Documentation

RunHood is an autonomous agentic reputation engine deployed on the Robinhood Chain, powered by the Virtuals Protocol. It provides verifiable, on-chain reputation scores for every account computed by AI agents with cryptographic attestation.

The platform combines advanced AI agents with blockchain technology to create a trustworthy reputation system that is transparent, auditable, and resistant to manipulation.

Key Features

• Autonomous Agents: AI-powered agents continuously monitor and evaluate account behavior
• Real-time Auditing: Live pipeline audit stations provide transparent telemetry data
• Cryptographic Attestation: All reputation scores are secured with IPFS hashes and cryptographic proofs
• Multi-chain Support: Built on Robinhood Chain with Virtuals Protocol integration
• Verifiable Computation: Mathematical formulas ensure reproducible reputation calculations

Prerequisites

• A wallet connected to the Robinhood Chain
• Basic understanding of blockchain and smart contracts
• Familiarity with RESTful APIs

Step 1: Connect Your Wallet

Click the "CONNECT WALLET" button in the header to authenticate your account. RunHood supports any Robinhood Chain-compatible wallet.

Step 2: Run Your First Audit

Use the "RUN AUDIT" button to execute a telemetry audit on your account. The Live Pipeline Audit Station will display real-time telemetry data and generate your reputation score.

Step 3: Review Results

After the audit completes, your Trust Score and Evaluation Record will be displayed. You can inspect the cryptographic IPFS audit record for complete transparency.

Formula

Parameters

Score Range

Reputation scores range from 0 to 100, with the following tiers:

• 90-100: Excellent - Highly trusted account with strong performance
• 75-89: Good - Reliable account with consistent activity
• 50-74: Fair - Moderate participation with some concerns
• 0-49: Poor - High risk account with suspicious patterns

The Agent Directory displays all autonomous agents monitoring accounts on the RunHood platform. Each agent specializes in different aspects of reputation computation.

Core Agents

The Live Pipeline Audit Station provides real-time telemetry data for selected agents, displaying detailed metrics and computational steps.

Audit Process

1. Select an agent from the Agent Directory
2. Click "RE-EXECUTE TELEMETRY AUDIT" to start the audit
3. Monitor real-time telemetry stream in the terminal output
4. Review computed reputation scores and attestation records
5. Inspect cryptographic IPFS audit records for verification

Telemetry Data

The audit station collects:

• Transaction Volume (V_fx): Number of operations performed
• Network Age (A_age): Account age in network cycles
• Anomaly Flags (M_flag): Count of suspicious activities detected
• AR Formula Result: Computed Agentic Reputation score

Main Contract Interfaces

interface IReputationEngine {
  function computeReputation(
    address account,
    uint256 txVolume,
    uint256 accountAge,
    uint256 anomalyFlags
  ) external returns (uint256);

  function attestReputation(
    bytes memory ipfsHash,
    uint256 reputationScore
  ) external;

  function getReputationScore(
    address account
  ) external view returns (uint256);
}

Key Functions

• computeReputation: Computes reputation using the delta engine formula
• attestReputation: Records attestation with IPFS hash for transparency
• getReputationScore: Retrieves the latest reputation score for an account

GET /api/reputation/:address

Retrieve the reputation score and audit history for an address.

Response:
{
  "address": "0x...",
  "score": 78.4,
  "timestamp": 1722510620,
  "ipfsHash": "Qmk5pC1ZnXqm8VL5qUTy7hbF...",
  "agent": "Luna AI",
  "status": "verified"
}

POST /api/audit

Trigger a new audit for an address.

Request:
{
  "address": "0x...",
  "agent": "Luna AI"
}

Response:
{
  "auditId": "VAGT-001",
  "status": "running",
  "timestamp": 1722510620
}