How I Built x402 Payment Rails on Both XRPL (RLUSD) and Base (USDC) for AI Agents

Most x402 implementations I've seen run exclusively on Base/USDC. That makes sense — Coinbase built the protocol and Base is the reference chain. But I needed something different: a payment rail that works on both Base (USDC) and the XRP Ledger (RLUSD) simultaneously, targeting autonomous AI agents as the primary consumers.
Here's what I built and why.
The Problem
I run Script Master Labs — an institutional-grade quantitative trading signal platform. My signals cover squeeze detection, FTD cycle mechanics, gamma exposure proxies, and dark pool flow. The data is valuable. But the monetization model was wrong.
Subscriptions require humans. API keys require account management. Neither works for an AI agent that needs to autonomously purchase a signal, consume it, and move on in under two seconds.
x402 solves this. An agent hits my endpoint, gets a 402 back, pays in USDC or RLUSD, and gets the signal. No accounts. No keys. No humans.
The Architecture
proof402-middleware is my x402 payment middleware layer. It sits in front of my signal APIs and handles the full HTTP 402 handshake.
The dual-rail setup works like this:
Base/USDC — Standard EIP-3009 flow, compatible with Coinbase's reference facilitator
XRPL/RLUSD — Custom facilitator using the XRPL SDK, settling RLUSD payments on-ledger
The resource server (my FastAPI backend) doesn't care which rail the agent uses. It validates the payment proof and serves the signal either way.
Why XRPL/RLUSD?
A few reasons:
Settlement speed — XRPL closes ledgers in 3-5 seconds, which matters for real-time signal delivery
RLUSD — Ripple's regulated stablecoin gives enterprise-grade payment assurance
Hooks — Xahau's native smart contract layer (Hooks) lets me build payment logic directly into the ledger without deploying EVM contracts
Agent diversity — Not every AI agent has a Base wallet. XRPL support widens the addressable market
CRAWLTOLL
On top of proof402-middleware, I built CRAWLTOLL — an x402-gated signal feed where every data request is a micropayment. AI agents crawl the feed, pay per signal, and consume the data autonomously.
The pricing is dynamic: squeeze signals cost more than standard price data. The agent's payment payload includes the signal type, and the middleware routes accordingly.
What's Next
NEXUS-402 at neuralosagent.com is the agent marketplace layer — a RAG orchestration system where agents discover, purchase, and consume signals through a unified MCP interface.
ghost-layer (on PyPI) is the ephemeral execution layer — agents spin up, execute, sign their residue with Ed25519, and disappear. The payment proof lives on-chain. The agent doesn't.
Repos
Signal API + x402 middleware: github.com/Timwal78/sml-x402-signal-api
XRPL Fee Forge: github.com/Timwal78/SML-XRPL-FEE-FORGE
Full stack: scriptmasterlabs.com
If you're building in the x402 space — especially on XRPL — I'm interested in connecting. The agent payment economy is early and the infrastructure layer is wide open.
Script Master Labs LLC — Service-Disabled Veteran-Owned Small Business, Kinston NC