Token导航 LogoToken导航TokenDH.com
研究检索需要联网github未标认证来源可访问clear审计未展示

claude-rules-authoringClaude rules authoring 搜索

Agent Skill

用于辅助安全审计、权限检查、凭据风险、认证流程和常见漏洞排查。它适合让 Agent 梳理敏感配置、检查依赖风险、分析鉴权逻辑或生成安全复核清单。使用时不能把工具输出直接当最终结论,涉及密钥、令牌、用户数据或生产系统时,应先确认最小权限、脱敏方式和操作边界。

总安装

643

周安装

26

GitHub Stars

公开资料未说明

下载量

202
CodexClaudeCursorGemini CLI

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

2

许可证

MIT

最后核验

2026-05-01

来源状态

来源可访问

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:claude-rules-authoring(Claude rules authoring 搜索)
来源仓库:https://github.com/outfitter-dev/agents
仓库路径:skills/claude-rules-authoring
安装命令:
npx skills add outfitter-dev/agents --skill "claude-rules-authoring"
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

AgentSkills.tonpx skills
npx skills add outfitter-dev/agents --skill "claude-rules-authoring"

简介

该技能用于安全规则编写与审计辅助,识别常见漏洞与权限问题。

  • 适合梳理敏感配置并生成合规性检查清单。claude-rules-authoring 属于研究检索类 Skill,可作为该场景下的辅助能力补充。
  • 不能单独作为安全结论依据,需结合人工复核与测试验证。
  • 涉及生产系统时应启用最小权限模式与操作日志追踪。
  • 建议定期更新规则库以应对新出现的威胁模式。

SKILL.md

Claude Rules Authoring

Create reusable instruction files in .claude/rules/ for project conventions.

Rules vs CLAUDE.md

AspectCLAUDE.md.claude/rules/
LoadingAutomatic at session startOn-demand via reference
ContentProject setup, key commandsReusable conventions
SizeConcise (~200-500 lines)Can be detailed
ScopeThis specific projectPatterns across files

Put in CLAUDE.md: One-off instructions, project-specific commands, key file locations.

Put in rules/: Formatting conventions, architecture patterns, workflow guidelines, commit standards.

File Conventions

Naming

  • UPPERCASE.md - All caps with .md extension
  • Topic-focused - One concern per file
  • Kebab-case for multi-word - API-PATTERNS.md, CODE-REVIEW.md

Good: FORMATTING.md, TESTING.md, COMMITS.md Bad: formatting.md, MyRules.md, everything.md

Structure

.claude/rules/
├── FORMATTING.md      # Code style, output conventions
├── TESTING.md         # Test patterns, coverage requirements
├── COMMITS.md         # Commit message format, PR conventions
├── ARCHITECTURE.md    # Component structure, file organization
└── SECURITY.md        # Security guidelines, auth patterns

Content Structure

Rules files should be scannable and actionable:

# Topic Name

Brief description of what this covers.

## Section 1

| Pattern | Example | Notes |
|---------|---------|-------|
| ... | ... | ... |

## Section 2

**Do:**
- Specific guideline

**Don't:**
- Anti-pattern to avoid

## Examples

{ concrete examples }

Referencing Rules

From CLAUDE.md

Reference rules explicitly - they're not auto-loaded:

# CLAUDE.md

## Code Style
Follow `.claude/rules/FORMATTING.md` for all code conventions.

## Testing
See `.claude/rules/TESTING.md` for TDD patterns.

Cross-file References

Use @ syntax to include content from other files:

# .claude/rules/FORMATTING.md

@../../baselayer/shared/rules/FORMATTING.md

This keeps rules DRY by pointing to authoritative sources.

Plugin Shared Rules

Plugins can provide shared rules in shared/rules/:

my-plugin/
├── .claude-plugin/
│   └── plugin.json
├── skills/
└── shared/
    └── rules/
        ├── FORMATTING.md
        └── PATTERNS.md

Projects reference via relative paths or @ includes.

Quick Start

Create a New Rule

  1. Identify the convention scope (formatting, testing, commits, etc.)
  2. Create TOPIC.md in .claude/rules/
  3. Write scannable content with tables, do/don't lists
  4. Reference from CLAUDE.md

Validate a Rule

  • Filename is UPPERCASE.md
  • Single focused topic
  • Scannable structure (tables, lists)
  • Referenced from CLAUDE.md
  • No duplicate content with CLAUDE.md

Anti-Patterns

Don't:

  • Create rules for one-off instructions (use CLAUDE.md)
  • Duplicate content between CLAUDE.md and rules/
  • Create catch-all files like EVERYTHING.md
  • Expect rules to auto-load (they must be referenced)

Do:

  • Keep each rule file focused on one topic
  • Use tables and lists for scannability
  • Reference shared rules via @ when available
  • Document why, not just what

Related Skills

  • claude-code-configuration - CLAUDE.md and settings.json
  • claude-plugin-development - Plugin structure including shared/rules/

适合场景

01

用户想查找某类 Agent Skill 时

02

需要根据任务场景推荐可安装能力包时

03

需要对比不同来源的安装命令和来源信息时

04

需要参考平台分布和安装热度时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

补充不同宿主或平台的使用分布数据

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

Claude Code

30.41%
按下载量换算61

windsurf

22.52%
按下载量换算45

trae

18.16%
按下载量换算37

OpenCode

14.46%
按下载量换算29

Cursor

7.66%
按下载量换算15

Codex

3.88%
按下载量换算8

安全审计

暂无安全审计结果可展示。

权限和风险

需要联网

该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。当前只有一个来源,正式发布前建议补源仓库或其他目录站核验。

来源信息

继续浏览同类 Skills