Get the FREE Ultimate OpenClaw Setup Guide →

goldrush

This project provides a MCP (Model Context Protocol) server that exposes Covalent's GoldRush APIs as MCP resources and tools. It is implemented in TypeScript using @modelcontextprotocol/sdk and @covalenthq/client-sdk.

Installation
Run this command in your terminal to add the MCP server to Claude Code.
Run in terminal:
Command
claude mcp add --transport stdio covalenthq-goldrush-mcp-server npx -y @covalenthq/goldrush-mcp-server@latest \
  --env GOLDRUSH_API_KEY="YOUR_GOLDRUSH_API_KEY"

How to use

GoldRush MCP Server exposes Covalent's GoldRush APIs as MCP tools and resources. It enables an LLM or client to call GoldRush endpoints through the MCP abstraction, making it easy to fetch token balances, chain info, quotes, and other on-chain data within a unified tool-API surface. The server supports both STDIO and HTTP transports, so you can integrate it with local CLIs, Claude Code, Cursor, Windsurf, or custom clients. Typical usage involves starting the MCP server with the required API key, then listing available tools and invoking them by name to obtain structured results from GoldRush.

In practice, you will configure an MCP entry (e.g., goldrush) with the npx command pointing at the published goldrush-mcp-server package. You must provide your GOLDRUSH_API_KEY in the environment for authentication. Once running, you can programmatically or via HTTP request discover available tools (such as token_balances, chain_info, quote_currency, etc.) and call them with the appropriate arguments (e.g., chainName, address, nft flag, etc.). The README examples show integration patterns for Claude Desktop, Claude Code CLI, Cursor, and Windsurf, plus a programmatic STDIO and HTTP transport approach for custom clients.

Tools exposed include a token_balances tool (to fetch token holdings for an address on a given chain), and other GoldRush endpoints exposed as MCP resources. You can access these tools by name and pass a structured arguments object, then handle the returned content in your application or within an LLM prompt workflow.

How to install

Prerequisites:

  • Node.js (LTS) and npm installed on your system
  • An API key from GoldRush (GOLDRUSH_API_KEY)
  1. Install the MCP server package (via npm or npx):
  • Using npm (install globally or locally in your project): npm install -g @covalenthq/goldrush-mcp-server

  • Or run directly with npx (no installation required): npx -y @covalenthq/goldrush-mcp-server@latest

  1. Set up environment variables:
  • Create a .env file or export GOLDRUSH_API_KEY in your environment: export GOLDRUSH_API_KEY=YOUR_API_KEY
  1. Run the server (Node.js runtime):
  • If installed locally: node dist/index.js --transport stdio
  • If running via npx or CLI examples, follow the transport setup in the README (STDIO or HTTP).
  1. Verify the server is reachable:
  • For HTTP transport, open http://localhost:3000 or the port you configured and send a sample MCP request.
  • For STDIO transport, interact via the provided client example to list and call tools.

Additional notes

Notes and tips:

  • You must provide a valid GoldRush API key (GOLDRUSH_API_KEY) for the tools to function.
  • The README demonstrates multiple transport options (STDIO and HTTP). Choose the one that best fits your integration.
  • When using Claude Desktop, Claude Code CLI, Cursor, or Windsurf, you can copy the exact mcpServers configuration example and replace the API key in the env block.
  • If you update to a newer version of the GoldRush MCP server, re-run with the latest tag (@latest) or a specific version to ensure compatibility.
  • For local testing, the HTTP transport example shows starting the Node server with dist/index.js and then performing HTTP POST calls to /mcp with the appropriate JSON-RPC parameters.

Related MCP Servers

Sponsor this space

Reach thousands of developers