IntelliConnect

IntelliConnect is a Java-based MCP (Model Context Protocol) server built on Spring Boot. It integrates a rich set of capabilities including an agent-based AI runtime, memory management, knowledge base (RAG) support, and various IoT/event features. The project emphasizes a modular MCP gateway (SSE MCP) for extending functionality and connecting with third-party services, along with role-based routing and tool invocation via an intelligent agent. The README outlines a docker-centric deployment approach, with a docker-compose setup in the docker directory to spin up required infrastructure and the Java application packaged as IntelliConnect-1.8-SNAPSHOT.jar. To run the server locally or in a container, ensure you have the prerequisites installed and adjust application.yaml to setDDL-auto to update for smooth schema management during development. Once running, you’ll be able to leverage the MCP access point, memory and knowledge management, and a suite of integrated tools and agents for intelligent IoT workflows.

Prerequisites:

• Docker and Docker Compose installed
• Java 21 runtime (JDK 21)
• MySQL, Redis, EMQX (with exhook) and InfluxDB available (via docker-compose in this project)

Installation steps:

1. Download or clone the IntelliConnect repository.
2. Use the provided docker-compose setup to bring up dependencies:
   • Navigate to the docker directory
   • Run: docker-compose up This initializes MySQL, Redis, EMQX (with exhook) and InfluxDB as defined by the project.
3. Install Java 21 if not already installed.
4. Configure application.yaml to set ddl-auto to update for development: ddl-auto: update
5. Run the application:
   • If you are running the jar directly: java -jar IntelliConnect-1.8-SNAPSHOT.jar
   • If using Docker, ensure the image tag and run command reflect your container configuration
6. Access the MCP endpoints via the configured gateway (as documented in the project docs) and verify the SSE MCP integration and tool access points are functioning.

Notes:

• The docker-compose setup is recommended for a quick start since it provisions MySQL, Redis, EMQX, and InfluxDB together.
• Ensure network connectivity between containers and proper exhook configuration in EMQX for MQTT handling.

Tips and common considerations:

• The project supports multiple AI models and agent capabilities; configure model endpoints in application.yaml as needed.
• Exhook must be correctly configured in EMQX to enable MQTT message handling for MCP interactions.
• If you modify the database schema during development, ddl-auto: update helps automatically apply changes, but review migrations for production.
• The MCP access point and SSE MCP integration allow extending capabilities with additional tools and services; explore role-based routing and semantic path configuration for advanced use cases.
• When deploying with Docker, prefer the docker-compose workflow described in the README to ensure all dependencies start in the correct order.
• If you encounter container networking issues, verify that ports are correctly exposed and that environment variables required by the server are provided (see application.yaml and docs for specifics).
• This server is Java-based; there is no npm package involved.