mcpserver.so LogoMCPFly
Submit
Quarkus + MCP = Agentic logo

Quarkus + MCP = Agentic

by jamesfalkner

AI/AgentsQuarkusLangChain4jAgenticJava

This project demonstrates a simple agentic application using Quarkus, LangChain4j, and the Model Context Protocol (MCP). It leverages multiple MCP servers to enable agent reasoning and tool invocation.

View on GitHub

Last updated: N/A

What is Quarkus + MCP = Agentic?

This project is a Quarkus-based application that utilizes the Model Context Protocol (MCP) to create an agentic system. It integrates with LangChain4j and various MCP servers (Brave search, OpenAI, Google Maps, Slack) to enable the agent to perform tasks based on user prompts.

How to use Quarkus + MCP = Agentic?

To use this application, you need to clone the repository, install Node.js and npm, and optionally a container environment (Podman or Docker). You'll also need API keys for the MCP services. Configure these keys in a .env file or application.properties. Run the application in dev mode using ./mvnw compile quarkus:dev. Access the application through http://localhost:8080 to interact with the agent.

Key features of Quarkus + MCP = Agentic

  • Integration with multiple MCP servers (Brave, OpenAI, Google Maps, Slack)

  • Agent reasoning and tool invocation

  • Quarkus Dev UI for LLM interaction

  • Simple frontend for testing the assistant

  • Support for creating ICS calendar files

  • Configuration via .env file or application.properties

Use cases of Quarkus + MCP = Agentic

  • Scheduling meetings based on team preferences and dietary needs

  • Finding restaurants based on specific criteria

  • Sending invitations via Slack

  • Generating calendar events

  • Answering questions about the reasoning behind recommendations

FAQ from Quarkus + MCP = Agentic

What is Quarkus?

Quarkus is a Supersonic Subatomic Java Framework.

What is the Model Context Protocol (MCP)?

The Model Context Protocol is a protocol for interacting with models.

How do I get API keys for the MCP services?

Links to obtain API keys for Brave Search, OpenAI, Google Maps, and Slack are provided in the README.

How do I run the application in dev mode?

Use the command ./mvnw compile quarkus:dev.

How do I test the application?

Access http://localhost:8080 and interact with the assistant through the chat interface.