Token导航 LogoToken导航TokenDH.com
研究检索敏感数据github未标认证来源可访问clear审计异常

create-master-skill创造大师技能

Agent Skill

create-master-skill 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。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

288

周安装

12

GitHub Stars

2

下载量

96
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:create-master-skill(创造大师技能)
来源仓库:https://github.com/abdullahbeam/nexus-design-abdullah
仓库路径:skills/create-master-skill
安装命令:
npx skills add https://github.com/abdullahbeam/nexus-design-abdullah --skill create-master-skill
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/abdullahbeam/nexus-design-abdullah --skill create-master-skill

简介

用于查找、检索和筛选相关信息,适合根据关键词或任务场景快速定位候选结果。

  • 适用于 Codex、Claude、Cursor、Gemini CLI 等宿主环境,支持主技能架构设计与构建。
  • 通过命令行安装,需结合来源仓库文档核验具体用法,确保项目先行与验证闭环完整。
  • 安装前建议确认权限范围,避免触发未授权的联网或文件操作。
  • create-master-skill 属于研究检索类 Skill,可作为该场景下的辅助能力补充。

SKILL.md

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ ⚠️ CRITICAL EXECUTION REQUIREMENTS ⚠️

WORKFLOW: Project FIRST, Research SECOND, Build THIRD

MANDATORY STEPS (DO NOT SKIP):

  1. ✅ Create TodoWrite with ALL phases
  2. ✅ Ask integration name (e.g., "airtable", "slack", "github")
  3. ✅ RUN create-project skill to create planning project
  4. ✅ PHASE 1: Web Research - comprehensive API/integration research
  5. ✅ PHASE 2: Architecture Design - define master skill structure
  6. ✅ PHASE 3: Build - create master skill from templates
  7. ✅ PHASE 4: Validate - test and document

ANTI-PATTERN (DO NOT DO THIS): ❌ Skip project creation ❌ Start building without research ❌ Create master skill without understanding the integration ❌ Copy from notion-master without adapting ❌ Skip validation phase

CORRECT PATTERN (DO THIS): ✅ Ask integration → Create project → Research → Design → Build → Validate ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

Create Master Skill

Build production-ready master skills (shared resource libraries) for any integration.

Purpose

Master skills solve the duplication problem: When 3+ skills share common content (setup, API docs, error handling), extract it into a master skill that all can reference.

Key Insight from notion-master:

  • 3 Notion skills had 950 lines of duplicated content
  • After extraction: 60% context reduction (950 → 370 lines)
  • Single source of truth for setup, API, errors, schemas

Quick Start

User says: "create master skill for Airtable"

AI does:

  1. Creates planning project: {ID}-airtable-master-skill
  2. Runs comprehensive web research on Airtable API
  3. Designs architecture based on research
  4. Builds master skill from templates
  5. Validates and documents

Workflow

Phase 0: Initiation

🔧 CREATE MASTER SKILL

What integration do you want to create a master skill for?
Examples: airtable, slack, github, linear, google-sheets, stripe

Integration name: ___________

After user provides name:

  1. Validate name (lowercase, hyphenated)
  2. Check if {integration}-master already exists
  3. Create planning project using create-project skill

Phase 1: Web Research (CRITICAL)

Purpose: Gather comprehensive information before building anything.

Run these searches:

## Research Tasks

1. **API Documentation**
   - Search: "{integration} API documentation official"
   - Capture: Base URLs, authentication, rate limits

2. **API Reference**
   - Search: "{integration} API endpoints reference"
   - Capture: Key endpoints, request/response formats

3. **Authentication Patterns**
   - Search: "{integration} API authentication OAuth API key"
   - Capture: Auth methods, token handling, refresh patterns

4. **Common Operations**
   - Search: "{integration} API CRUD operations examples"
   - Capture: Create, read, update, delete patterns

5. **Error Handling**
   - Search: "{integration} API error codes troubleshooting"
   - Capture: Error codes, messages, recovery patterns

6. **Rate Limits**
   - Search: "{integration} API rate limits throttling"
   - Capture: Limits, backoff strategies, best practices

7. **SDK/Libraries**
   - Search: "{integration} Python SDK library"
   - Capture: Official SDK, popular libraries, installation

8. **Best Practices**
   - Search: "{integration} API best practices integration"
   - Capture: Patterns, anti-patterns, optimization tips

Save research to: 02-projects/{ID}-{integration}-master-skill/02-resources/research.md


Phase 2: Architecture Design

Based on research, design the master skill structure:

  1. Identify Child Skills - What specific skills will use this master?

- Query/search operations - Import/export operations - Management operations - Automation/workflow operations

  1. Define Shared Resources

- What setup steps are common? - What API patterns are reused? - What errors occur across all operations? - What schemas/types need documentation?

  1. Plan Scripts

- Configuration validation script - Resource discovery script - Common utility functions - Rate limiting (if needed)

Document architecture in: 02-projects/{ID}-{integration}-master-skill/01-planning/plan.md


Phase 3: Build Master Skill

Use templates from: create-master-skill/templates/

  1. Create folder structure: 00-system/skills/{integration}-master/ ├── SKILL.md ├── references/ │ ├── setup-guide.md │ ├── api-reference.md │ ├── error-handling.md │ └── [domain-specific].md ├── scripts/ │ ├── check_{integration}_config.py │ ├── discover_resources.py │ └── [operation scripts].py └── tests/ ├── README.md └── run_tests.py
  2. Populate from templates (adapting to research)
  3. Fill in integration-specific content

Phase 4: Validate

  1. Run config check: python scripts/check_{integration}_config.py
  2. Test discovery: python scripts/discover_resources.py
  3. Verify documentation completeness
  4. Create at least one child skill that uses the master

Resources

templates/

SKILL.md.template - Master skill declaration template

setup-guide.md.template - Setup wizard template

api-reference.md.template - API documentation template

error-handling.md.template - Error handling template

check_config.py.template - Config validation script template

references/

master-skill-patterns.md - Architecture patterns from notion-master

research-checklist.md - Comprehensive research checklist


Example: Creating Airtable Master Skill

User: "create master skill for airtable"

AI: "I'll help you create an Airtable master skill. First, let me create
     a planning project to organize this work."

     → Runs create-project skill
     → Creates: 04-airtable-master-skill project

AI: "Project created. Now I'll research the Airtable API comprehensively."

     → Runs 8 web searches
     → Saves findings to research.md

AI: "Research complete. Based on findings, I recommend these child skills:
     - airtable-connect (query any base/table)
     - airtable-sync (import/export records)
     - airtable-automate (trigger-based workflows)

     The master skill will provide:
     - Setup guide (API key, base IDs)
     - API reference (REST endpoints, field types)
     - Error handling (rate limits, validation errors)
     - Scripts (config check, base discovery)

     Shall I proceed with building?"

User: "Yes"

AI: → Creates 00-system/skills/airtable-master/
    → Populates from templates
    → Adapts content based on research
    → Validates setup

Success Criteria

A master skill is complete when:

  • SKILL.md clearly says "DO NOT load directly"
  • setup-guide.md covers first-time setup end-to-end
  • api-reference.md documents key operations
  • error-handling.md covers common errors with solutions
  • check_config.py validates all required configuration
  • At least one child skill successfully references it
  • Context reduction achieved (measure before/after)

Why This Design?

Why Project First?

  • Complex work deserves proper planning
  • Research findings need a home
  • Progress tracked via project tasks
  • Validates the approach before building

Why Research First?

  • Every integration is different
  • API patterns vary significantly
  • Prevents wrong assumptions
  • Ensures comprehensive coverage

Why Templates?

  • Consistent structure across master skills
  • Proven patterns from notion-master
  • Reduces cognitive load
  • Accelerates development

Version: 1.0 Created: 2025-12-11 Based on: notion-master architecture analysis

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

32.57%
按下载量换算31

Antigravity

22.78%
按下载量换算22

windsurf

17.02%
按下载量换算16

Codex

12.72%
按下载量换算12

OpenCode

7.79%
按下载量换算7

Gemini CLI

4.04%
按下载量换算4

安全审计

Gen Agent Trust Hub

未通过

Socket

通过

Snyk

通过

权限和风险

敏感数据

该 Skill 可能接触密钥、Token、环境变量或敏感配置,应进入高风险复核队列,默认不自动发布。

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。

来源信息

继续浏览同类 Skills