Cheqd MCP Toolkit

The Cheqd MCP Toolkit provides a modular framework to run a Model Context Protocol (MCP) server that helps AI agents securely interact with decentralized identifiers (DIDs), verifiable credentials, and trust registries on the Cheqd network. It enables you to configure, deploy, and operate an MCP server with flexible options for remote, local NPX, or Docker-based setups.

How to use

Connect to the MCP server in the way that fits your workflow. You can use the hosted MCP server for immediate access, run a local MCP server with NPX, or deploy via Docker-Compose for a repeatable environment.

How to install

Prerequisites you need before running an MCP server:

• Node.js 20 or higher

• pnpm 8 or higher

Remote MCP server (easiest)

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

Local server via NPX

Run the MCP server locally using NPX. This approach passes your own environment into the running server.

{
  "mcpServers": {
    "cheqd": {
      "command": "npx",
      "args": [
        "-y",
        "@cheqd/mcp-toolkit-server"
      ],
      "env": {
        "TOOLS": "credo,<other available tools>",
        "CREDO_PORT": "3000",
        "CREDO_NAME": "faber",
        "CREDO_ENDPOINT": "http://faber:3000",
        "CREDO_CHEQD_TESTNET_MNEMONIC": "your-mnemonic-phrase",
        "TRAIN_ENDPOINT": "https://dev-train.trust-scheme.de/tcr/v1/",
        "PORT": "5000"
      }
    }
  }
}

Local server via docker-compose

Use Docker-Compose to run the MCP server in a containerized environment. Update the path to the docker-compose file as needed for your setup.

{
  "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 toolkit

The credo toolkit helps AI agents manage DIDs and verifiable credentials, including issuing and revoking credentials, managing schemas and definitions, and authenticating with DIDs. It also supports trust registry verification with TRAIN.

Environment and configuration notes

You will configure environment variables for Credo and related tooling when running local or remote MCP servers. The following variables are commonly used:

TOOLS="credo"        # Tools to enable
CREDO_PORT="3000"        # Credo agent port
CREDO_NAME="faber"        # Credo agent name
CREDO_ENDPOINT="http://faber:3000"  # Exposed endpoint
CREDO_CHEQD_TESTNET_MNEMONIC="your-mnemonic-phrase"
TRAIN_ENDPOINT="https://dev-train.trust-scheme.de/tcr/v1/"    
PORT="5000"

Developer setup

If you are developing or extending the toolkit locally, install pnpm, clone the project, and build the packages.

npm install -g pnpm

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

pnpm install

pnpm build

Notes on usage and security

Operate your MCP server in a controlled environment. Use the remote server for quick trials, and prefer local NPX or Docker-Compose for reproducible setups. Keep credentials and mnemonics secure and avoid exposing endpoints in public channels.

Troubleshooting

If you encounter connectivity issues, verify that the MCP server URL is reachable and that the correct port is exposed. Check that environment variables are set as shown for local runs and that dependencies are installed before starting the server.

Available tools

credo

Manage DIDs and verifiable credentials, issue/revoke credentials, manage schemas/definitions, and authenticate with DIDs. Includes TRAIN-based trust registry verification.

mcp_toolkit_server

Local MCP server runtime exposed via NPX command to serve MCP endpoints and integrate with toolkit tooling.

train_verification

Verify trust registry entries via TRAIN endpoint to ensure credential trust.