rjtmahinay/loan-application-mcp-server

Loan Application MCP Server

This is a Model Context Protocol (MCP) server built with Spring AI MCP that provides tools for integrating with loan application services.

Features

The MCP server exposes tools for four main service areas:

1. Loan Service

• create_loan_application: Create a new loan application
• get_loan_application: Retrieve a loan application by ID
• update_loan_status: Update the status of a loan application
• list_loan_applications: List all loan applications with optional filtering
• calculate_loan_eligibility: Calculate loan eligibility based on income and requested amount

2. Credit Bureau Service

• get_credit_score: Get credit score for an applicant
• get_credit_report: Get detailed credit report for an applicant
• validate_identity: Validate identity of an applicant
• check_fraud_alerts: Check for fraud alerts on SSN

3. Risk Engine Service

• assess_risk: Assess risk for a loan application
• calculate_risk_score: Calculate risk score based on financial data
• analyze_fraud_risk: Analyze fraud risk for an applicant

4. Collateral Service

• add_collateral: Add collateral to a loan
• get_collateral: Get collateral information by ID
• valuate_collateral: Valuate collateral
• verify_ownership: Verify ownership of collateral

Service Endpoints

The server integrates with the following external services:

• Loan Service: https://loan-service-git-rjtmahinay-dev.apps.rm1.0a51.p1.openshiftapps.com
• Credit Bureau Service: https://credit-bureau-service-git-rjtmahinay-dev.apps.rm1.0a51.p1.openshiftapps.com
• Risk Engine Service: https://internal-risk-engine-service-git-rjtmahinay-dev.apps.rm1.0a51.p1.openshiftapps.com
• Collateral Service: https://collateral-service-git-rjtmahinay-dev.apps.rm1.0a51.p1.openshiftapps.com

Prerequisites

• Java 17 or higher
• Maven 3.6 or higher

Running the Server

1. Clone the repository
2. Navigate to the project directory
3. Run the following command:

./mvnw spring-boot:run

The server will start on port 8080.

Configuration

The server can be configured through application.properties:

# MCP Server Configuration
server.port=8080
spring.application.name=loan-application-mcp-server

# MCP Server WebFlux Configuration
spring.ai.mcp.server.name=loan-application-mcp-server
spring.ai.mcp.server.version=1.0.0
spring.ai.mcp.server.description=MCP Server for Loan Application Services

Error Handling

The server includes graceful error handling. When external services are unavailable, the tools will return descriptive error messages indicating the service is temporarily unavailable along with the request parameters that were attempted.

Project Structure

src/
├── main/
│   ├── java/com/rjtmahinay/mcp/
│   │   ├── LoanApplicationMcpServerApplication.java
│   │   ├── config/
│   │   │   └── McpServerConfig.java
│   │   ├── controller/
│   │   │   └── McpServerController.java
│   │   └── tools/
│   │       ├── LoanServiceTools.java
│   │       ├── CreditBureauServiceTools.java
│   │       ├── RiskEngineServiceTools.java
│   │       └── CollateralServiceTools.java
│   └── resources/
│       └── application.properties

Dependencies

The project uses Spring AI MCP with the following key dependencies:

• Spring Boot 3.5.6
• Spring AI MCP Server WebFlux Starter
• Spring WebFlux
• Reactor Core

Usage with MCP Clients

This server can be used with any MCP-compatible client. The server exposes the tools via the Model Context Protocol, allowing AI assistants and other applications to interact with the loan application services.

Development

To add new tools or modify existing ones:

1. Create or modify the service classes in the tools/ package
2. Update the McpServerController to expose the new functions
3. Restart the server

Troubleshooting

If you encounter issues:

1. Check that all external services are accessible
2. Verify the server is running on the expected port
3. Check the logs for detailed error information
4. Ensure proper network connectivity to external services