home / mcp / quarkus mcp servers

Quarkus MCP Servers

A collection of MCP servers built with Quarkus to extend MCP-enabled LLM applications.

Installation

Add the following to your MCP client configuration file.

Configuration

{
  "mcpServers": {
    "quarkiverse-quarkus-mcp-servers": {
      "command": "jbang",
      "args": [
        "jdbc@quarkiverse/quarkus-mcp-servers"
      ]
    }
  }
}

You can run Model Context Protocol (MCP) servers to extend your MCP-enabled LLM applications. These Java-based servers let you store data, inspect running processes, serve files, draw with JavaFX, interact with Kubernetes, manage containers, or access Wolfram Alpha capabilities, all in a scalable MCP-friendly way.

How to use

Start any MCP server locally using the provided command line interface. Each server runs as a separate process that you can connect to from MCP-enabled clients. You typically start a server and then point your MCP-enabled application to that server to perform its specialized tasks, such as data storage, JVM inspection, file access, or external knowledge queries.

How to install

Prerequisites you need before starting any server:

Java 17+ is installed on your machine
A compatible command launcher is available (the environment used here is jbang)
An MCP-enabled client or application is ready to connect to a running server

Install and start each server using its dedicated command. Run these commands in your terminal one by one as needed.

jbang jdbc@quarkiverse/quarkus-mcp-servers
```

```bash
jbang jvminsight@quarkiverse/quarkus-mcp-servers
```

```bash
jbang filesystem@quarkiverse/quarkus-mcp-servers [path1] [path2] ...
```

```bash
jbang jfx@quarkiverse/quarkus-mcp-servers
```

```bash
jbang kubernetes@quarkiverse/quarkus-mcp-servers
```

```bash
jbang containers@quarkiverse/quarkus-mcp-servers
```

```bash
jbang wolfram@quarkiverse/quarkus-mcp-servers

Further setup notes

If you need to serve specific files or directories with the filesystem server, provide the paths when starting that server. The server will then expose those files to MCP-enabled clients for retrieval or processing.

Available tools

data_store

Store and retrieve data using a JDBC-backed data source exposed through the MCP server.

jvm_inspect

Inspect a running JVM process to collect metrics, threads, and runtime information.

file_server

Serve files from the local filesystem to MCP-enabled clients for retrieval and processing.

canvas_draw

Expose a JavaFX canvas for interactive drawing and rendering operations via MCP.

k8s_integration

Interact with Kubernetes clusters to manage resources and queries from MCP clients.

container_engine

Interact with Docker/Podman/OCI-compatible containers through an MCP interface.

wolfram_search

Perform web-optimized searches and queries via the Wolfram Alpha API for MCP-enabled apps.