API.

Can AI agents build and deploy indexers through SQD?

Yes, by driving the sqd CLI. Agents can scaffold a new Squid SDK project, generate entities and batch-processor logic, and run sqd deploy to push to SQD Cloud, all coordinated through natural language in their host environment (Claude Code, Cursor, and similar). The MCP server itself is read-only against Portal; deployment to Cloud goes through the standard sqd CLI tool.