New community server: openclaw-trust-scorer — Solana agent trust scoring

[baronsengir007]

Community Server Announcement

I've built a MCP server for Solana agent trust scoring that I'd like to share with the MCP community.

openclaw-trust-scorer

GitHub: https://github.com/baronsengir007/openclaw-trust-scorer
Remote endpoint: http://178.104.131.84:8082/mcp

What it does

Given a Solana wallet address, queries the mainnet-beta RPC and returns a structured trust profile:

{
  "wallet_address": "9WzDXwBbmkg8ZTbNMqUxvQRAyrZzDsGYdLVL9zYtAWWM",
  "trust_score": 0.784,
  "tier": "verified",
  "signals": {
    "transaction_count": 847,
    "last_activity": "2026-04-11T14:22:31Z",
    "balance_sol": 6.19,
    "account_exists": true
  },
  "related_agents": [...]
}

Tools

• get_trust_score(wallet_address) — on-chain trust profile
• list_recent_queries(limit?) — anonymized ecosystem overview

Use case

In the emerging agent economy, AI agents need to evaluate counterparty credibility before delegating tasks or making payments. On-chain activity is a verifiable, tamper-resistant signal. This MCP server makes that signal accessible from any MCP client.

Quick start

# Add to Claude Code
claude mcp add --transport http "openclaw-trust-scorer" http://178.104.131.84:8082/mcp

# Test via curl
curl -X POST http://178.104.131.84:8082/mcp \
  -H "Content-Type: application/json" \
  -H "Accept: application/json, text/event-stream" \
  -d '{"jsonrpc":"2.0","id":1,"method":"tools/list","params":{}}'

Tech stack

• MCP TypeScript SDK v1.29.0
• Streamable HTTP transport (remote) + stdio mode (local)
• Node.js 23 / Docker
• Solana mainnet-beta public RPC

Built as part of the OpenClaw Research pipeline exploring trust and verification in agent ecosystems.