Open App

agents_trust_score_single

>-

agents_trust_score_single

Returns a full in-depth trust profile for a single ERC-8004 registered AI agent. Requires agent_id and chain_id - obtain these from agents_trust_score_list.

MCP Endpoint: https://prediction.mcp.chainaware.ai/sse

Input Schema

Field Type Required Description
agent_id integer Yes Agent ID from agents_trust_score_list
chain_id integer Yes Chain ID where the agent is registered (e.g. 1 = ETH, 56 = BNB)

Output Schema

{
  "agent_id": 0,
  "chain": "string",
  "chain_id": 1,
  "owner_address": "string",
  "agent_wallet": "string",
  "wallet_verified": true,
  "agent_uri": "string",
  "registered_at": "ISO-8601",
  "fetched_at": "ISO-8601",
  "error": "string | null",
  "meta_name": "string",
  "meta_description": "string",
  "meta_image": "string",
  "metadata_json": {
    "type": "string",
    "name": "string",
    "description": "string",
    "image": "string",
    "active": true,
    "supportedTrust": ["string"]
  },
  "registration": {
    "agent_name": "string",
    "agent_desc": "string",
    "fetch_status": "string",
    "fetched_at": "ISO-8601"
  },
  "wallets": [
    {
      "wallet_chain_id": 1,
      "wallet_address": "string",
      "source": "registry",
      "fetched_at": "ISO-8601"
    }
  ],
  "reputation_score": "string",
  "trust_score": 0,
  "trust_tier": "string",
  "trust_flags": ["string"]
}

Key Fields

Field Description
trust_score 0-1000 composite trust score. Higher = more trustworthy.
trust_tier Human-readable tier label (see Tier Interpretation below)
trust_flags Array of specific risk or trust signals flagged for this agent
wallet_verified Whether the agent's declared wallet matches on-chain registration. If false, treat the score with extra caution.
owner_address The human or entity who registered and controls the agent
agent_wallet The agent's operational wallet address
supportedTrust Trust frameworks the agent declares support for (from ERC-8004 metadata)
error Non-null if metadata fetch failed - agent may have an invalid or unreachable URI

Trust Tier Interpretation

Score Range Trust Tier Meaning
800-1000 Elite Exceptionally strong on-chain history - high confidence to transact
600-799 High Strong track record - suitable for most agentic commerce use cases
400-599 Moderate Some history present - proceed with transaction limits
200-399 Low Thin or mixed history - enhanced scrutiny recommended
1-199 Very Low Insufficient or negative history - avoid high-value interactions
0 Fraud / New Flagged as fraudulent or no verifiable on-chain history

Code Examples

Node.js

const profile = await client.callTool({
  name: "agents_trust_score_single",
  arguments: {
    agent_id: 123,
    chain_id: 1
  }
});

console.log("Agent:", profile.meta_name);
console.log("Trust Score:", profile.trust_score);
console.log("Trust Tier:", profile.trust_tier);
console.log("Flags:", profile.trust_flags);
console.log("Wallet verified:", profile.wallet_verified);

Python

profile = await session.call_tool("agents_trust_score_single", {
    "agent_id": 123,
    "chain_id": 1
})

print(f"Score: {profile['trust_score']}")
print(f"Tier: {profile['trust_tier']}")
print(f"Flags: {profile['trust_flags']}")
print(f"Wallet verified: {profile['wallet_verified']}")

Example Agent Prompts

Deep profile for agent ID 123 on Ethereum
What is the trust score for agent 42 on BNB Chain?
What trust flags does this agent have? agent_id: 89 chain_id: 8453
Is this ERC-8004 agent safe to transact with? agent_id: 456
Show me the full trust profile for agent 101 on Ethereum

Common Chain IDs

chain_id Chain
1 Ethereum
56 BNB Chain
8453 Base
137 Polygon
42161 Arbitrum

Use Cases

  • Agentic commerce platforms - verify an agent's trustworthiness before allowing it to execute purchases on behalf of users
  • AI agent marketplaces - surface trust_flags prominently for low-score agents
  • DeFi protocols - gate high-value operations (flash loans, large swaps) based on agent trust tier
  • Multi-agent pipelines - screen counterparty agents before delegating tasks or sharing funds

Composability

  • Use predictive_fraud on agent_wallet or owner_address for a deeper behavioral fraud check
  • Use predictive_behaviour on agent_wallet to understand the agent's full on-chain behavioral profile

Error Codes

Code Meaning
400 Malformed agent_id or chain_id
500 Temporary backend failure - retry after a short delay

Further Reading

See also: agents_trust_score_list | Prediction MCP Overview | Setup Guide