loaditout.ai
SkillsPacksTrendingLeaderboardAPI DocsBlogSubmitRequestsCompareAgentsXPrivacyDisclaimer
{}loaditout.ai
Skills & MCPPacksBlog

claude-rules

MCP Tool

pegregenerate417/claude-rules

Build project-specific coding rules for AI assistants with base, language, and framework layers to keep generated code clean and consistent

Install

$ npx loaditout add pegregenerate417/claude-rules

Platform-specific configuration:

.claude/settings.json
{
  "mcpServers": {
    "claude-rules": {
      "command": "npx",
      "args": [
        "-y",
        "claude-rules"
      ]
    }
  }
}

Add the config above to .claude/settings.json under the mcpServers key.

About

Claude Rules

[](LICENSE) [](https://github.com/lifedever/claude-rules/stargazers) [](https://github.com/lifedever/claude-rules/pulls) [](https://lifedever.github.io/claude-rules/)

English | 中文

Coding standards template library for AI coding assistants. Combine base + language + framework layers to generate a project-specific rules file that keeps AI-generated code clean and consistent.

The Problem

AI coding assistants tend to mimic existing code style in legacy projects — including bad habits. The core principle of this library:

> Don't imitate legacy code. Refactor according to the standards.

Every rule is a concrete, actionable directive (not vague "use proper XX"), with "Bad / Good" code comparison examples.

Directory Structure
claude-rules/
├── base/                   # Universal (required)
│   ├── core.md             # Core principles: legacy code attitude, quality metrics, naming, architecture
│   └── git.md              # Git commit message conventions
│
├── languages/              # Pick by language
│   ├── typescript.md       # No any/enum/barrel exports, as const, import type
│   ├── javascript.md       # ES2022+, JSDoc type annotations, ESM only
│   ├── java.md             # Java 17+ record/sealed/pattern matching, Optional
│   ├── kotlin.md           # Null safety, structured concurrency, sealed class
│   ├── swift.md            # guard let, async/await, actor, Protocol
│   ├── python.md           # ruff, type annotations, Protocol, uv/poetry
│   ├── html.md             # Semantic tags, accessibility, no div soup
│   ├── css.md              # Custom properties, Flexbox/Grid, BEM, modern features
│   ├── go.md               # Error 

Tags

agentai-toolsbuild-in-publicclaude-apiclaude-code-marketplaceclaude-desktopclaude-promptsclaude-skillsclaudecodeclinecodex-clillmmcprulessecurity

Reviews

Loading reviews...

Quality Signals

0
Installs
Last updated15 days ago
Security: AREADME

Safety

Risk Levelmedium
Data Access
read
Network Accessnone

Details

Sourcegithub-crawl
Last commit4/2/2026
View on GitHub→

Embed Badge

[![Loaditout](https://loaditout.ai/api/badge/pegregenerate417/claude-rules)](https://loaditout.ai/skills/pegregenerate417/claude-rules)