Server data from the Official MCP Registry
Pay-per-call AI + crypto + data microservices via x402. 22 endpoints. USDC on Base.
About
Pay-per-call AI + crypto + data microservices via x402. 22 endpoints. USDC on Base.
Security Report
Valid MCP server (1 strong, 0 medium validity signals). No known CVEs in dependencies. Imported from the Official MCP Registry.
10 files analyzed · No issues found
Security scores are indicators to help you make informed decisions, not guarantees. Always review permissions before connecting any MCP server.
Documentation
View on GitHubFrom the project's GitHub README.
AgentPay — agentpay.help
Machine-payable AI microservices via the 402 Payment Required protocol (x402 / MPP)
No accounts. No API keys. No OAuth. Pay per call in USDC on Base.
AgentPay is an open-source reference implementation of the Machine Payments Protocol — wrapping local AI models behind an HTTP 402 paywall so that AI agents (and humans) can pay for compute on a per-request basis using stablecoins.
Built with Express 5, @x402/express, and Ollama-served Gemma models. Live on Base mainnet with the PayAI facilitator.
Table of Contents
- Quick Start
- Architecture
- Services & Pricing
- Tech Stack
- Deployment
- API Reference
- Configuration
- Contributing
- License
Quick Start
Prerequisites
- Node.js ≥ 20
- Ollama running locally with the required model pulled
- A wallet private key (for receiving payments)
1. Clone & install
git clone https://github.com/your-org/AgentPay.git
cd AgentPay
npm install
2. Pull the AI model
ollama pull gemma3:1b
# Or use a larger model for better quality:
# ollama pull gemma4:31b-cloud
3. Configure
cp .env.example .env
# Edit .env — set SELLER_ADDRESS to your wallet address
4. Start the server
npm start
# AgentPay listening on :4021
# payTo: 0xYourWalletAddress
# network: eip155:84532 (Base Sepolia testnet)
# facilitator: https://x402.org/facilitator
5. Test a paid request
# Unpaid request → HTTP 402 (paywall)
curl -s -o /dev/null -w "%{http_code}" -X POST https://agentpay.help/v1/summarize \
-H 'Content-Type: application/json' \
-d '{"text":"Machine Payments Protocol lets AI agents pay for API calls using the HTTP 402 status code."}'
# → 402
# Automated test (requires buyer wallet with USDC)
npm run test:402
6. Buy a service (buyer client)
# Set your buyer private key in .env
echo "BUYER_PK=0xYourPrivateKey" >> .env
# Run the buyer script
npm run buyer -- /v1/summarize ./payload.json
Architecture
┌─────────────────────────────────────────────────────────────────┐
│ AgentPay Architecture │
├─────────────────────────────────────────────────────────────────┤
│ │
│ ┌──────────┐ HTTP POST ┌───────────────────────────┐ │
│ │ Client │ ──────────────► │ Express 5 Server │ │
│ │ (Agent / │ (no auth) │ (port 4021) │ │
│ │ Human) │ │ │ │
│ └──────────┘ │ ┌─────────────────────┐ │ │
│ │ │ │ Payment Middleware │ │ │
│ │ │ │ (@x402/express) │ │ │
│ │ │ │ │ │ │
│ │ ◄── HTTP 402 ───────│ │ • Validates x402 │ │ │
│ │ (paywall) │ │ payment headers │ │ │
│ │ │ │ • Verifies on-chain │ │ │
│ │ ──── signed payment ►│ │ via facilitator │ │ │
│ │ (USDC) │ │ │ │ │
│ │ │ └──────────┬──────────┘ │ │
│ │ ◄── 200 OK ─────────│ │ │ │
│ │ (result JSON) │ ┌──────────▼──────────┐ │ │
│ │ │ │ Service Handlers │ │ │
│ │ │ │ │ │ │
│ │ │ │ /v1/summarize │ │ │
│ │ │ │ /v1/classify-ins │ │ │
│ │ │ │ /v1/extract │ │ │
│ │ │ └──────────┬──────────┘ │ │
│ │ └─────────────┼─────────────┘ │
│ │ │ │
│ │ ┌──────▼──────┐ │
│ │ │ Ollama │ │
│ │ │ (local LLM) │ │
│ │ │ gemma3:1b │ │
│ │ └─────────────┘ │
│ │ │
│ ┌────▼────────────────────────────────────────────────────┐ │
│ │ Payment Flow (x402) │ │
│ │ │ │
│ │ Client ──► HTTP 402 ──► Facilitator ──► On-Chain ──► │ │
│ │ │ (PayAI) Base Mainnet │ │
│ │ ▼ (USDC) │ │
│ │ Payment Required │ │
│ │ (price + accepts[]) │ │
│ └─────────────────────────────────────────────────────────┘ │
│ │
│ ┌─────────────────────────────────────────────────────────┐ │
│ │ Free Endpoints (no paywall) │ │
│ │ • / — Landing page (HTML) │ │
│ │ • /health — Health check │ │
│ │ • /stats — Revenue & usage stats │ │
│ │ • /.well-known/x402 — Machine-readable service catalog │ │
│ └─────────────────────────────────────────────────────────┘ │
│ │
│ ┌─────────────────────────────────────────────────────────┐ │
│ │ Data Layer │ │
│ │ • data/ledger.json — Append-only payment ledger │ │
│ └─────────────────────────────────────────────────────────┘ │
└─────────────────────────────────────────────────────────────────┘
How it works:
- Client sends
POST /v1/summarize(or any paid endpoint) — no auth headers needed - Payment middleware intercepts, returns HTTP 402 with pricing info (
accepts[]) - Client constructs a USDC payment, signs it, attaches
X-PAYMENTheader - Facilitator verifies the payment on Base mainnet
- Middleware grants access → request proceeds to the service handler
- Handler calls Ollama, returns AI-generated result as JSON
Services & Pricing
| Endpoint | Price | Description |
|---|---|---|
POST /v1/summarize | $0.01 | Summarize text (200–20,000 chars). Returns a ~250-word summary. |
POST /v1/classify-insurance | $0.02 | Classify insurance leads: intent, urgency, line-of-business, confidence score. |
POST /v1/sentiment | $0.02 | Sentiment analysis: positive/negative/neutral with emotions and keywords. |
POST /v1/extract | $0.03 | Extract structured key-value fields from raw text (emails, forms, documents). |
POST /v1/translate | $0.03 | Text translation to any language. |
POST /v1/code-review | $0.05 | AI code review: bugs, security issues, performance, quality analysis. |
POST /v1/insurance-analysis | $0.10 | ⭐ FULL BUNDLE — classification + extraction + summary in one call. |
All services accept USDC on Base mainnet (chain ID 8453) via the exact payment scheme. Testnet (Base Sepolia) is available via configuration.
Tech Stack
| Component | Technology |
|---|---|
| Runtime | Node.js ≥ 20 (ESM) |
| HTTP Server | Express 5.2 |
| Payment Protocol | @x402/express 2.22, @x402/evm, @x402/fetch, @x402/extensions |
| Blockchain | Base (OP Stack L2), USDC stablecoin |
| Facilitator | PayAI x402 facilitator (x402.org/facilitator) |
| AI Runtime | Ollama (local inference) |
| LLM | Gemma 3 1B (default) / Gemma 4 31B (recommended) |
| Wallet | viem (Ethereum client library) |
| Config | dotenv |
Deployment
Local Development
# Base Sepolia testnet (recommended for development)
cp .env.example .env
# Edit .env: PAYMENT_NETWORK=eip155:84532
npm start
Production (Base Mainnet)
# Edit .env for mainnet
PAYMENT_NETWORK=eip155:8453 # Base mainnet
SELLER_ADDRESS=0xYourMainnetWallet
OLLAMA_URL=http://127.0.0.1:11434
MODEL_SUMMARIZE=gemma4:31b-cloud # Use larger model for quality
MODEL_CLASSIFY=gemma4:31b-cloud
MODEL_EXTRACT=gemma4:31b-cloud
Docker (recommended for production)
FROM node:20-slim
WORKDIR /app
COPY package*.json ./
RUN npm ci --omit=dev
COPY src/ ./src/
COPY data/ ./data/
EXPOSE 4021
HEALTHCHECK CMD curl -f https://agentpay.help/health || exit 1
CMD ["node", "src/server.js"]
Environment Variables
| Variable | Required | Default | Description |
|---|---|---|---|
PORT | No | 4021 | Server port |
SELLER_ADDRESS | Yes | — | Wallet address to receive USDC payments |
PAYMENT_NETWORK | No | eip155:84532 | Blockchain network (eip155:8453 for mainnet) |
FACILITATOR_URL | No | https://x402.org/facilitator | x402 facilitator endpoint |
OLLAMA_URL | No | http://127.0.0.1:11434 | Ollama API base URL |
MODEL_SUMMARIZE | No | gemma3:1b | Model for summarize endpoint |
MODEL_CLASSIFY | No | gemma3:1b | Model for classify-insurance endpoint |
MODEL_EXTRACT | No | gemma3:1b | Model for extract endpoint |
PUBLIC_URL | No | — | Public URL for discovery metadata |
API Reference
Free Endpoints
GET /
Landing page with service catalog and usage stats.
GET /health
Health check.
{ "ok": true, "ts": "2025-01-01T00:00:00.000Z" }
GET /stats
Revenue and usage statistics.
{
"requests_paid": 42,
"gross_usd": 0.84,
"by_service": { "summarize": 0.42, "classify-insurance": 0.28, "extract": 0.14 },
"last_20": [...]
}
GET /.well-known/x402
Machine-readable service catalog (Bazaar discovery extension). Use this for automated service discovery by AI agents.
{
"name": "AgentPay",
"description": "Pay-per-call AI microservices (x402 / MPP)",
"endpoints": [
{ "path": "/v1/summarize", "method": "POST", "price": "$0.01", "description": "Summarize text (200-20k chars)" },
{ "path": "/v1/classify-insurance", "method": "POST", "price": "$0.02", "description": "Insurance lead classification" },
{ "path": "/v1/extract", "method": "POST", "price": "$0.03", "description": "Structured field extraction" }
]
}
Paid Endpoints
All paid endpoints require a valid x402 payment in the X-PAYMENT header. Unpaid requests receive HTTP 402 Payment Required.
POST /v1/summarize — $0.01
Summarize text into a concise ~250-word output.
Request:
{
"text": "Your text to summarize (200-20000 characters)..."
}
Response (200 OK):
{
"summary": "The text discusses...",
"words": 247
}
Errors:
400— Missingtextfield or text exceeds 20,000 characters402— Payment required (see x402 protocol)502— Upstream AI model failed
POST /v1/classify-insurance — $0.02
Classify an insurance lead or customer message.
Request:
{
"text": "I was in a car accident last week and need to file a claim urgently..."
}
Response (200 OK):
{
"intent": "claim",
"urgency": "high",
"line": "auto",
"confidence": 0.92
}
Possible values:
intent:quote_request|renewal|claim|complaint|otherurgency:low|medium|highline:auto|home|life|health|commercial|other
POST /v1/extract — $0.03
Extract structured fields from raw text (emails, forms, documents).
Request:
{
"text": "From: john@example.com\nSubject: Policy #12345 renewal\nDear customer, your auto policy expires on March 15...",
"fields": ["email", "policy_number", "expiry_date"]
}
Response (200 OK):
{
"email": "john@example.com",
"policy_number": "12345",
"expiry_date": "March 15"
}
If fields is omitted, all extractable key-value pairs are returned.
Client Library (Buyer)
Use @x402/fetch to automatically handle the 402 → payment → retry flow:
import { wrapFetchWithPayment, x402Client } from "@x402/fetch";
import { ExactEvmScheme } from "@x402/evm/exact/client";
import { privateKeyToAccount } from "viem/accounts";
const signer = privateKeyToAccount(process.env.BUYER_PK);
const client = x402Client.fromConfig({
schemes: [{ network: "eip155:*", client: new ExactEvmScheme(signer) }],
});
const payFetch = wrapFetchWithPayment(globalThis.fetch, client);
// This automatically handles the 402 → payment → retry flow
const res = await payFetch("https://agentpay.help/v1/summarize", {
method: "POST",
headers: { "Content-Type": "application/json" },
body: JSON.stringify({ text: "Your text here..." }),
});
const result = await res.json();
console.log(result.summary);
Contributing
Contributions are welcome! This is an open-source reference implementation of the x402 / MPP protocol.
Development Setup
git clone https://github.com/your-org/AgentPay.git
cd AgentPay
npm install
cp .env.example .env
# Edit .env with your test wallet and Base Sepolia settings
npm start
Adding a New Service
- Define the payment middleware entry in
src/server.jsunder thepaymentMiddleware()call - Add the route handler after the middleware block
- Register the endpoint in
/.well-known/x402discovery - Add to the landing page HTML
Guidelines
- Keep it simple. This is a reference implementation — clarity over complexity.
- Test on Base Sepolia first. Use the testnet before going to mainnet.
- Use
@x402/packages. Don't reinvent payment verification. - Append-only ledger. Never modify
data/ledger.json— only append.
Reporting Issues
Open a GitHub issue with:
- Steps to reproduce
- Expected vs. actual behavior
- Environment (Node version, OS, model used)
License
MIT
Resources
- x402 Protocol Spec — The Machine Payments Protocol
- PayAI Facilitator — Payment verification service
- Base Network — OP Stack L2 where USDC payments settle
- Ollama — Local LLM inference engine
- viem — TypeScript Ethereum client
Reviews
No reviews yet
Be the first to review this server!
More Developer Tools MCP Servers
Git
Freeby Modelcontextprotocol · Developer Tools
Read, search, and manipulate Git repositories programmatically
Fetch
Freeby Modelcontextprotocol · Developer Tools
Web content fetching and conversion for efficient LLM usage
Toleno
Freeby Toleno · Developer Tools
Toleno Network MCP Server — Manage your Toleno mining account with Claude AI using natural language.
mcp-creator-python
Freeby mcp-marketplace · Developer Tools
Create, build, and publish Python MCP servers to PyPI — conversationally.
MCP Marketplace
Freeby mcp-marketplace · Developer Tools
Search and install MCP servers from inside your AI client.
MarkItDown
Freeby Microsoft · Content & Media
Convert files (PDF, Word, Excel, images, audio) to Markdown for LLM consumption
