Integration: SATP agent trust scores via agentfolio-mcp

[0xbrainkid]

Hey! Love what you're building with aixyz — the payment-native agent framework with ERC-8004 identity.

We built agentfolio-mcp — an MCP server for AI agent trust verification via SATP (Solana Agent Trust Protocol). It complements your ERC-8004 identity with cross-platform trust scores.

What it adds to aixyz agents

• Trust score verification before agent-to-agent payments (via x402)
• On-chain identity on Solana mainnet (200+ agents registered)
• 14 verification types — GitHub, X, Solana wallet, ETH, MCP endpoint, domain, and more
• Embeddable trust badges for agent READMEs

Quick integration

npx agentfolio-mcp

8 MCP tools — check_trust, verify_identity, browse_agents, assess_agent, etc.

The stack

Your ERC-8004 handles identity registration. SATP handles cross-platform trust verification. Together: identity + reputation.

Would love to explore an integration where aixyz agents can query trust scores before executing payments.

• npm: agentfolio-mcp
• GitHub: 0xbrainkid/satp-mcp
• Platform: agentfolio.bot

[linear]

XYZ-113

[0xbrainkid]

Integration Guide: SATP Trust Scores in aixyz

Thanks for tracking this on Linear (XYZ-113) — here's a concrete integration path.

Install

npm install agentfolio-mcp
# or run directly: npx agentfolio-mcp

8 MCP tools available: check_trust, verify_identity, search_agents, get_attestations, get_registry, browse_agents, assess_agent, get_programs.

Trust-Check Before Payment (copy-pasteable)

The key integration point — verify agent trust before executing ERC-8004 payments:

async function checkAgentTrust(walletAddress, minScore = 50) {
  const res = await fetch(
    `https://agentfolio.bot/api/satp/scores/${walletAddress}`
  );
  if (!res.ok) return { trusted: false, reason: "not-registered" };
  
  const { data } = await res.json();
  return {
    trusted: (data.reputationScore ?? 0) >= minScore,
    score: data.reputationScore ?? 0,
    level: data.verificationLevel ?? 0,
    rank: data.reputationRank || "Unknown",
    attestations: data.attestationCount ?? 0
  };
}

// Before agent-to-agent payment
const trust = await checkAgentTrust(receiverWallet, 50);
if (!trust.trusted) throw new Error(`Trust too low: ${trust.score}`);

// High-value payments → require higher level
if (amount > 100 && trust.level < 3) {
  throw new Error(`High-value requires Level 3+ (got ${trust.level})`);
}

REST API (no MCP required)

curl https://agentfolio.bot/api/satp/scores/WALLET    # Trust scores
curl https://agentfolio.bot/api/satp/identity/WALLET   # Identity data
curl https://agentfolio.bot/api/satp/registry           # Full registry (200+ agents)

How the stack fits

• ERC-8004 → Agent identity registration (your layer)
• SATP → Cross-platform trust verification + reputation (our layer)
• Together → Identity + reputation before payment execution

Full integration guide with more examples: docs/integration-guide.md

Happy to pair on a deeper integration if there's interest — e.g. a native checkTrust() hook in the aixyz payment flow.

[0xbrainkid]

Quick follow-up — saw this is tracked on Linear (XYZ-113), wanted to check if the integration guide above covers what you need or if there are questions.

A few things that might help scope the work:

1. Minimal integration (~20 lines): Just the checkAgentTrust() function above, called before executePayment(). Catches unverified agents before any funds move.

2. MCP-native (zero code): If aixyz agents already use MCP, they can call check_trust directly via agentfolio-mcp — no API integration needed, just add the MCP server to the agent's config.

3. Full integration: Trust-gated payment tiers, automatic reputation updates after successful transactions, attestation chain between agents.

We're shipping updates to the MCP server this week — happy to prioritize anything that makes the aixyz integration smoother. What's the best way to coordinate? Discord, or should I open a draft PR against the payment flow?