PayRouter
Universal agentic payment router that gives AI agents the ability to pay across multiple payment rails through a single MCP interface, with cross-rail budget enforcement and a unified audit trail.
README
PayRouter
Universal agentic payment router. One MCP server, every payment rail.
PayRouter gives any AI agent — Claude, ChatGPT, Cursor, or custom — the ability to pay across multiple payment rails through a single interface, with cross-rail budget enforcement and a unified audit trail.
What it does
┌─────────────────────────────────────────────┐
│ PayRouter Server │
│ │
│ MCP Endpoint REST API Dashboard │
│ /mcp /api/v1/* :3000 │
│ │
│ ┌─────────────────────────────────────┐ │
│ │ Core Engine │ │
│ │ Mandate Engine → Router → Adapter │ │
│ └─────────────────────────────────────┘ │
│ │ │ │ │ │
│ x402 Lightning Stripe UPI │
│ (USDC) (BTC/LN) (Card) (INR) │
└─────────────────────────────────────────────┘
- MCP server (Streamable HTTP) — any MCP client connects and gets payment tools
- Mandate engine — cross-rail budget limits, category/merchant restrictions, deny-by-default
- Smart routing — picks the best rail, falls back on failure
- Dashboard — live transactions, budget bars, rail config, setup wizard
- Docs site — Fumadocs + Next 16, Stripe-like documentation
Supported rails
| Rail | Protocol | Currency | Status |
|---|---|---|---|
| x402 | HTTP 402 + stablecoins | USDC on Base | Working |
| Lightning | L402 + NWC | BTC (sats) | Working (see examples/l402) |
| Stripe | Shared Payment Tokens | USD/EUR/GBP | Planned |
| UPI | Reserve Pay (Razorpay) | INR | Planned |
Quick start
git clone https://github.com/ram0verflow/payrouter.git
cd payrouter
nvm use # Node 22
pnpm install
pnpm -r build
pnpm --filter @payrouter/server start
If no rails are configured, the setup wizard opens at http://localhost:3000/setup.
Connect an agent
Add to your Claude Desktop config:
{
"mcpServers": {
"payrouter": {
"url": "http://localhost:3001/mcp"
}
}
}
The agent gets 5 tools: request_payment, check_budget, list_transactions, discover_merchant_rails, revoke_all.
MCP tools
request_payment
merchant: "https://api.example.com"
amount: 0.01
currency: "USD"
preferred_rail: "auto"
Checks mandate → routes to best rail → executes payment → returns normalized receipt.
check_budget
Returns remaining budget across all rails for the current agent.
list_transactions
Recent transaction history with rail/category/date filters.
discover_merchant_rails
Probes a URL to detect which payment rails the merchant accepts (HTTP 402 headers, etc).
revoke_all
Emergency kill switch — revokes all payment capabilities immediately.
Project structure
payrouter/
├── packages/
│ ├── core/ # Mandate engine, router, receipt schema, types
│ ├── adapters/
│ │ └── x402/ # x402 stablecoin adapter
│ └── server/ # MCP + REST API + SQLite
├── apps/
│ ├── dashboard/ # React + Vite + Tailwind
│ └── docs/ # Fumadocs + Next 16
├── examples/
│ └── l402/ # L402 Lightning payment example (testnet)
Configuration
Create payrouter.config.json or use the dashboard setup wizard:
{
"server": {
"port": 3001,
"dashboard_port": 3000,
"db_path": "./payrouter.db"
},
"rails": {
"x402": {
"enabled": true,
"wallet_private_key": "0x...",
"network": "base-sepolia"
}
},
"mandates": {
"default": {
"daily_limit": 100,
"per_transaction_limit": 50,
"allowed_categories": ["*"],
"allowed_merchants": ["*"]
}
}
}
Environment variables also work — see .env.example.
Development
pnpm --filter @payrouter/server start # API server on :3001
pnpm --filter dashboard dev # Dashboard on :3000
pnpm --filter docs dev # Docs on :3002
Tech stack
- TypeScript, pnpm workspaces
- MCP:
@modelcontextprotocol/sdk(Streamable HTTP) - Database: better-sqlite3
- x402:
@x402/fetch+@x402/evm+ viem - Lightning:
@getalby/sdk(NWC) - Dashboard: React 19 + Vite + Tailwind CSS
- Docs: Fumadocs + Next.js 16 + Turbopack
License
MIT
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.