The Future of Credit Scoring is AI-Native
Today, we're excited to announce MCP Services — a new way for AI agents to access Cred Protocol's credit scoring, financial reporting, and identity verification tools through the Model Context Protocol.
For the first time, AI assistants like Claude, GPT, and custom LLM applications can natively understand and evaluate the creditworthiness of any Ethereum address — no custom integration required.
What is MCP?
The Model Context Protocol (MCP) is an open standard that allows AI systems to securely discover and use external tools. Instead of building custom API integrations, AI agents can simply ask: "What tools do you have?" and immediately start using them.
This is a fundamental shift in how AI interacts with the world. Rather than treating APIs as rigid interfaces that require code, MCP treats them as capabilities that AI can reason about and use conversationally.
Why This Matters for Web3
Consider a lending protocol that wants to offer better rates to creditworthy borrowers. Today, they'd need to:
- Integrate with Cred Protocol's REST API
- Build custom logic to interpret scores
- Create user interfaces to display results
- Maintain the integration as APIs evolve
With MCP Services, an AI agent can do all of this through natural language:
User: "Should we approve a $5,000 loan for vitalik.eth?"
AI Agent: [Calls get_credit_score, get_financial_summary, get_identity_attestations]
"Based on my analysis, vitalik.eth has an Excellent credit score of 920 with $2.3M in verified assets, a clean repayment history, and identity verification through ENS and Gitcoin Passport. I recommend approval with standard terms."
No code. No integration. Just conversation.
Available MCP Tools
Our initial release includes six powerful tools:
Credit Scoring
- get_credit_score — Get the Cred Score (300-1000) for any Ethereum address or ENS name, with optional factor breakdown explaining the score
- get_credit_scores_batch — Evaluate multiple addresses simultaneously for portfolio risk analysis
Financial Reporting
- get_financial_summary — Comprehensive report including net worth, DeFi positions, transaction history, and credit events
- get_portfolio_value — Total USD value across all supported chains
- get_chain_portfolio_value — Chain-specific asset values
Identity Verification
- get_identity_attestations — Verified credentials including ENS, Gitcoin Passport, POAPs, and more
How to Use MCP Services
Option 1: HTTP Endpoints (Works Anywhere)
curl https://api.credprotocol.com/mcp/sandbox/score/vitalik.ethNo authentication required. Instant responses with sandbox data.
Option 2: AI Agent Integration
Add Cred Protocol to your AI agent's MCP configuration and start asking questions naturally:
- "What's the credit risk of this wallet?"
- "Compare the creditworthiness of these three addresses"
- "Does this user meet our identity requirements?"
Use Cases
1. AI-Powered Lending Decisions
Let AI agents evaluate loan applications by analyzing credit scores, collateral ratios, and repayment history.
2. Automated Risk Monitoring
Build AI systems that continuously monitor portfolio risk and alert on credit score changes.
3. Conversational Underwriting
Enable users to ask natural language questions about their creditworthiness and get actionable insights.
4. Identity-Gated Access
Use AI to verify identity attestations before granting access to premium features.
Sandbox-First Development
All MCP endpoints use sandbox data by default — deterministic mock responses that are perfect for development and testing. The same address always returns the same score, making it easy to build and debug integrations.
When you're ready for production, our REST API provides real-time blockchain data with the same scoring models.
What's Next
This is just the beginning. We're working on:
- Real-time MCP endpoints with live blockchain data
- Streaming responses for large reports
- Custom tool definitions for specific use cases
- Multi-chain identity aggregation
Get Started
Ready to give your AI credit superpowers?