About ScanPay — Code Security Oracle
ScanPay is a remote security oracle for agents and machine-to-machine workflows. It combines code scanning, transaction preflight simulation, and URL scraping, with structured verdicts and x402 payments.
Key features
- 01AST-based vulnerability scanning
- 02Deterministic code patches and execution verdicts
- 03EVM and Solana transaction simulation
- 04URL scraping with cleaned Markdown output
- 05x402 micropayments and prepaid sessions
Use cases
- 01Checking code before execution
- 02Reviewing transactions before broadcast
- 03Detecting risky token approvals
- 04Preparing scraped page content for LLMs
MCP support
ScanPay runs a remote Streamable MCP server over HTTP and SSE, with tools including `scan_code`, `simulate_tx`, and `scrape_url`. It also offers a zero-install stdio CLI.
Configuration
Remote servers need no local install. Pick your client, then copy the config or install in one click.
https://scanpay-ob3w.onrender.com/mcpclaude mcp add --transport http misterio070-scanpay https://scanpay-ob3w.onrender.com/mcp{"mcpServers":{"misterio070-scanpay":{"type":"http","url":"https://scanpay-ob3w.onrender.com/mcp"}}}Open in Cursor{"servers":{"misterio070-scanpay":{"type":"http","url":"https://scanpay-ob3w.onrender.com/mcp"}}}Open in VS Code{"mcpServers":{"misterio070-scanpay":{"type":"http","url":"https://scanpay-ob3w.onrender.com/mcp"}}}{"mcpServers":{"misterio070-scanpay":{"serverUrl":"https://scanpay-ob3w.onrender.com/mcp"}}}Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.
https://scanpay-ob3w.onrender.com/mcphttps://scanpay-ob3w.onrender.com/sseTools
Details
From the project READMEScanPay — Security Oracle & Agent Infrastructure Suite
ScanPay is an autonomous security guardrail and oracle infrastructure suite designed specifically for the Agent-to-Agent (A2A) and Machine-to-Machine (M2M) economy with x402 v2 micropayments.
- Production API:
https://scanpay-ob3w.onrender.com - Remote Streamable MCP:
https://scanpay-ob3w.onrender.com/mcp(HTTP) /https://scanpay-ob3w.onrender.com/sse(SSE) - Solana Merchant Wallet:
JDKXvegmW5j4sAJPB6YCA9ffJbN422WLMmCWCcpy1vm4 - Base (EVM) Merchant Wallet:
0xb295E606F38C363b476FC546c27226B5fBcAAEFA(USDC:0x833589fCD6eDb6E08f4c7C32D4f71b54bdA02913)
🛠️ The 3 Core Agent Services
| Service | Endpoint | MCP Tool | Pricing (Base USDC / SOL) | Purpose |
|---|---|---|---|---|
| 1. Code Security Scanner | POST /v2/scan | scan_code | $0.10 USDC (0.0007 SOL) | AST vulnerability analysis, deterministic patches, and execution gating (PASS/WARN/BLOCK). |
| 2. Pre-Flight Tx Shield | POST /v2/preflight/simulate | simulate_tx | $0.05 USDC (0.00035 SOL) | Simulates EVM/Solana transactions before broadcast, detecting honeypots, drainers, and MEV slippage traps. |
| 3. Anti-Bot Scraper Oracle | POST /v2/oracle/scrape | scrape_url | $0.02 USDC (0.00015 SOL) | Stealth browser scraping bypassing Cloudflare Turnstile/WAF, returning clean LLM Markdown. |
| 4. 10-Operation Session | POST /v2/session/deposit | create_session_deposit | $0.85 USDC (0.006 SOL) | Prepaid micro-voucher valid across all tools with sub-millisecond latency and 0ms RPC bypass. |
🎯 Machine Experience (MX) & A2A Capabilities
Traditional developer tools output verbose human prose. ScanPay produces structured Machine Responses:
- Execution Gate Verdicts: Unambiguous
ALLOWorBLOCKbooleans with 0-100risk_scoreand cryptographic Ed25519 oracle attestations. - Deterministic Auto-Remediation: Generates instant
remediated_codeandunified_diffpatches ready forgit apply. - Tx Pre-Flight Simulation: Decodes ERC-20 approvals, detects unlimited allowances to untrusted spenders, honeypot functions (
setTrading,blacklist), and SolanaSetAuthorityaccount takeovers. - Anti-Bot Scraper Oracle: Strips navigation, ads, cookies, and scripts, saving up to 90% of LLM context window tokens.
🌐 Live Discovery Endpoints
Autonomous agent frameworks discover ScanPay automatically via spec-compliant endpoints:
| Endpoint | Standard / Protocol | Purpose |
|---|---|---|
GET /llms.txt | LLMs Discovery Standard | Machine documentation for LLMs, agents, and Cursor |
GET /robots.txt | AI Crawler Standard | Explicit index permissions for GPTBot, Claude-Web, PerplexityBot |
GET /.well-known/x402 | x402 v2 Spec | Protocol discovery, CAIP-2 network, multi-chain pricing & endpoints |
GET /bazaar.json | Coinbase CDP / Onyx Bazaar | Registry catalog with extensions.bazaar.discoverable: true |
GET /.well-known/mcp.json | Model Context Protocol | Remote MCP manifest with 5 registered tools |
GET /.well-known/agent-card.json | ERC-8004 / ERC-8183 | Agent identity card declaring Oracle capabilities |
GET /.well-known/ai-plugin.json | OpenAI Plugin Standard | OpenAPI integration manifest for autonomous function callers |
GET /SKILL.md | Agent Skills | Native instructions for Claude Code, Hermes, and Codex |
GET /PAY.md | Multi-Chain Settlement | Settlement routes, recipient wallets, and spend caps |
POST /mcp & GET /sse | MCP Remote Transports | Stateless JSON-RPC tool invocation over HTTP & SSE |
🤖 Remote Model Context Protocol (MCP)
ScanPay runs as a native remote streamable MCP server.
Zero-Install stdio (CLI)
npx github:Misterio070/scanpay
Add to Claude Desktop / Cursor / Windsurf
{
"mcpServers": {
"scanpay": {
"url": "https://scanpay-ob3w.onrender.com/sse"
}
}
}
📦 Agent Framework Integrations
- Coinbase AgentKit (Base USDC):
packages/agentkit-action-scanpay(@scanpay/agentkit-action) - Solana Agent Kit v2:
packages/plugin-solana-agent-kit(@scanpay/plugin-solana-agent-kit) - ElizaOS:
packages/plugin-elizaos(@elizaos/plugin-scanpay) - Universal Guardrail Hook:
packages/scanpay-guardrail/scanpay-hook.py(Claude CodePreToolUsehook, OpenCode, Codex, LangGraph) - ERC-8183 Commerce Evaluator:
contracts/ScanPayEvaluatorV2.sol(Base ERC-8004 Validation Registry bridge)
🧪 Testing & Verification
# Run 125+ automated unit and integration tests
python3 -m pytest
📄 License
MIT © Alex Taghiashvili
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.