MCP.so
Sign In
Servers

CCXT MCP Server

@doggybee

High-performance CCXT MCP server for cryptocurrency exchange integration

Overview

What is CCXT MCP Server?

It is a high‑performance server that connects LLMs (like Claude and GPT) to 20+ cryptocurrency exchanges through the Model Context Protocol (MCP) and the CCXT library, enabling real‑time market data access and trading operations.

How to use CCXT MCP Server?

Install globally via npm install -g @mcpfun/mcp-server-ccxt and start with mcp-server-ccxt, or run directly with npx @mcpfun/mcp-server-ccxt. Create a .env file with exchange API keys (optional for public data) and proxy settings (optional). Then add the server to your claude_desktop_config.json and restart Claude for Desktop.

Key features of CCXT MCP Server

  • Connects to 20+ cryptocurrency exchanges
  • Supports spot, futures, and swap markets
  • Proxy configuration for exchange access
  • Optimized caching and adaptive rate limiting
  • MCP‑standard compliant with LLMs

Use cases of CCXT MCP Server

  • Get current ticker prices for any trading pair
  • Retrieve order books, OHLCV candlestick data, and recent trades
  • Compare asset prices across multiple exchanges
  • Check account balances (requires API keys)
  • Place market/limit orders and manage positions (requires API keys)

FAQ from CCXT MCP Server

What exchanges does the server support?

It connects to over 20 cryptocurrency exchanges via the CCXT library.

Do I need API keys to use the server?

API keys are optional for public market data tools, but required for private tools such as account balances and order placement.

How do I configure a proxy?

Set USE_PROXY=true and PROXY_URL=http://username:password@your-proxy-server:port in your .env file.

How are API keys secured?

Best practices recommend creating dedicated API keys with minimal permissions, disabling withdrawals, using IP whitelisting, and never committing keys to version control.

What performance optimizations exist?

The server uses an LRU caching system with different TTLs (e.g., ticker: 10s, order book: 5s, market data: 1h) and adaptive rate limiting with exponential backoff.

Tags

More from Other