Token导航 LogoToken导航TokenDH.com
研究检索external-servicegithub未标认证来源可访问clear审计提醒

claude-docs-consultantClaude 文档 consultant

Agent Skill

用于辅助文档、README、Markdown、说明文和内容稿件的整理与改写。它适合让 Agent 提炼结构、补齐章节、统一术语、检查链接或把零散材料整理成可读文档。使用时应保留项目已有事实、命令和路径,不要把未确认的信息写成确定结论;涉及对外文案时,还需要控制语气,避免过度营销或夸大能力。

总安装

218

周安装

9

GitHub Stars

2

下载量

71
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:claude-docs-consultant(Claude 文档 consultant)
来源仓库:https://github.com/zpankz/mcp-skillset
仓库路径:skills/claude-docs-consultant
安装命令:
npx skills add https://github.com/zpankz/mcp-skillset --skill claude-docs-consultant
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/zpankz/mcp-skillset --skill claude-docs-consultant

简介

用于高效查阅 Claude Code 官方文档与扩展开发资料。

  • 适合创建钩子、调试技能或理解子代理参数时使用。
  • 使用时需按需拉取特定文档而非加载全部内容。
  • 涉及对外发布时应避免夸大功能描述与营销用语。
  • claude-docs-consultant 属于研究检索类 Skill,可作为该场景下的辅助能力补充。

SKILL.md

Claude Docs Consultant

Overview

This skill enables efficient consultation of official Claude Code documentation by fetching only the specific docs needed for the current task. Instead of loading all documentation upfront, determine which docs are relevant and fetch them on-demand.

When to Use This Skill

Invoke this skill when:

  • Creating or modifying Claude Code hooks
  • Building or debugging skills
  • Working with subagents or understanding subagent parameters
  • Implementing MCP server integrations
  • Understanding any Claude Code feature that requires official documentation
  • Troubleshooting Claude Code functionality
  • Verifying correct API usage or parameters

Common Documentation

For the most frequently referenced topics, fetch these detailed documentation files directly:

Hooks Documentation

  • hooks-guide.md - Comprehensive guide to creating hooks with examples and best practices

- URL: https://docs.claude.com/en/docs/claude-code/hooks-guide.md - Use for: Understanding hook lifecycle, creating new hooks, examples

  • hooks.md - Hooks API reference with event types and parameters

- URL: https://docs.claude.com/en/docs/claude-code/hooks.md - Use for: Hook event reference, available events, parameter details

Skills Documentation

  • skills.md - Skills creation guide and structure reference

- URL: https://docs.claude.com/en/docs/claude-code/skills.md - Use for: Creating skills, understanding SKILL.md format, bundled resources

Subagents Documentation

  • sub-agents.md - Subagent types, parameters, and usage

- URL: https://docs.claude.com/en/docs/claude-code/sub-agents.md - Use for: Available subagent types, when to use Task tool, subagent parameters

Workflow for Selective Fetching

Follow this process to efficiently fetch documentation:

Step 1: Identify Documentation Needs

Determine which documentation is needed based on the task:

  • Hook-related task → Fetch hooks-guide.md and/or hooks.md
  • Skill-related task → Fetch skills.md
  • Subagent-related task → Fetch sub-agents.md
  • Other Claude Code feature → Proceed to Step 2

Step 2: Discover Available Documentation (If Needed)

For features not covered by the 4 common docs above, fetch the docs map to discover available documentation:

URL: https://docs.claude.com/en/docs/claude-code/claude_code_docs_map.md

The docs map lists all available Claude Code documentation with descriptions. Identify the relevant doc(s) from the map.

Step 3: Fetch Only Relevant Documentation

Use WebFetch to retrieve only the specific documentation needed:

WebFetch:
  url: https://docs.claude.com/en/docs/claude-code/[doc-name].md
  prompt: "Extract the full documentation content"

Fetch multiple docs in parallel if the task requires information from several sources.

Step 4: Apply Documentation to Task

Use the fetched documentation to:

  • Verify correct API usage
  • Understand available parameters and options
  • Follow best practices and examples
  • Implement the feature correctly

Examples

Example 1: Creating a New Hook

User request: "Help me create a pre-tool-use hook to log all tool calls"

Process:

  1. Identify need: Hook creation requires hooks documentation
  2. Fetch hooks-guide.md for creation process and examples
  3. Fetch hooks.md for pre-tool-use event reference
  4. Apply: Create hook following guide, using correct event parameters

Example 2: Debugging a Skill

User request: "My skill isn't loading - help me fix SKILL.md"

Process:

  1. Identify need: Skill structure requires skills documentation
  2. Fetch skills.md for SKILL.md format requirements
  3. Apply: Validate frontmatter, structure, and bundled resources

Example 3: Using Subagents

User request: "Which subagent should I use to search the codebase?"

Process:

  1. Identify need: Subagent selection requires subagent documentation
  2. Fetch sub-agents.md for subagent types and capabilities
  3. Apply: Recommend appropriate subagent (e.g., Explore or code-searcher)

Example 4: Unknown Feature

User request: "How do I configure Claude Code settings.json?"

Process:

  1. Identify need: Not covered by the 4 common docs
  2. Fetch docs map: claude_code_docs_map.md
  3. Discover: Find relevant doc (e.g., settings.md)
  4. Fetch specific doc: https://docs.claude.com/en/docs/claude-code/settings.md
  5. Apply: Configure settings.json correctly

Best Practices

Token Efficiency

  • Fetch only the documentation actually needed for the current task
  • Fetch multiple docs in parallel if needed (single message with multiple WebFetch calls)
  • Do not fetch documentation "just in case" - fetch when required

Staying Current

  • Always fetch from docs.claude.com (live docs, not cached copies)
  • Documentation may be updated by Anthropic - fetching ensures latest information
  • If documentation seems outdated or unclear, verify URL is correct

Selective vs Comprehensive

  • Selective (preferred): Fetch hooks-guide.md for hook creation task
  • Comprehensive (avoid): Fetch all 4 common docs for every task
  • Discovery-based: Use docs map when common docs don't cover the need

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

OpenCode

30.09%
按下载量换算21

Claude Code

25.76%
按下载量换算18

windsurf

17.19%
按下载量换算12

Codex

13.26%
按下载量换算9

kiro-cli

8.22%
按下载量换算6

mcpjam

3.27%
按下载量换算2

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

可疑

权限和风险

external-service

该 Skill 可能调用第三方服务、云服务或外部模型 API,使用前需要确认账号、额度、数据发送范围和服务条款。

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