PayCrow provides instant trust scoring and secure escrow services for AI agent transactions on the Base mainnet.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install paycrow
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 paycrow 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 PayCrow?
PayCrow is a reputation and security layer designed specifically for the agentic economy, offering a robust way to verify the credibility of any Ethereum address before sending funds. By integrating this into your Openclaw Skills library, you gain access to a multi-source trust score that analyzes over 24,000 agents based on real-time on-chain behavior and historical data. This skill empowers users and developers to mitigate risk when interacting with autonomous agents, ensuring that transactions are safe, transparent, and backed by verifiable data.
The system focuses on the Base mainnet, providing a seamless way to check trust without requiring an API key for basic queries. It serves as a critical safety net for anyone building or using automated services where financial value is exchanged. By leveraging Openclaw Skills like PayCrow, you can build more resilient AI workflows that prioritize security and reputation management.
PayCrow Use Cases
- Verifying the reputation of an AI agent before paying for a task or API call.
- Automating safety checks within coding agents to prevent funds from being sent to high-risk addresses.
- Implementing secure USDC escrow for high-value deliveries to ensure funds are only released upon successful task completion.
- Identifying unknown agents with insufficient data to trigger manual user approval workflows.
How PayCrow Works
- The skill identifies the destination Ethereum address for a proposed transaction.
- It queries the PayCrow Trust API to retrieve a real-time reputation report based on four on-chain sources.
- The system calculates a trust score ranging from 0 to 100, weighted by reputation, identity, social standing, and chain activity.
- The agent evaluates the score and recommendation (e.g., safe, caution, or risky) to determine the next action.
- If the trust score is high, the transaction proceeds; if low or flagged, the system warns the user or blocks the payment to prevent loss.
PayCrow Setup
To perform a basic trust check using this Openclaw Skills component, no complex installation is required for simple API queries. Use the following command to check an address:
curl -s "https://paycrow-app.fly.dev/trust/ADDRESS" | jq '{score: .preview.score, confidence: .preview.confidence, recommendation: .preview.recommendation}'
For advanced escrow and safe payment features, you can integrate the PayCrow MCP server or install the package via npm:
npm install paycrow
PayCrow Data Schema & Taxonomy
PayCrow aggregates data into a weighted trust score based on the following taxonomy:
| Source | Weight | Metric Measured |
|---|---|---|
| PayCrow Reputation | 40% | Escrow completion rates and historical dispute data |
| ERC-8004 Identity | 25% | Cross-ecosystem feedback and agent identity verification |
| Moltbook Social | 15% | Social standing, account age, and karma |
| Base Chain Activity | 20% | Transaction volume, wallet age, and USDC activity |
The API returns a JSON response containing a score, a confidence interval, and a clear recommendation string.
PayCrow Advanced Features
- Safe Pay Integration: Combines trust verification and USDC escrow into a single tool call for maximum security.
- Automated Dispute Resolution: Utilizes smart contracts on Base to hold funds until delivery is verified.
- Reputation-Based Decision Logic: Allows agents to automatically refuse payments to entities with a caution flag or high dispute history.
- Scalable Agent Scoring: Provides access to a database of over 24,000 scored agents within the Openclaw Skills ecosystem.
SKILL.md
Loading
Related Openclaw Skills
Dialogue Next Steps
A sophisticated conversational framework that adds intent analysis and structured next-step suggestions to AI agent responses.
Prime Directive
A foundational ethical framework and partnership covenant designed to align AI agent behavior with human operator trust through transparency and accountability.
NetPing
A comprehensive network diagnostic utility for checking host connectivity, port status, and DNS records through a unified interface.
APIMon
A comprehensive command-line toolkit for monitoring, validating, and reporting on API health and lifecycle tasks.
Data Chart Tool
A professional-grade data visualization tool that converts raw data files into presentation-ready charts with ease.
DeadLink
A robust URL validator and website crawler designed to find and report broken links in files and live environments.
Featured*
