This extension allow usage of the HiveMQ MQTT Client inside a Quarkus App, in JVM and Native mode.

Added with the SmallRye Reactive Messaging MQTT allows usage of a new connector type smallrye-mqtt-hivemq that will use HiveMQ MQTT Client instead of Vertx MQTT client.

This adds some benefits to the original SmallRye MQTT:

• Battle tested MQTT Client outside of Vertx landscape.
• Management of external CA file for secure connections with self-signed certificates
• Backpressure support integrated with MQTT QOS.
• Automatic and configurable reconnect handling and message redelivery.
• Real Health Check against a configurable topic (defaults to the standard MQTT $SYS/broker/uptime) integrated in Quarkus HealthReport.
• Many others you can read in official documentation here.

For more information about installation and configuration please read the documentation here.

This project is configured with Claude Code, an AI-powered development assistant optimized for Java backend development and maintenance tasks.

The project includes a specialized .claude/ directory structure with:

• 7 Backend-Focused Agents: Master Orchestrator, Backend Engineer, Quality Engineer, Performance Optimizer, Security Auditor, Debugger Detective, Deep Researcher, and Content Copywriter
• Java/Quarkus Patterns: Best practices, API design, testing strategies, and performance optimization
• Session-Based Workflow: Coordinated task management with automatic tracking and documentation

All technical requests automatically trigger the master-orchestrator agent:

# Example maintenance request
"Fix the connection timeout issue in HiveMQ client"

# Claude Code automatically:
# 1. Spawns master-orchestrator to plan the task
# 2. Creates a session file in .claude/tasks/
# 3. Coordinates specialized agents (debugger-detective + deep-researcher)
# 4. Tracks progress with TodoWrite
# 5. Auto-commits when all tasks complete

Backend Engineer - Primary Java/Quarkus development, bug fixes, feature implementation

• Use for: Code changes, API modifications, Maven build updates
• Pattern files: best-java-patterns.md, quarkus.md, api-auth-patterns.md

Quality Engineer - Testing and quality assurance

• Use for: Unit tests, integration tests, test coverage
• Pattern files: performance-testing-patterns.md

Debugger Detective + Deep Researcher - Debugging (ALWAYS used together in parallel)

• Use for: Complex bugs, production issues, performance problems
• Automatically researches solutions while investigating

Performance Optimizer - JVM tuning and optimization

• Use for: Performance issues, memory leaks, message throughput optimization
• Focus: JVM profiling, HiveMQ client optimization

Security Auditor - Security reviews and vulnerability scanning

• Use for: Security audits, dependency updates, vulnerability fixes
• Focus: REST API security, authentication, compliance

Dependency Updates:

"Update HiveMQ client dependency to version X.Y.Z"
→ master-orchestrator plans update
→ backend-engineer updates pom.xml
→ quality-engineer runs tests
→ Auto-commit when complete

Bug Fixes:

"Fix connection pool exhaustion under load"
→ master-orchestrator creates session
→ debugger-detective + deep-researcher investigate (parallel)
→ backend-engineer implements fix
→ quality-engineer adds regression test
→ Auto-commit when complete

Performance Optimization:

"Optimize message throughput for high-volume scenarios"
→ master-orchestrator plans optimization
→ performance-optimizer analyzes bottlenecks
→ backend-engineer implements changes
→ quality-engineer validates performance
→ Auto-commit when complete

Security Updates:

"Run security audit and fix vulnerabilities"
→ master-orchestrator coordinates validation mode
→ security-auditor + quality-engineer + performance-optimizer (parallel)
→ backend-engineer fixes issues
→ Auto-commit when complete

For comprehensive quality checks before releases:

"Run validation" or "validate security and performance"
→ Triggers parallel execution of:
   - security-auditor (vulnerability scanning, API security)
   - performance-optimizer (JVM profiling, throughput analysis)
   - quality-engineer (test coverage, integration tests)
→ Generates comprehensive validation report
→ Creates remediation tasks for any issues

All work is tracked in session files (.claude/tasks/):

• session-current.md - Active development session
• session-NNN.md - Archived completed sessions
• Each session includes: tasks, decisions, progress, quality gates

When all TodoWrite tasks are completed, Claude Code automatically:

1. Runs pre-commit checks (Maven verify if available)
2. Stages all modified files
3. Creates commit with format: "feat: [summary] - Auto-commit after completing N tasks 🤖"
4. Notifies completion

Think Hard Directive: Always include "think hard" in requests for maximum analytical depth

Parallel Execution: Multiple independent tasks run simultaneously for faster completion

Evidence-Based: All decisions backed by research and validation

Security-First: All API changes reviewed by security-auditor

Quality Gates: Testing and validation required before completion

The .claude/context/rules/ directory contains specialized patterns:

• best-java-patterns.md - Java best practices and coding standards
• java-patterns.md - Advanced Java patterns and functional programming
• quarkus.md - Quarkus framework patterns and extensions
• api-auth-patterns.md - REST API design and authentication
• performance-testing-patterns.md - JUnit, integration tests, performance testing
• project-organization-patterns.md - Maven project structure
• git-workflow-patterns.md - Git protocols and workflows

# Dependency update
"Update Quarkus to version 3.8.0"

# Bug investigation
"Debug why MQTT reconnection fails after network timeout"

# Feature addition
"Add support for MQTT 5 shared subscriptions"

# Performance issue
"Investigate high memory usage during message processing"

# Security review
"Audit REST endpoints for security vulnerabilities"

# Documentation update
"Update README with new configuration options"

# Test coverage
"Add integration tests for connection pool manager"

Claude Code uses a session-based workflow where:

1. Master Orchestrator creates a strategic plan
2. Specialized Agents execute tasks in parallel or sequentially
3. TodoWrite tracks real-time progress
4. Session Files maintain complete development history
5. Auto-Commit preserves all completed work

All maintenance tasks are documented, tracked, and validated through this coordinated system.

Thanks goes to these wonderful people (emoji key):

masini 💻 🚧

pjgg 💻 🚧

This project follows the all-contributors specification. Contributions of any kind welcome!