Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.xenarch.com/llms.txt

Use this file to discover all available pages before exploring further.

The Xenarch MCP server (@xenarch/agent-mcp) gives AI agents three tools to discover, pay for, and track x402-gated content. It handles wallet management, on-chain USDC payments, and access token retrieval.

​
What is an MCP server?

The Model Context Protocol (MCP) is an open standard from Anthropic that lets AI agents access external tools. An MCP server exposes a set of tools over stdio or HTTP; agents (like Claude Desktop and Claude Code) call them as needed. The Xenarch MCP server exposes x402 payment tools — check a gate, pay in USDC on Base, retrieve payment history — so any MCP-compatible agent can autonomously pay for gated APIs and content.

​
What it does

ToolDescription
xenarch_check_gateCheck if a URL or domain has a payment gate. Returns price, contract details, and payment instructions.
xenarch_payExecute a USDC micropayment on Base L2 and return a Bearer access token.
xenarch_get_historyList past payments with tx hashes, amounts, and timestamps.

​
Supported MCP clients

  • Claude Desktop
  • Claude Code
  • Any MCP-compatible client (Cursor, Cline, etc.)

​
Quick setup

claude mcp add xenarch -- npx @xenarch/agent-mcp
That’s it for Claude Code. See Installation for Claude Desktop and other clients.

​
How agents use it

An agent with the Xenarch MCP server can:
  1. Discover: “Check if example.com has a paywall” → calls xenarch_check_gate
  2. Pay: “Pay for access” → calls xenarch_pay, sends USDC on Base
  3. Access: Uses the returned Bearer token to fetch gated content
  4. Review: “Show my payment history” → calls xenarch_get_history
The agent decides when to pay based on the task context. No hardcoded payment logic needed.