GoldRush MCP Server

The GoldRush MCP Server provides a Model Context Protocol (MCP) implementation that exposes GoldRush APIs as MCP resources and tools. It enables LLMs to interact with blockchain data across multiple chains. One connector gives you 37 tools: 28 blockchain data tools covering balances, transactions, NFTs, prices and approvals, plus 9 protocol analytics tools covering volume, fees, TVL, liquidations and active users.

You’ll need a GoldRush API key to use this server.

Get Started

Setup Instructions

Step 1: Installation

No installation is needed. The server is hosted at https://mcp.goldrushdata.com/mcp. There is nothing to clone and no config file to edit unless your client requires one.

Step 2: Add the connector

Claude: add from the connector directory

GoldRush is listed in Claude’s connector directory as a community connector. This is the shortest path and it works on both Claude.ai and Claude Desktop.

Go to Customize > Connectors.
Find GoldRush in the directory and click Connect.
Paste your GoldRush API key on the sign-in page.

The connector authenticates over OAuth, and the API key you paste is the credential it carries. You still need a GoldRush API key.

Claude: add as a custom connector (fallback)

Use this only if the directory listing is unavailable in your Claude build.

Go to Customize > Connectors and click Add custom connector.
Enter https://mcp.goldrushdata.com/mcp. If asked for an OAuth client, choose Register automatically.
Click Add, then paste your GoldRush API key on the sign-in page.

For more details, see Add a connector that isn’t in the directory.

Usage with Claude Code CLI

$ claude mcp add --transport http goldrush https://mcp.goldrushdata.com/mcp --header "Authorization: Bearer <GOLDRUSH_API_KEY>"

For more details, see Set up Model Context Protocol (MCP).

Usage with Cursor

Open Cursor Settings
Go to Features > MCP
Click + Add new global MCP server
Add this to your ~/.cursor/mcp.json:

{
  "mcpServers": {
    "goldrush": {
      "url": "https://mcp.goldrushdata.com/mcp",
      "headers": {
        "Authorization": "Bearer <GOLDRUSH_API_KEY>"
      }
    }
  }
}   

For project-specific configuration, add the above to a .cursor/mcp.json file in your project directory.

Usage with Windsurf

Add this to your mcp_config.json file:

{
  "mcpServers": {
    "goldrush": {
      "serverUrl": "https://mcp.goldrushdata.com/mcp",
      "headers": {
        "Authorization": "Bearer <GOLDRUSH_API_KEY>"
      }
    }
  }
}   

Any other MCP client

The server is a standard remote MCP server over streamable HTTP. Any client that supports remote MCP servers can connect with two values:

URL: https://mcp.goldrushdata.com/mcp
Auth: either the Authorization: Bearer <GOLDRUSH_API_KEY> header, or the OAuth flow if your client registers an OAuth client automatically.

Follow your client’s own instructions for adding a remote MCP server or custom connector.

Step 3: Programmatic Usage

Here’s an example using the MCP TypeScript SDK:

import { Client } from "@modelcontextprotocol/sdk/client/index.js";
import { StreamableHTTPClientTransport } from "@modelcontextprotocol/sdk/client/streamableHttp.js";

const transport = new StreamableHTTPClientTransport(
  new URL("https://mcp.goldrushdata.com/mcp"),
  { requestInit: { headers: { Authorization: "Bearer <GOLDRUSH_API_KEY>" } } }
);

const client = new Client(
  {
    name: "example-client",
    version: "1.0.0"
  },
  {
    capabilities: {
      tools: {}
    }
  }
);

await client.connect(transport);

// List tools
const tools = await client.listTools();
console.log("Available tools:", tools.tools.map(tool => tool.name).join(", "));

// Now you can call tools
const result = await client.callTool({
  name: "getAllChains",
  arguments: {}
});

console.log(result);

Available Tools

The GoldRush MCP Server exposes 37 tools. The 28 blockchain data tools below are the full set; the 9 protocol analytics tools are listed in the next section.

Cross Chain

multichain_address_activity: Gets a summary of wallet activity across all supported blockchains.
multichain_balances: Gets token balances for a wallet address across multiple blockchains.
multichain_transactions: Gets transactions for multiple wallet addresses across multiple blockchains.

Balances

token_balances: Commonly used to fetch the native and fungible (ERC20) tokens held by an address.
historical_token_balances: Commonly used to fetch the historical native and fungible (ERC20) tokens held by an address at a given block height or date.
native_token_balance: Get the native token balance (ETH, BNB, MATIC, etc.) for a specified wallet address on a blockchain.
erc20_token_transfers: Commonly used to render the transfer-in and transfer-out of a token along with historical prices from an address.
historical_portfolio_value: Commonly used to render a daily portfolio balance for an address broken down by the token.
token_holders: Used to get a paginated list of current or historical token holders for a specified ERC20 or ERC721 token.

Transaction

transaction: Commonly used to fetch and render a single transaction including its decoded log events.
transaction_summary: Commonly used to fetch the earliest and latest transactions, and the transaction count for a wallet.
transactions_for_address: Commonly used to fetch and render the most recent transactions involving an address.
transactions_for_block: Commonly used to fetch all transactions including their decoded log events in a block and further flag interesting wallets or transactions.

NFT

nft_for_address: Commonly used to get all NFTs owned by a specific wallet address on a blockchain.
nft_check_ownership: Commonly used to verify ownership of NFTs (including ERC-721 and ERC-1155) within a collection.
nft_check_ownership_token_id: Verifies ownership of a specific token (ERC-721 or ERC-1155) within a collection.

Security

token_approvals: Commonly used to get a list of approvals across all token contracts categorized by spenders for a wallet’s assets.

Bitcoin

bitcoin_hd_wallet_balances: Fetch balances for each active child address derived from a Bitcoin HD wallet.
bitcoin_non_hd_wallet_balances: Fetch Bitcoin balance for a non-HD address.
bitcoin_transactions: Fetch transactions for a specific Bitcoin address with full transaction details.

Prices

historical_token_prices: Commonly used to get historic prices of a token between date ranges.
pool_spot_prices: Get the spot token pair prices for a specified pool contract address. Supports pools on Uniswap V2, V3 and their forks.
gas_prices: Get real-time gas estimates for different transaction speeds on a specific network.

Utility

log_events_by_address: Commonly used to get all the event logs emitted from a particular contract address.
log_events_by_topic: Commonly used to get all event logs of the same topic hash across all contracts within a particular chain.
block: Commonly used to fetch and render a single block for a block explorer.
block_heights: Commonly used to get all the block heights within a particular date range.
resolve_address: Resolves ENS, RNS and Unstoppable Domains addresses. Only supports the resolution of a registered domain to an address.

Every tool’s parameters and their descriptions are served by the connector itself. Call tools/list over MCP, or open your client’s tool inspector, for the authoritative, always-current schema.

Protocol Analytics Tools

Nine of the 37 tools answer protocol-level analytics questions (volume, fees, TVL, liquidations, active users) rather than fetching a single address or transaction. They run over a semantic layer, and the set of datasets they can reach is discovered at runtime, so start with list_entities or search_catalog rather than guessing a dataset name.

Discovery

list_entities: The complete list of datasets with live analytics, each with its metric ids, concepts, breakdown and filter dimensions, depth and semantic quality. Anything not listed here cannot be queried.
search_catalog: Typed, fuzzy keyword search across every entity, metric, concept and dimension. Use it first when a question names a dataset, provider, measure or ticker you have not confirmed exists. Each match carries the next call to make.
get_metadata: Everything about one dataset: what each metric means, which breakdowns and filters it supports, its coverage and caveats, and an explicit statement of what it cannot answer.

Members

list_members: The values a dimension holds (chains, markets, builder codes, coins, tickers, deployers), ranked and searchable, with the exact strings to pass back as filters. Values are matched exactly and case-sensitively.

Analytics

get_metric: One metric of one dataset as a day, week or month series, with a summary of the window.
get_metrics: Several metrics of one dataset in a single call, each independently executed with its own success or no-data state.
get_breakdown: Split one metric by a dimension it supports (chain, asset, market), with each member’s share when a total is meaningful.
compare_entities: Compare exactly the 2 to 10 datasets you name, on one metric, with a leader. No other dataset is ever added.
rank_entities: Rank every eligible protocol that serves a metric, without naming them. The server discovers the candidates and checks comparability.

These tools refuse rather than approximate. If a comparison is not semantically valid, for example Aave against Morpho on liquidations, the tools return a structured refusal explaining why the comparison does not hold instead of a number that looks right and is not. Refusals are final: incomparable_metrics, metric_not_available, unsupported and requires_breakdown will not be worked around by rephrasing the question. Follow the next_actions in the response instead. Windows end on the last complete UTC day, and a null bucket means missing data, never zero.

Available Resources

Resources are a core primitive in the Model Context Protocol that allow servers to expose data and content for use as context in LLM interactions. The GoldRush MCP Server provides both static and dynamic resources:

Static Resources

config://supported-chains: The chain names the connector’s tools accept
config://quote-currencies: Provides a list of all supported quote currencies for price conversions

Dynamic Resources

status://all-chains: Provides real-time synchronization status for all chains
status://chain/: Provides detailed status information for a specific chain

Dynamic resources fetch real-time data from the GoldRush API on each request, ensuring current information.

Chains the Connector Accepts

The chainName parameter on every blockchain data tool is a fixed enum, so the connector reaches a narrower set of chains than the Foundational API does. It currently accepts 65 chain names, 44 of them mainnets. config://supported-chains serves exactly that list, and reading it is the reliable way to check whether a chain is callable through MCP. A chain that the REST API supports but that is absent from that resource cannot be requested through the connector.

Example LLM Interaction

Once you’ve set up the GoldRush MCP Server with your preferred LLM client, you can start asking blockchain-related questions:

“What are the top tokens in this Ethereum wallet: 0xeefB13C7D42eFCc655E528dA6d6F7bBcf9A2251d?”
“Show me the transaction history for this wallet address on Polygon.”
“What NFTs does this wallet own across multiple chains?”
“Get the gas prices for ERC20 transfers on Ethereum mainnet.”

The protocol analytics tools answer a different shape of question:

“Which protocols have the highest fees over the last 30 days?”
“Break down Hyperliquid volume by market for the last week.”
“What datasets do you have live analytics for?”

The LLM will automatically use the appropriate GoldRush MCP tools to retrieve the blockchain data and provide you with informative responses.

GoldRush Agent Skills — Install skills for AI agents (Claude, Cursor, Gemini)
GoldRush SDKs & CLI — TypeScript, Python, and CLI tools
Powered by GoldRush — See how projects use our APIs
What is GoldRush — Learn about all GoldRush products

Learn about GoldRush

Components

Get Started

Setup Instructions

Step 1: Installation

Step 2: Add the connector

Step 3: Programmatic Usage

Available Tools

Protocol Analytics Tools

Available Resources

Chains the Connector Accepts

Example LLM Interaction

Get Started

​Setup Instructions

​Step 1: Installation

​Step 2: Add the connector

​Step 3: Programmatic Usage

​Available Tools

​Protocol Analytics Tools

​Available Resources

​Chains the Connector Accepts

​Example LLM Interaction

​Related Resources

Setup Instructions

Step 1: Installation

Step 2: Add the connector

Step 3: Programmatic Usage

Available Tools

Protocol Analytics Tools

Available Resources

Chains the Connector Accepts

Example LLM Interaction

Related Resources