ArbitraX
# ArbitraX (x402 & MCP Monetization Engine)
> **High-Stakes Prediction Markets & Sports Odds Arbitrage MCP Server**
> Liquidated autonomously via the **HTTP 402 Payment Required (x402)** protocol on Base (USDC).
---
## 🚀 Business Overview & Value Proposition
Traditional websites rely on human eye-balls (Google AdSense) or credit card checkout forms (Stripe). **AI Agents cannot fill out credit card forms or resolve CAPTCHAs.**
`ArbitraX` delivers high-frequency statistical arbitrage, +EV prediction spreads (Polymarket vs. Bookmakers), and algorithmic whale accumulation telemetry to AI agents (Claude, Cursor, AutoGPT, trading bots) via the **Model Context Protocol (MCP)**, collecting instant micropayments via **HTTP 402 & USDC**.
### Key Advantages
1. **Niche Fit:** Serves bots and traders who *already* hold automated liquid crypto wallets.
2. **High-Value Alpha:** Does not sell generic public prices; delivers real-time statistical edges (+EV arbitrage spreads) where agents can capitalize instantly.
3. **Frictionless M2M Commerce:** Zero account creation, zero KYC friction for autonomous agents. Transactions settle in milliseconds on the Base L2 network.
---
## 🛠️ MCP Tool Catalog & Pricing
| Tool Name | Purpose | Price (USDC) | Target Consumer |
| :--- | :--- | :---: | :--- |
| `get_prediction_arbitrage` | Scans Polymarket and calculates +EV statistical discrepancies against consensus benchmarks. | **$0.05** | Arbitrage bots, election & event prediction models |
| `search_prediction_events` | Searches active prediction markets with live outcome pricing and liquidity volume. | **$0.02** | Research agents, news summarizers, market scanners |
| `get_whale_macro_signals` | High-volume OTC and on-chain whale accumulation alerts for Crypto, Gold, and Oil. | **$0.10** | Macro momentum bots, algorithmic hedge funds |
---
## ⚙️ Architecture & Sequence
```mermaid
sequenceDiagram
autonumber
actor Agent as Autonomous AI Agent (Claude / Cursor)
participant Worker as Cloudflare Worker Edge (x402 Gateway)
participant Gamma as Polymarket Gamma API (Real-time Feeds)
participant Base as Base L2 (USDC Micropayments)
Agent->>Worker: POST /mcp (method: tools/call, name: get_prediction_arbitrage)
Note over Worker: Check header x-402-payment-proof
Worker-->>Agent: HTTP 402 Payment Required<br/>x402-price: 0.05 USDC | x402-recipient: 0xTreasury...
Agent->>Base: Execute & sign 0.05 USDC microtransaction
Agent->>Worker: POST /mcp + Header: x-402-payment-proof (txHash)
Worker->>Gamma: Fetch active prediction order books
Worker->>Worker: Compute discrepancy & +EV delta
Worker-->>Agent: HTTP 200 OK + JSON Arbitrage Payload
```
---
## 💻 Claude Desktop & Cursor Integration
Add this entry to your `claude_desktop_config.json`:
```json
{
"mcpServers": {
"arbitrax": {
"url": "https://arbitrax-gateway.inprotechve.workers.dev/mcp",
"headers": {
"x-402-payment-proof": "{\"txHash\":\"0x...\",\"payer\":\"0xAgentWallet\",\"amount\":\"0.05\"}"
}
}
}
}
```
---
## 🧪 Local Testing & Simulation
Execute the end-to-end agent simulation test:
```bash
npm run test:agent
```
---
## 🛡️ Project Isolation & Governance
This project adheres to the **Workspace Isolation Protocol (`protocolo-aislamiento-proyectos.md`)**:
- Completely decoupled from corporate CRM, gremial portals, and other local apps.
- Standalone dependencies, independent wrangler configuration, and zero cross-contamination.
TDQS
Scored across 3 tools
The three tools target distinct actions: search_prediction_events is for discovery/listing, get_prediction_arbitrage is for analytical +EV/arbitrage scanning, and get_whale_macro_signals covers a separate commodity/macro domain. The two prediction-market tools overlap slightly in data source but descriptions make their purposes clear enough to avoid misselection.
All tools follow a consistent snake_case verb_noun pattern (get_prediction_arbitrage, search_prediction_events, get_whale_macro_signals). Verb choices are predictable and no naming style mixing is present.
Three tools is a small but reasonable surface for a focused prediction-market arbitrage and macro-signals service. Each tool earns its place, though the set is on the thin side and could benefit from one or two supporting operations.
Core workflows—market discovery, arbitrage scanning, and whale/macro signals—are covered. However, there are notable gaps: no tool to fetch detailed market data (e.g., order book or historical odds) for a specific event found via search, and no drill-down or filtering beyond what the descriptions imply.