Introduction

Welcome to the Open Agent Trust SDK documentation!

What is the Open Agent Trust SDK?

The Open Agent Trust SDK is a TypeScript library that enables autonomous agents to build verifiable, portable reputation on the blockchain. It provides a decentralized infrastructure for:

• Recording interactions between agents
• Issuing attestations about agent behavior and capabilities
• Building trust networks through weighted relationships
• Calculating reputation based on on-chain data

The Problem

As autonomous AI agents become more prevalent in the digital economy, a critical challenge emerges: How can agents and users trust each other?

Traditional solutions fall short:

• Centralized reputation systems lock data in silos
• Ratings can be manipulated or censored
• Reputation doesn't transfer between platforms
• No cryptographic proof of claims

The Solution

The Trust SDK solves this by putting reputation on the blockchain:

graph LR
    A[Agent A] -->|Interaction| B[Agent B]
    B -->|Attestation| C[On-Chain Record]
    C -->|Verifiable| D[Trust Score]
    D -->|Portable| E[Any Platform]

Key Benefits:

• Immutable: Records can't be altered or deleted
• Verifiable: Cryptographic proof of authenticity
• Portable: Works across all platforms
• Transparent: Anyone can verify claims
• Decentralized: No single point of control

Core Components

1. Interactions

Record when agents work together:

await sdk.registerInteraction({
  counterparty: agentAddress,
  metadata: 'task-completed',
  score: 95
});

2. Attestations

Make detailed claims about agents:

await sdk.submitAttestation({
  subject: agentAddress,
  schemaId: 'skill-verification',
  data: { skill: 'solidity', level: 'expert' },
  tags: ['verified', 'blockchain']
});

3. Trust Graph

Build networks of trusted agents:

await sdk.setTrust({
  trustee: agentAddress,
  level: 85,
  context: 'code-review'
});

4. Reputation

Calculate weighted scores:

const reputation = await sdk.getWeightedReputation({
  subject: agentAddress,
  context: 'code-review'
});

Use Cases

Freelance Marketplaces

Agents can prove their track record and clients can verify quality before hiring.

Multi-Agent Systems

Agents can discover reliable collaborators and delegate tasks with confidence.

Decentralized Networks

Build permissionless agent ecosystems with built-in trust mechanisms.

Certification Programs

Issue verifiable credentials that agents can use across platforms.

How It Works

sequenceDiagram
    participant Agent A
    participant Trust SDK
    participant Smart Contracts
    participant Agent B

    Agent A->>Trust SDK: Initialize
    Trust SDK->>Smart Contracts: Connect
    Agent A->>Agent B: Work Together
    Agent A->>Trust SDK: Record Interaction
    Trust SDK->>Smart Contracts: Store On-Chain
    Agent A->>Trust SDK: Submit Attestation
    Trust SDK->>Smart Contracts: Store Attestation
    Agent B->>Trust SDK: Query Reputation
    Trust SDK->>Smart Contracts: Read Data
    Trust SDK->>Agent B: Return Score

Technology Stack

• Blockchain: Base Sepolia (EVM-compatible Layer 2)
• Language: TypeScript
• Library: ethers.js for blockchain interaction
• Contracts: Solidity smart contracts (verified)

Network Details

Currently deployed on Base Sepolia testnet:

Contract	Address
Interaction Registry	0x12F5C3fD1893bf9b2DeaA43AE1A2CCb122C3E707
Attestation Schema Registry	0x64DaE82fE64D2fE96f90017FE51069C107BFe9d5
Trust Graph	0x8DC39B04A9C32e16DD7bd8906a8ea0d9DE6cCbDF

View full deployment details →

Getting Started

Ready to build trusted agent networks?

1. Check Prerequisites - Wallet, testnet ETH, RPC endpoint
2. Install the SDK - Quick start guide
3. Learn Core Concepts - Understand the system
4. Integrate with Agents - Add to your workflows

Quick Example

import { AgentTrustSDK } from '@open-agent-economy/trust-sdk';

// Initialize
const sdk = new AgentTrustSDK({
  rpcUrl: 'https://sepolia.base.org',
  privateKey: process.env.AGENT_PRIVATE_KEY,
  contracts: {
    interactionRegistry: '0x12F5C3fD1893bf9b2DeaA43AE1A2CCb122C3E707',
    attestationSchemaRegistry: '0x64DaE82fE64D2fE96f90017FE51069C107BFe9d5',
    trustGraph: '0x8DC39B04A9C32e16DD7bd8906a8ea0d9DE6cCbDF'
  }
});

// Record interaction
await sdk.registerInteraction({
  counterparty: '0x742d35Cc6634C0532925a3b844Bc9e7595f0bEb',
  metadata: 'task-12345',
  score: 95
});

// Submit attestation
await sdk.submitAttestation({
  subject: '0x742d35Cc6634C0532925a3b844Bc9e7595f0bEb',
  schemaId: 'task-completion-v1',
  data: { quality: 'excellent' },
  tags: ['verified']
});

// Build trust
await sdk.setTrust({
  trustee: '0x742d35Cc6634C0532925a3b844Bc9e7595f0bEb',
  level: 85,
  context: 'general'
});

Community & Support

• GitHub: Open-Agent-Economy/open-agent-trust
• NPM: @open-agent-economy/trust-sdk
• Issues: Report bugs or request features

Next Steps

Choose your path:

New to the concept? → Why Trust SDK?

Ready to code? → Getting Started

Want to understand deeply? → Core Concepts

Need integration help? → Agentic Flows