mcpField
M
Mia-Platform Console MCP Server

Mia-Platform Console MCP Server

Remote
github.com/mia-platformCloud & Infra3Updated 2w ago
Repository

About Mia-Platform Console MCP Server

Mia-Platform Console MCP Server connects developers' tools and clients to Mia-Platform Console APIs. It supports service-account authentication as well as OAuth 2.1 with Dynamic Client Registration.

Key features

  1. 01Connect to Mia-Platform Console APIs
  2. 02Service-account authentication
  3. 03OAuth 2.1 Dynamic Client Registration
  4. 04Run with Docker or locally

Use cases

  1. 01Interact with Console APIs through tools
  2. 02Authenticate machine-to-machine with a service account
  3. 03Connect to Console through OAuth login

MCP support

The MCP server makes Mia-Platform Console API capabilities available to MCP clients. It supports OAuth 2.1 with Dynamic Client Registration and service-account authentication.

Configuration

Remote servers need no local install. Pick your client, then copy the config or install in one click.

Endpointhttps://console.cloud.mia-platform.eu/console-mcp-server/mcp
Copy install command
Claude CodeTerminal
Cursor.cursor/mcp.json
VS Code.vscode/mcp.json
Claude Desktopclaude_desktop_config.json
Windsurf~/.codeium/windsurf/mcp_config.json

Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.

All endpoints
Streamable HTTPhttps://console.cloud.mia-platform.eu/console-mcp-server/mcpAuth required
Packages
npm@mia-platform/console-mcp-server1.2.2

Tools

Details

From the project README

Mia-Platform Console MCP Server

pipeline status license

Introduction

The Mia-Platform Console MCP Server is a Model Context Protocol (MCP) server that provides seamless integration with Mia-Platform Console APIs, enabling advanced automation and interaction capabilities for developers and tools.

Prerequisites

To use the Mia-Platform Console MCP Server in your client (such as Visual Studio Code, Claude Desktop, Cursor, Gemini CLI or others), you first need to have a valid account on the Mia-Platform Console instance you want to communicate with. You will be required also to include the instance host address you in the environment variable named CONSOLE_HOST.

You may decide to access via:

  • Service Account to perform machine-2-machine authentication and have full access to the MCP capabilities to perform operations on the Company where the S.A. has been created (for more information, visit our official documentation on how to create a Mia-Platform Service Account). If you do so, you need to include the environment variables MIA_PLATFORM_CLIENT_ID and MIA_PLATFORM_CLIENT_SECRET.
  • Using your own credentials: Mia-Platform Console MCP Server follows the Model Context Protocol specifications on authentication using OAuth2.1 and Dynamic Client Registration: clients that follow that specifications will be able to discover the authentication endpoints of the selected Mia-Platform instance you want to access to and guide you to perform the log in.
How to Run

You can run stable versions of the Mia-Platform Console MCP Server using Docker. You can get detailed guide in the related page of the Mia-Platform documentation.

If you don't have Docker installed, or you simply wish to run it locally, you can use NPM and Node.js. Once you have cloned the project you can run the commands:

npm ci
npm run build

These commands will install all the dependencies and then transpile the typescript code in the build folder.

NOTE

The server automatically loads environment variables from a .env file if present in the project root. You can create one by copying default.env to .env and updating the values as needed.

Once these steps are completed you can setup the MCP server using the node command like the following:

{
  "mcp": {
    "servers": {
      "mia-platform-console": {
        "command": "node",
        "args": [
          "${workspaceFolder}/mcp-server",
          "start",
          "--stdio",
          "--host=https://console.cloud.mia-platform.eu"
        ]
      }
    }
  }
}

TIP

Alternatively, you start the service after the build with the following command:

node mcp-server start

Then add the mcp server to your client simply including the url. As example for VS Code:

{
  "mcp": {
    "servers": {
      "mia-platform-console": {
        "type": "http",
        "url": "http://localhost:3000/console-mcp-server/mcp"
      }
    }
  }
}

Instead of 3000, please include the port defined in the environment variable PORT. More detail in the Environment Variables section.

Environment Variables

Environment variables located inside a file named .env are automatically included at service startup.

Variable NameDescriptionRequiredDefault Value
LOG_LEVELLog level of the applicationNoinfo
PORTPort number for the HTTP serverNo3000
CONSOLE_HOSTThe host address of the Mia-Platform Console instanceYes-
MIA_PLATFORM_CLIENT_IDClient ID for Service Account authenticationNo-
MIA_PLATFORM_CLIENT_SECRETClient secret for Service Account authenticationNo-
CLIENT_EXPIRY_DURATIONDuration in seconds of clients generated with the DCR authentication flow. After this time, the client will be expired and cannot be used anylonger.No300

Local Development

To help with the development of the server you need Node.js installed on your machine.
The recommended way is to use a version manager like nvm or mise.

Once you have setup your environment with the correct Node.js version declared inside the .nvmrc file you can run the following command:

npm ci

Once has finished you will have all the dependencies installed on the project, then you have to prepare an environment file by copying the default.env file and edit it accordingly.

cp default.env .env

Finally to verify everything works, run:

npm run local:test

If you are not targeting the Console Cloud installation you can use the --host flag and specify your own host

npm run local:test  -- --host https://CONSOLE_HOST

This command will download and launch the MCP inspector on http://localhost:6274 where you can test if the implementation will work correctly testing the discovery of tools and prompts without the needs of a working llm environment.

To run tests for new implementations you can use:

npm test

Or running a test for a single file run:

node --test --import tsx <FILE_PATH>

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.