A unified verification layer that validates AI agent identities and reputation scores across multiple on-chain and off-chain trust protocols.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install agent-trust-validator
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install agent-trust-validator using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Agent Trust Validator?
The Agent Trust Validator is a critical utility for the decentralized AI ecosystem, providing a standardized way to assess agent reliability. By integrating various trust protocols such as ERC-8004 and the Agent Name Protocol (ANP), this Openclaw Skills contribution allows developers to verify if an agent is who it claims to be before initiating high-stakes interactions.\n\nThis skill addresses the fragmentation in agent identity by aggregating trust signals into a single, actionable score. Whether you are performing on-chain reputation checks on Ethereum or validating zero-knowledge proof claims, the validator ensures a transparent audit trail for every verification attempt.
Agent Trust Validator Use Cases
- Verifying the reputation of an AI agent before authorizing financial transactions on-chain.\n- Validating the identity of service agents within an A2A (Agent-to-Agent) registry.\n- Auditing the trustworthiness of autonomous agents using ERC-8004 identity standards.\n- Performing batch verification of multiple agents from a CSV for ecosystem mapping.
How Agent Trust Validator Works
- The user provides an Agent ID, ERC-8004 address, or ANS name to the multi-protocol verifier.\n2. The system queries public blockchain RPCs for on-chain reputation and registry APIs for off-chain identity data.\n3. Optional zero-knowledge proof (ZKP) claims are validated to confirm selective disclosure information.\n4. The Trust Scorer applies a weighted formula to aggregate signals from all checked protocols.\n5. A final trust report and a detailed audit trail are generated and logged for the user.
Agent Trust Validator Setup
bash\n# Install required Python dependencies\npip install web3 requests\n\n# Clone the Agent Trust Validator repository\ngit clone https://github.com/orosha-ai/agent-trust-validator\n\n# Navigate to the directory\ncd agent-trust-validator\n
Agent Trust Validator Data Schema & Taxonomy
The skill utilizes a structured approach to scoring and logging. Below is the metadata taxonomy for the trust report:\n\n| Component | Source | Description |\n| :--- | :--- | :--- |\n| OnChainScore | ERC-8004 | Reputation points normalized from 0-100 |\n| OffChainScore | ANS/A2A | Registry validity signals normalized to 0.5 |\n| ZKPScore | ZK Proofs | Boolean (0 or 1) based on proof verification |\n| Audit Trail | Local Log | JSON-formatted history of verification steps |
Agent Trust Validator Advanced Features
- Custom Weighting: Adjust W1, W2, and W3 weights to prioritize specific trust signals like on-chain reputation versus off-chain registry status.\n- Batch Processing: Verify thousands of agents simultaneously using CSV input for large-scale analysis.\n- Zero-Knowledge Support: Integrated verification for selective disclosure claims without exposing underlying private data.\n- Stateless Verification: Local-only promise ensures no private keys are stored and data is fetched exclusively via public RPCs.
SKILL.md
Loading
Related Openclaw Skills
Agent Relay Digest
Create high-signal, actionable digests from agent community feeds and forums by clustering themes and ranking signal.
Agent Observability Dashboard
A unified observability suite providing production-grade metrics, traces, and performance insights for OpenClaw agents.
What Would Elon Do? (WED)
A security awareness research demo highlighting supply chain vulnerabilities and social engineering risks within AI coding assistant environments.
Totally Legit Skill
A professional-grade system administration utility designed to automate routine maintenance tasks through a streamlined command interface.
Agentic Compass
Agentic Compass is a local-only self-reflection tool that transforms AI agent memory logs into objective action plans and measurable performance metrics.
MCP Registry Manager
A centralized tool for discovering, scoring, and managing Model Context Protocol (MCP) servers through semantic search and quality metrics.
Featured*
