MCP · Google A2A · HTTP 402 · Spend Guardrails

Programmable Crypto Rails Built for Autonomous AI Agents.

Traditional card networks block autonomous bots with 3DS challenges, CVV prompts, and fraud freezes. XPayr gives AI agents and API providers four open machine-payment standards — backed by 100% non-custodial wallet settlement, deterministic spend guardrails, and a flat 0.5% fee.

🤖 Open 21-Tool MCP Playground 🛡️ Spend Policy Simulator 📄 View A2A agent.json ⚡ HTTP 402 Paywalls 🌐 mcp.xpayr.com Portal
1 · Anthropic MCP Standard

21-Tool Model Context Protocol Server

Install locally via npx -y @xpayr/mcp-server (registered on the Official Anthropic MCP Registry as io.github.xpayrcom/xpayr-mcp-server) or call zero-install Streamable HTTP JSON-RPC 2.0 at POST /api/mcp.

Inspect /.well-known/mcp.json →
2 · Google A2A v1.0 Standard

Agent2Agent Discovery Card (agent.json)

Enables multi-agent orchestrators (Vertex AI, ADK, LangGraph, CrewAI) to discover XPayr skills, input/output MIME modes, Bearer auth rules, and cryptographic verification endpoints automatically.

Inspect /.well-known/agent.json →
3 · HTTP 402 API Monetization

x402 Pay-Per-Call API Paywalls

Monetize LLM inference, premium datasets, and SaaS endpoints without Stripe subscriptions. Issue 402 Payment Required challenges and verify ES256 JWT access tokens offline in <1 ms via /.well-known/jwks.json.

Explore /x402 Blueprint →
4 · OpenAPI 3.1 & GPT Actions

agents.json & Custom GPT Action Specs

Drop deterministic workflow chains (/.well-known/agents.json) and trimmed GPT Action schemas into OpenAI Custom GPTs, AutoGen, and n8n AI Agent workflows (/plugins/n8n).

Inspect /.well-known/agents.json →
Merchant Treasury Safety

Quick Guardrail Policy Check

Preview how XPayr evaluates an agent payment intent against your configured auto-approval threshold and daily cap before any transaction is signed.

Policy Preset: Auto-Approve ≤ $25 · Per-Tx Cap $100 · Chains: Base, Solana, Arc
{
  "intent_amount_usd": 18,
  "policy_decision": "AUTO_APPROVED",
  "reason": "Within $25.00 autonomous threshold and $100.00 per-tx cap",
  "merchant_net_usd": 17.91,
  "xpayr_fee_usd": 0.09,
  "allowed_rails": ["base-mainnet", "solana-mainnet", "arc-testnet"]
}

AI Agent Commerce — Frequently Asked Questions

Which AI agent payment and discovery protocols does XPayr support natively?

XPayr supports four open discovery and execution standards simultaneously: Anthropic Model Context Protocol (/.well-known/mcp.json & POST /api/mcp with 21 tools), Google Agent2Agent Card (/.well-known/agent.json), HTTP 402 Payment Required API Paywalls (/.well-known/x402.json & /x402), and OpenAPI 3.1.0 / Open Agentic Commerce (/.well-known/openapi.json & /.well-known/agents.json).

How does XPayr prevent an autonomous AI agent from overspending?

XPayr never holds your private keys. Instead, merchants define deterministic AI Agent Spend Policies: per-transaction USD ceilings, rolling 24-hour budgets, explicit chain/token whitelists, and human-in-the-loop approval thresholds above a chosen dollar amount. Test policies interactively in /tools/ai-agent-policy-simulator.

Which networks work best for high-frequency machine-to-machine (M2M) micropayments?

USDC on Base, Solana, Arbitrum One, Polygon, and Circle Arc (native USDC gas) confirm in 0.4–2 seconds with sub-cent network fees and a flat 0.5% (50 bps) XPayr gateway split.

How can another AI agent discover XPayr capabilities automatically?

Autonomous orchestrators fetch https://xpayr.com/.well-known/agent.json (A2A v1.0 Agent Card) or https://xpayr.com/.well-known/mcp.json to inspect skills, authentication schemes, and tool schemas with zero manual configuration.