AI Agent Identity and Verifiable Data

What is AI Agent Identity and Verifiable Data?

This is an MCP (Model Context Protocol) server that acts as an adapter between AI models and the Empe Self-Sovereign Identity (SSI) Issuer Service. It enables AI agents to create and issue verifiable credentials within the End-to-End Verifiable Data Infrastructure (EVDI) ecosystem, supporting interactions between agents, humans, and systems.

How to use AI Agent Identity and Verifiable Data?

Configure the MCP server with your Issuer Service base URL, client secret, and schema templates in application.properties or application.yml. Build with Maven (mvn clean package) and run the JAR. Connect AI models (e.g., OpenAI Custom GPT or Anthropic Claude) to the MCP server endpoint (http://your-mcp-server:8090/api/v1). A one-click deployment portal and a demo environment (access code required) are available.

Key features of AI Agent Identity and Verifiable Data

• Schema management tools (create, list, get, delete)
• Check schema existence and retrieve latest by type
• Create targeted credential offerings for specific DIDs
• Create open credential offerings claimable by anyone
• Spring Boot and Spring AI MCP server foundation
• Tool-based integration for AI-driven credential workflows

Use cases of AI Agent Identity and Verifiable Data

• AI agents prove their identity and capabilities to other agents or systems
• Trusted authorities issue verifiable credentials (e.g., membership, proof of purchase)
• AI assistants help users manage credential schemas and offerings via natural language
• Human-to-agent and system-to-agent credential exchange with privacy and security

FAQ from AI Agent Identity and Verifiable Data

What is the difference between EVDI and this MCP server?

EVDI is the broader decentralized identity infrastructure. This MCP server implements the Issuer component, enabling AI models to create and issue verifiable credentials. Verification and agent identity management are future capabilities.

What are the prerequisites to run this server?

Java 23 or later, Maven 3.8+, and access to an instance of the Empe Issuer Service (base URL and client secret required). The README also provides a one-click portal and a demo environment with an access code.

How does the MCP server communicate with AI models?

The server exposes SSI operations as MCP tools. AI models connect to the server, discover available tools, and make tool calls. The server executes the operation on the Issuer Service and returns results.

Where does credential data live?

Credentials are managed by the Empe Issuer Service backend.