Token导航 LogoToken导航TokenDH.com
开发规范需要联网github未标认证来源可访问clear审计通过

best-practices最佳实践

Agent Skill

best-practices 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

297

周安装

12

GitHub Stars

160

下载量

93
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

3

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:best-practices(最佳实践)
来源仓库:https://github.com/yonatangross/orchestkit
仓库路径:skills/best-practices
安装命令:
npx skills add https://github.com/yonatangross/orchestkit --skill best-practices
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。不同来源提供的安装方式可能略有差异;本站展示可直接复制的安装命令,安装前请核对来源页面。

skills.shnpx skills
npx skills add https://github.com/yonatangross/orchestkit --skill best-practices

简介

用于记录任务执行中的错误、用户纠正和经验缺口,适合持续优化 Agent 能力。

  • 可沉淀问题、修正和改进最佳实践,支持多宿主环境使用。
  • 通过 npx skills add 命令从 GitHub 仓库安装,需结合原始 README 确认具体用法。
  • 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操作。
  • best-practices 属于开发规范类 Skill,可作为该场景下的辅助能力补充。

SKILL.md

Best Practices - View Your Pattern Library

Display your aggregated best practices library, showing successful patterns and anti-patterns across all projects.

Usage

/best-practices                     # Show full library
/best-practices <category>          # Filter by category
/best-practices --warnings          # Show only anti-patterns
/best-practices --successes         # Show only successes
/best-practices --stats             # Show statistics only

Options

  • <category> - Filter by specific category (pagination, database, authentication, etc.)
  • --warnings - Show only anti-patterns (failed patterns)
  • --successes - Show only successful patterns
  • --stats - Show statistics summary without individual patterns

Workflow

1. Query mem0 for Best Practices

Use the mem0 CLI script:

python3 ${CLAUDE_PLUGIN_ROOT}/src/skills/mem0-memory/scripts/crud/search-memories.py \
  --query "patterns outcomes" \
  --user-id "project-decisions" \
  --limit 100

2. Aggregate Results

Group patterns by category, then by outcome:

{
  "pagination": {
    "successes": [...],
    "failures": [...]
  },
  "authentication": {
    "successes": [...],
    "failures": [...]
  }
}

3. Calculate Statistics

For each pattern:

  • Count occurrences across projects
  • Calculate success rate: successes / (successes + failures)
  • Note which projects contributed

4. Display Output

Full Library View:

📚 Your Best Practices Library
═══════════════════════════════════════════════════════════════

PAGINATION
─────────────────────────────────────────────────────────────
  ✅ Cursor-based pagination (3 projects, always worked)
     "Scales well for large datasets"

  ❌ Offset pagination (failed in 2 projects)
     "Caused timeouts on tables with 1M+ rows"
     💡 Lesson: Use cursor-based for large datasets

AUTHENTICATION
─────────────────────────────────────────────────────────────
  ✅ JWT + httpOnly refresh tokens (4 projects)
     "Secure and scalable for web apps"

  ⚠️ Session-based auth (mixed: 1 success, 1 failure)
     "Works but scaling issues in high-traffic scenarios"

───────────────────────────────────────────────────────────────
📊 Summary: 8 patterns | 5 ✅ successes | 3 ❌ anti-patterns
💡 Use `/remember --success` or `/remember --failed` to add more

Stats Only View (--stats):

📊 Best Practices Statistics
═══════════════════════════════════════════════════════════════

Total Patterns: 15
├── ✅ Successful: 10 (67%)
├── ❌ Anti-patterns: 5 (33%)
└── ⚠️ Mixed: 2

Categories:
├── pagination: 3 patterns (2 ✅, 1 ❌)
├── authentication: 4 patterns (3 ✅, 1 ⚠️)
├── database: 5 patterns (4 ✅, 1 ❌)
└── api: 3 patterns (1 ✅, 2 ❌)

Projects Contributing: 7
Last Updated: 2 days ago

Filtered View (by category):

📚 Best Practices: PAGINATION
═══════════════════════════════════════════════════════════════

  ✅ Cursor-based pagination (3 projects, always worked)
     "Scales well for large datasets"
     Projects: project-a, project-b, project-c

  ❌ Offset pagination (failed in 2 projects)
     "Caused timeouts on tables with 1M+ rows"
     💡 Lesson: Use cursor-based for large datasets
     Projects: project-a, project-d

Pattern Confidence Indicators

IconMeaning
Strong success (3+ projects, 100% success rate)
Moderate success (1-2 projects or some failures)
⚠️Mixed results (both successes and failures)
Anti-pattern (only failures)
🔴Strong anti-pattern (3+ projects, all failed)

Empty Library

📚 Your Best Practices Library is empty

Start building it with:
• /remember --success "Pattern that worked well"
• /remember --failed "Pattern that caused problems"

Your patterns will be tracked across all projects and help
Claude warn you before repeating past mistakes.

Proactive Integration

See references/proactive-warnings.md for automatic anti-pattern detection.

Related Skills

  • code-review-playbook: Review best practices
  • api-design-framework: API design best practices
  • testing-strategy: Testing best practices
  • security-hardening: Security best practices

Related Commands

  • /remember --success <text> - Add a successful pattern
  • /remember --failed <text> - Add an anti-pattern
  • /ork:memory search <query> - Search all memories (not just best practices)

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

能力 5

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

Claude Code

31.52%
按下载量换算29

Gemini CLI

23.29%
按下载量换算22

Antigravity

18.09%
按下载量换算17

windsurf

11.68%
按下载量换算11

trae

8.09%
按下载量换算8

OpenCode

3.65%
按下载量换算3

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

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。

来源信息

继续浏览同类 Skills