Token导航 LogoToken导航TokenDH.com
研究检索只读github未标认证来源可访问许可证需确认审计提醒

best-practice-guide最佳实践指南

Agent Skill

best-practice-guide 用于记录任务执行中的错误、用户纠正、经验和能力缺口,适合在 Codex、Claude、Cursor、Gemini CLI 中希望让 Agent 持续沉淀问题、修正和最佳实践时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

总安装

570

周安装

24

GitHub Stars

1

下载量

38
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

unknown

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:best-practice-guide(最佳实践指南)
来源仓库:https://github.com/wojons/skills
仓库路径:skills/best-practice-guide
安装命令:
npx skills add https://github.com/wojons/skills --skill best-practice-guide
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。

skills.shnpx skills
npx skills add https://github.com/wojons/skills --skill best-practice-guide

简介

best-practice-guide 分析项目文档缺口,生成可落地的最佳实践指南。

  • 适用于新项目启动、知识断层填补或复杂任务前的准备工作。
  • 结合网络搜索获取权威资料,输出结构化指导或独立技能模块。
  • 使用时应限定在公开信息范围内,避免处理敏感配置或私有数据。
  • 建议定期更新指南以反映最新行业标准和内部实践经验。

SKILL.md

Best Practice Guide Generator

Analyze project documentation (AGENTS.md, CLAUDE.md, specifications) to identify missing context, critical problems, and knowledge gaps. Use web search when available to find authoritative best practices, then generate and store guides as either documentation links or standalone skills for easy agent discovery.

When to use me

Use this skill when:

  • Starting work on a new project with incomplete documentation
  • Notice repeated questions or confusion about specific topics
  • Preparing for complex tasks requiring specialized knowledge
  • Building comprehensive project onboarding materials
  • Identifying areas where agents lack sufficient context
  • Creating reusable knowledge assets for future agent sessions
  • Validating project practices against industry standards

What I do

1. Documentation Analysis & Gap Detection

  • Scan project files for key documentation (AGENTS.md, CLAUDE.md, README.md, SPECS.md, etc.)
  • Identify knowledge gaps by analyzing:

- Missing or incomplete explanations - Undocumented assumptions and conventions - Areas requiring specialized domain knowledge - Security, performance, or compliance considerations not addressed - Tool-specific configurations not documented

  • Map dependencies between documented and undocumented areas
  • Prioritize gaps by impact on agent effectiveness and project success

2. Research-Driven Best Practice Discovery

  • Use available search tools (searxng_searxng_web_search, searxng_web_url_read) when possible
  • Find authoritative sources for identified gaps:

- Official documentation and style guides - Industry best practice articles and whitepapers - Framework/library-specific recommendations - Security compliance requirements - Performance optimization patterns

  • Cross-reference multiple sources to validate information
  • Document sources for traceability and credibility

3. Guide Generation & Storage

  • Create comprehensive guides addressing identified gaps
  • Structure guides for easy consumption by both agents and humans
  • Generate multiple output formats:

- Documentation links: Add to AGENTS.md/CLAUDE.md with clear sections - Standalone skills: Create new skills in the skills directory for agent discovery - Reference files: Store in references/ or guides/ directories

  • Include practical examples and implementation patterns
  • Add verification steps to confirm guide effectiveness

4. Integration & Maintenance

  • Update project documentation with guide references
  • Create skill metadata for easy agent discovery
  • Establish maintenance procedures for keeping guides current
  • Set up validation checks to ensure guide accuracy over time
  • Document guide creation process for future updates

Output Strategy

Option 1: Documentation Integration (When guides are project-specific)

  • Add to AGENTS.md: New sections with clear headings and references
  • Update CLAUDE.md: Context-specific guidance for Claude Code
  • Create docs/guides/ directory: Organized by topic
  • Link from existing documentation: Cross-references between related topics

Option 2: Skill Creation (When guides are reusable across projects)

  • Create new skill directory in skills/<topic>-guide/
  • Include SKILL.md with guide content and usage instructions
  • Add references/README.md with detailed explanations
  • Create scripts/ for practical implementation
  • Register in skills catalog for agent discovery

Option 3: Hybrid Approach

  • Core best practices as reusable skills
  • Project-specific adaptations in project documentation
  • Cross-references between skills and project docs

Examples

# Analyze current project documentation for gaps
npm run best-practice:analyze -- --project-dir . --output gaps.json

# Generate guide for identified React performance gaps
npm run best-practice:generate -- --topic "react-performance" --format skill --search true

# Create documentation links for identified security gaps
npm run best-practice:integrate -- --gaps gaps.json --target AGENTS.md --format documentation

# Comprehensive analysis and guide generation
npm run best-practice:full -- --project-dir . --search true --output-formats "skill,documentation"

Output format

Gap Analysis Report:

project: example-project
analysis_date: 2026-02-26
documents_analyzed:
  - AGENTS.md
  - CLAUDE.md
  - README.md
  - package.json
critical_gaps:
  - topic: "database-migration-best-practices"
    impact: "High (data loss risk)"
    evidence: "No documentation on migration procedures"
    sources_needed: ["Official framework docs", "Industry patterns"]
  - topic: "authentication-security-considerations"
    impact: "Critical (security risk)"
    evidence: "Authentication mentioned but no security guidelines"
    sources_needed: ["OWASP guidelines", "Framework security docs"]
  - topic: "performance-monitoring-setup"
    impact: "Medium (operational visibility)"
    evidence: "Monitoring mentioned but no implementation details"
    sources_needed: ["Monitoring tool docs", "SRE practices"]
research_required: 3 topics
search_tools_available: true

Generated Guide (as Skill):

skills/database-migration-guide/
├── SKILL.md
├── references/
│   └── README.md (with detailed migration patterns)
└── scripts/
    └── safe-migration.sh

Documentation Integration:

## Database Migration Best Practices

[See skills/database-migration-guide/ for comprehensive guidance]

Key principles:
1. Always backup before migration
2. Use transactional migrations where possible
3. Test migrations in staging environment
4. Monitor performance during migration

*Generated by best-practice-guide skill - Last updated: 2026-02-26*

Notes

  • Research is critical: Always use available search tools to find authoritative sources
  • Balance specificity and reusability: Create guides that are useful for current project but potentially reusable
  • Document sources: Always cite sources for credibility and future reference
  • Prioritize by impact: Focus on high-impact gaps first (security, data integrity, critical functionality)
  • Validate generated content: Test guides for clarity and usefulness
  • Maintain over time: Schedule periodic review and updates of generated guides
  • Respect intellectual property: Properly attribute sources and follow licensing requirements
  • Adapt to project maturity: Different guidance needed for POC vs production projects
  • Consider agent capabilities: Create guides that work within agent context limits

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

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

展示第三方安全扫描或审计结果

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

Codex

32.79%
按下载量换算12

Claude

29.09%
按下载量换算11

Cursor

20.23%
按下载量换算8

Gemini CLI

10.33%
按下载量换算4

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

可疑

权限和风险

只读

该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

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。来源安全扫描存在 warning/failed 结果,不能写成本站确认安全。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