rrezartprebreza/spring-boot-skills

What it solves

AI coding agents often struggle with Spring Boot, frequently hallucinating outdated patterns, ignoring project-specific architectural conventions, or using incorrect dependency versions. This project provides a library of "skills"—structured markdown files—that guide AI agents to follow specific production-grade patterns and conventions instead of generic or outdated advice.

How it works

Users drop specific skill folders into their AI agent's configuration directory (such as .claude/skills/ for Claude Code or .codex/skills/ for Codex). These skills contain triggers, conventions, and "gotchas" (lists of common AI mistakes and their corrections). When a user asks the agent to perform a task, the agent matches the request to the skill's trigger and loads the relevant guidelines to ensure the generated code matches the team's architecture and standards.

Who it’s for

Java developers and architects using Spring Boot (versions 3.x or 4.x) who use AI coding agents like Claude Code or Codex to accelerate development while maintaining high code quality and architectural consistency.

Highlights

• Version-Specific Guidance: Separate skill trees for Spring Boot 3 and 4 to ensure compatibility with the latest framework versions.
• Coded Conventions: Covers a wide range of domains including Layered/Hexagonal architecture, REST API design, Spring Data JPA, and Spring AI integration.
• Anti-Hallucination: Includes a dedicated "Gotchas" section in each skill to explicitly tell the agent what not to do based on common AI failures.
• Plugin Support: Available as marketplace plugins for Claude Code and Codex for easier installation.
• MCP Integration: Provides guidance on building Model Context Protocol (MCP) servers to connect Spring Boot apps to AI agents.