Cheqd MCP Toolkit

What is Cheqd MCP Toolkit about?

The toolkit offers a ready‑to‑use environment for running an MCP server that connects AI agents with Cheqd’s decentralized identity infrastructure. It abstracts the complexity of DID, credential issuance, revocation, and trust‑registry verification, allowing developers to focus on AI‑driven identity workflows.

How to use Cheqd MCP Toolkit?

1. Remote option – add the provided JSON snippet to Claude Desktop or Cursor configuration to point to the hosted MCP endpoint; no local setup required.
2. Local NPX – run the server locally with npx -y @cheqd/mcp-toolkit-server and supply environment variables (TOOLS, CREDO_* and TRAIN_ENDPOINT) to customise the integrated tools.
3. Docker‑Compose – use the docker/docker-compose.yml together with an env.example file to spin up the server inside a container.

Key Features

• Configurable MCP server implementation (remote, local NPX, Docker).
• Built‑in integration with @cheqd/mcp-toolkit-credo for DID management, credential issuance, revocation, and schema handling.
• Support for TRUST Registry verification via TRAIN endpoint.
• Simple environment‑variable driven configuration (port, agent name, mnemonic, etc.).
• TypeScript first, compatible with Node 20+ and pnpm.

Use Cases

• AI assistants that need to prove identity or request verifiable credentials on‑chain.
• Automated credential issuance pipelines for education, supply‑chain, or healthcare.
• Trust‑registry lookup services for compliance or reputation systems.
• Prototyping decentralized identity solutions without building a full‑stack node.

FAQ

Q: Do I need to run my own MCP server? A: No. You can connect to the public remote server instantly, but running a local or containerised instance gives full control over data and custom tools.

Q: Which tools are currently supported? A: The toolkit ships with the credo toolset; other tools can be added by extending the TOOLS environment variable when they become available.

Q: What Node version is required? A: Node.js 20 or higher.

Q: How do I secure the mnemonic used for the Credo agent? A: Store it in a secret manager or environment file that is not committed to version control; the toolkit only reads the variable at runtime.

Q: Can I customize the server port? A: Yes, set the PORT environment variable (default 5000).

Cheqd MCP Toolkit

ℹ️ Overview

The @cheqd/mcp-toolkit is a modular framework built around the Model Context Protocol (MCP) which allows AI agents to interact with the Cheqd network. MCP standardizes AI agent interactions by providing a structured way to handle identity-related workflows. This toolkit enables AI agents to securely manage decentralized identities (DIDs), verifiable credentials, and trust registries, making it an essential component for AI-driven identity systems. This repository allows developers to configure and deploy an MCP server with the available toolkits.

🌐 Remote MCP Server

For the quickest way to get started, you can connect to our hosted MCP server. Simply add the following configuration to your Claude Desktop or Cursor settings file:

• For Claude Desktop: ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows)

• For Cursor: .cursor/mcp.json

{
    "mcpServers": {
        "cheqd-mcp": {
            "command": "npx",
            "args": [
                "mcp-remote",
                "https://remote-mcp.cheqd.io/mcp"
            ],
        }
    }
}

The remote option requires no local setup and provides immediate access to cheqd identity tools.

Prerequisites

• Node.js 20 or higher
• pnpm 8 or higher
• Basic knowledge of TypeScript and MCP

📦 Packages

@cheqd/mcp-toolkit

The @cheqd/mcp-toolkit package allows you to configure and host an MCP (Model Context Protocol) server within an environment. It integrates with tools from this repository to provide a customizable infrastructure for managing identity-related operations.

Features:

• Configurable MCP server setup
• Integration with various tools from this repository

Usage Options

1. Remote Server (Easiest)

Connect to our hosted MCP server - no local setup required:

{
    "mcpServers": {
        "cheqd-mcp": {
            "command": "npx",
            "args": [
                "mcp-remote",
                "https://remote-mcp.cheqd.io/mcp"
            ],
        }
    }
}

2. Local server via NPX

Use this to run the Cheqd MCP Server locally and pass your own environment variables.

{
    "mcpServers": {
        "cheqd": {
            "command": "npx",
            "args": [
                "-y",
                "@cheqd/mcp-toolkit-server"
            ],
            "env": {
                "TOOLS": "credo,<other available tools>",
                ...
            }
        }
    }
}

3. Local Server via docker-compose

Use the env.example file and update the appropriate variables.

{
  "mcpServers": {
    "cheqd-docker-compose": {
      "command": "docker",
      "args": [
        "compose",
        "-f",
        "/path/to/repo/mcp-toolkit/docker/docker-compose.yml",
        "run",
        "--rm",
        "-p", 
        "3000:3000",
        "--name",
        "faber",
        "-T",
        "mcp-server"
      ]
    }
  }
}

@cheqd/mcp-toolkit-credo

The @cheqd/mcp-toolkit-credo package is one of the toolkits that integrate with @openwalletfoundation/credo-ts, allowing an AI agent to manage DIDs and verifiable credentials. It provides tools for:

• Issuing and revoking credentials
• Schema and credential definition management
• DID-based authentication
• TRAIN for trust registry verification

🌍 Environment Variables

TOOLS="credo"       # Comma separated list of Tools, as of now only 'credo'
CREDO_PORT="3000"   # Port on which the Credo agent will run
CREDO_NAME="faber"  # Name of the Credo Agent
CREDO_ENDPOINT="http://faber:3000"    # Endpoint which Credo Agent is accessible externally
CREDO_CHEQD_TESTNET_MNEMONIC="your-mnemonic-phrase"   # Your Testnet mnemonic phrase
TRAIN_ENDPOINT="https://dev-train.trust-scheme.de/tcr/v1/"    # The TRAIN endpoint for verification of trust registry 
PORT="5000"   # The Port where the Remote MCP Server will run

Developer Options

Development Setup

1. Install pnpm

If you don't already have pnpm installed:

npm install -g pnpm

2. Clone the repository

git clone https://github.com/cheqd/mcp-toolkit.git
cd mcp-toolkit

3. Install dependencies

pnpm install

4. Build the packages

pnpm build

📚 Documentation

For comprehensive details on the Cheqd MCP Toolkit, usage examples, and AI agent integrations, please refer to our official documentation:

👉 MCP Documentation on cheqd.io

The documentation covers advanced topics including:

• Architecture and design of MCP
• Integrating MCP with AI agents
• Real-world use cases and patterns
• Security and trust model
• Deployment recommendations

💬 Community

Our Discord server is the primary chat channel for our open-source community, software developers, and node operators.

Please reach out to us there for discussions, help, and feedback on the project.

🙋 Find us elsewhere