@xrplriskscore/mcp
MCP (Model Context Protocol) server for xrplriskscore.ai. Exposes XRPL wallet risk scoring, RWA compliance checks, and escrow counterparty analysis as tools for Claude Desktop, Cline, and other MCP-compatible AI clients.
The server scores and informs — it does not approve or execute transactions. That decision stays with the agent or human.
Quick start (Claude Desktop)
Run this once:
npx -p @xrplriskscore/mcp xrplriskscore-mcp-setup
Then restart Claude Desktop. The tools will be available in your next conversation.
Available tools
check_xrpl_wallet_risk — full risk assessment on an XRPL wallet
quick_xrpl_prescore — lightweight pre-transaction score
explain_xrpl_risk_score — human-readable breakdown of a verdict
check_xrpl_rwa_compliance — RLUSD, tokenized asset, and escrow compliance check
check_xrpl_credential_eligibility — XRPL credential-based eligibility check
check_xrpl_escrow_counterparty — counterparty risk analysis for escrow transactions
list_xrpl_risk_endpoints — list all available risk scoring endpoints
Each scoring tool returns a verdict: ALLOW, CHALLENGE, or BLOCK.
Payment
The free demo tier is enabled by default — good for testing and light use.
For full paid analysis, set the XRPL_SEED environment variable and use the @xrplriskscore/client SDK. Paid requests are settled on XRPL mainnet via the x402 protocol.
Paid endpoints:
/score — full risk assessment (1 XRP)
/prescore — lightweight pre-score (0.1 XRP)
/rwa-check — RWA compliance check (0.5 XRP)
/credential-check — Permissioned Domain credential screening (0.5 XRP)
/escrow-check — XLS-85 Token Escrow counterparty screening (0.5 XRP)
POST /score-batch — batch 10/25/50 wallets at 20% discount (8/20/40 XRP)
How it works
When a tool is called, the MCP server contacts the xrplriskscore.ai API. For free-tier calls, results return immediately. For paid calls, the client SDK handles the x402 payment flow: HTTP 402 → sign XRPL payment with invoice ID → retry with payment proof → receive scored result.
Links
License
MIT