[Skill] Hyperf 3.1 Code style

[maskxu]

Hi everyone,

I've created an OpenClaw skill that helps AI generate code following Hyperf 3.1 framework conventions.

The Problem

When using AI to generate Hyperf code, I noticed it often:

• Uses YAML config instead of annotations
• Hardcodes config() calls instead of #[Value] / #[ConfigBean]
• Returns code: 0 or message instead of the standard { code: 200, msg, data }
• Misses proper dependency injection with #[Inject]
• Ignores coroutine safety (using global/static variables)

The Solution

This skill injects Hyperf best practices into every AI response:

Topic	Rule
Routing	Use #[AutoController], no manual registration
DI	Use #[Inject], no make() or new
Config	Use #[Value] / #[ConfigBean], no scattered config()
Response	Unified { code: 200, msg, data } format
Structure	Controller → Service → Model layers
Coroutine	No global/static vars, use Hyperf\Context\Context

Installation

git clone https://github.com/maskxu/hyperf-codestyle.git ~/.qclaw/skills/hyperf-codestyle