AI models are trained on the internet. The internet contains plenty of low-quality, insecure, and outdated code. To get consistently better output from AI coding assistants, you need to guide them with tools and instructions.
I quickly discovered that a single CLAUDE.md file is not the right approach for implementing all my rules and techniques. The ecosystem already has many excellent skills, plugins, and hooks -- but they are scattered across dozens of repositories, marketplaces, and blog posts. Finding, evaluating, and combining them takes hours.
That's why I've published my entire setup as a curated, ready-to-use collection with an installer script: github.com/alpham8/claude-webdev
What's Included
| Path | Purpose |
|---|---|
CLAUDE.md |
Global engineering baseline (coding standards, workflow rules) |
rules/ |
18 rule files referenced from CLAUDE.md |
skills/ |
Domain-specific reference skills for Claude |
hooks/ |
4 PostToolUse / PreToolUse hook scripts |
settings.json |
Claude Code settings (plugins, MCP servers, hooks, permissions) |
install.sh |
One-command installer into ~/.claude |
Skills in Detail
The collection covers my entire technology stack:
- Shopware 5 and 6 -- Plugin architecture, DAL, event system, admin components, Storefront JS, DDEV setup, dustin/shopware-utils
- Symfony -- Full framework coverage plus project scaffolding blueprint (Vite, DDEV, i18n, CAPTCHA, deployment)
- PHP -- Modern PHP 8.0-8.4, strict typing, security (OWASP Top 10), PSR standards
- TypeScript -- Type system, generics, utility types, async patterns, decorators
- Vue 2/3 and Svelte 5 -- Composition API, runes, stores, TypeScript integration
- Frontend Design -- Animation engineering (timing, easing, springs), UI audit and redesign workflow (anti-AI-slop patterns)
- Accessibility -- WCAG 2.1 AA checklist, keyboard navigation, ARIA, form accessibility
18 Coding Rules
The rules are organised as individual files in rules/ and cover: PSR-12 as universal style standard, security (input validation, output escaping, CSRF, rate limiting), strict type system, clean code principles, low complexity, maintainability, JS/TS tooling, Vue/Svelte conventions, testing (no real API calls in unit tests), environment (DDEV containers), boundaries, performance (no N+1 queries), database migrations, dependency management, observability, accessibility, i18n, and git workflow (conventional commits).
Safety Hooks
Four hooks prevent common mistakes automatically: format-on-save, static analysis before commit, type checking before build, and a destructive command guard that warns before potentially dangerous operations.
MCP Servers
The settings.json configures five MCP servers: context7 (up-to-date library documentation), Playwright for Chromium and Firefox (browser automation and E2E tests), MySQL (read-only MariaDB queries), and GitHub (API for issues, PRs, repos).
Installation
git clone https://github.com/alpham8/claude-webdev.git
cd claude-webdev
./install.sh
The installer copies all files to ~/.claude and configures Claude Code for immediate use. It works on Linux and macOS.
Why Open Source?
I use Claude Code daily for client projects. The quality of AI output depends directly on how good the instructions are. With this repository, any PHP and web developer can benefit from my configuration -- and adapt it for their own stack.
The repository is deliberately modular: you can pick individual rules or skills without adopting the entire setup. The CLAUDE.md references rules via @rules/ syntax, so Claude Code loads them automatically.
Feedback and pull requests are welcome: github.com/alpham8/claude-webdev
Kommentare
Kommentare werden von Remark42 bereitgestellt. Beim Laden werden Daten an unseren Kommentar-Server übertragen.