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审计异常

reference-generator参考发生器

Agent Skill

reference-generator 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

642

周安装

27

GitHub Stars

65

下载量

225
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:reference-generator(参考发生器)
来源仓库:https://github.com/dmccreary/claude-skills
仓库路径:skills/reference-generator
安装命令:
npx skills add https://github.com/dmccreary/claude-skills --skill reference-generator
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/dmccreary/claude-skills --skill reference-generator

简介

reference-generator 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。

  • 适用于研究检索类任务,如信息搜集、资料筛选和知识整理,尤其适合多源数据聚合与初步分析。
  • 通过关键词、任务描述或来源仓库提供查询条件,返回结构化候选结果列表供进一步处理。
  • 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。
  • 可结合来源仓库、安装命令和原始 README 继续核验具体用法。

SKILL.md

Reference Generator

Overview

Generate high-quality, curated reference lists for educational textbooks. Each chapter receives exactly 10 references, prioritizing Wikipedia articles first for reliability, followed by authoritative online resources. References are stored in separate references.md files for token-efficient maintenance by AI agents.

When to Use This Skill

Use this skill when:

  • Creating a new intelligent textbook that needs chapter references
  • Adding references to an existing textbook
  • Updating or expanding references for educational content
  • A user explicitly requests reference generation

Key Principles

10 References Per Chapter

Every chapter receives exactly 10 high-quality references:

  • References 1-3: Wikipedia articles (most reliable, always accessible)
  • References 4-5: Authoritative textbooks (title, author, publisher - no URLs that break)
  • References 6-10: Online resources (tutorials, courses, verified working links)

Wikipedia First

Wikipedia links are placed first because they:

  • Have stable, reliable URLs that rarely break
  • Provide comprehensive, well-sourced overviews
  • Are freely accessible to all students
  • Cover most technical topics thoroughly
  • Are maintained and updated by the community

Separate Reference Files

References are stored in separate references.md files (not inline in chapters) for:

  • Token efficiency: ~150 tokens vs ~6,000 tokens to read/update
  • Batch processing: Agents can glob **/references.md for bulk updates
  • Separation of concerns: Chapter edits don't risk breaking references

Reference Generation Workflow

Step 1: Analyze the Course Description

Read the /docs/course-description.md file to determine:

  • Subject matter - determines reference topics
  • Target audience - guides complexity of descriptions
  • Learning objectives - ensures references support goals

Step 2: Identify Chapter Structure

Locate all chapter directories:

# Find chapter folders
ls docs/chapters/

For each chapter, read the chapter index.md to understand:

  • Chapter title and topic
  • Key concepts covered
  • Learning objectives

Step 3: Generate 10 References Per Chapter

For each chapter, create exactly 10 references following this priority:

Positions 1-3: Wikipedia Articles

Find the most relevant Wikipedia articles for the chapter's main concepts:

1. [Concept Name](https://en.wikipedia.org/wiki/Concept_Name) - Wikipedia - Description of article content and relevance to chapter.

Positions 4-5: Textbooks (No URL)

Reference authoritative textbooks without URLs (which often break):

4. Textbook Title (Edition) - Author Name - Publisher - Description of relevant chapters and why this textbook is valuable.

Positions 6-10: Online Resources

Verified online tutorials, courses, and educational sites:

6. [Resource Title](https://verified-url.com/path) - Source Name - Description of content and specific relevance to chapter topics.

Step 4: Verify Online URLs

For references 6-10, verify each URL:

WebFetch(url=reference_url, prompt="Is this page accessible? What is the main topic?")

Preferred sources (stable, educational):

  • All About Circuits (allaboutcircuits.com)
  • Electronics Tutorials (electronics-tutorials.ws)
  • ChipVerify (chipverify.com)
  • HDLBits (hdlbits.01xz.net)
  • Nandland (nandland.com)
  • MIT OpenCourseWare (ocw.mit.edu)
  • GeeksforGeeks (geeksforgeeks.org)
  • Digilent Reference (digilent.com/reference)

Step 5: Write Reference Files

Create a references.md file in each chapter directory:

File location: docs/chapters/XX-chapter-name/references.md

File format:

# References: [Chapter Title]

1. [Wikipedia Article](https://en.wikipedia.org/wiki/Topic) - Wikipedia - Comprehensive overview of [topic] including [specific aspects relevant to chapter].

2. [Wikipedia Article 2](https://en.wikipedia.org/wiki/Topic2) - Wikipedia - Detailed explanation of [concept] with [relevant details].

3. [Wikipedia Article 3](https://en.wikipedia.org/wiki/Topic3) - Wikipedia - Coverage of [related topic] including [applications/examples].

4. Textbook Title (Edition) - Author Name - Publisher - Chapter X covers [specific topics] with [teaching approach/examples].

5. Textbook Title 2 - Author Name - Publisher - Provides [specific value] for understanding [chapter concepts].

6. [Online Resource](https://verified-url.com) - Source - Tutorial on [topic] with [specific features like examples, exercises, visualizations].

7. [Online Resource 2](https://verified-url.com) - Source - [Description of content and relevance].

8. [Online Resource 3](https://verified-url.com) - Source - [Description of content and relevance].

9. [Online Resource 4](https://verified-url.com) - Source - [Description of content and relevance].

10. [Online Resource 5](https://verified-url.com) - Source - [Description of content and relevance].

Step 6: Update Chapter Files

Add a link at the end of each chapter's index.md:

[See Annotated References](./references.md)

Important: Replace any existing ## References section with this single link.

Step 7: Update mkdocs.yml Navigation

Add references to the navigation for each chapter:

- Chapters:
  - 1. Chapter Name:
    - Content: chapters/01-chapter-folder/index.md
    - Quiz: chapters/01-chapter-folder/quiz.md
    - Annotated References: chapters/01-chapter-folder/references.md

Reference Quality Guidelines

Description Requirements

Each reference description should:

  • Explain what the resource covers (1 sentence)
  • Explain why it's relevant to this chapter (1 sentence)
  • Mention specific features (examples, exercises, visualizations)
  • Be 20-40 words total

Good example:

1. [Karnaugh map](https://en.wikipedia.org/wiki/Karnaugh_map) - Wikipedia - Detailed explanation of K-map theory, grouping rules, and don't-care conditions. Essential foundation for the simplification techniques covered in this chapter.

Bad example:

1. [Karnaugh map](https://en.wikipedia.org/wiki/Karnaugh_map) - Wikipedia - About K-maps.

Wikipedia Article Selection

Choose Wikipedia articles that:

  • Cover the chapter's primary concepts
  • Have substantial content (not stubs)
  • Include diagrams, examples, or formulas
  • Link to related topics students might explore

Textbook Selection

Reference textbooks that:

  • Are widely used in the field
  • Have strong reputations
  • Cover the chapter's topics in depth
  • Are reasonably accessible to students

Online Resource Selection

Choose online resources that:

  • Have verified, working URLs
  • Come from educational or reputable technical sources
  • Provide practical examples or exercises
  • Complement (not duplicate) Wikipedia content

Reference Quality Checklist

Before finalizing references, ensure:

  • Exactly 10 references per chapter
  • References 1-3 are Wikipedia articles
  • References 4-5 are textbooks (no URLs)
  • References 6-10 have verified working URLs
  • All descriptions are 20-40 words
  • Descriptions explain relevance to chapter
  • No duplicate sources across references
  • Proper formatting throughout

File Structure After Generation

docs/chapters/
├── 01-chapter-name/
│   ├── index.md          # Ends with: [See Annotated References](./references.md)
│   ├── quiz.md
│   └── references.md     # 10 curated references
├── 02-chapter-name/
│   ├── index.md
│   ├── quiz.md
│   └── references.md
...

Token Efficiency Benefits

TaskInline ReferencesSeparate Files
Read one chapter's refs~6,000 tokens~200 tokens
Update all 15 chapters~90,000 tokens~3,000 tokens
Batch reference checkNot feasibleGlob + process

Finish

After generating references:

  1. Report the number of chapters processed
  2. List any URLs that failed verification
  3. Confirm mkdocs.yml was updated
  4. Remind user to verify the site builds: mkdocs serve

Resources

This skill uses web-based verification tools built into Claude Code:

  • WebSearch: Find authoritative sources on topics
  • WebFetch: Verify URLs are accessible and extract metadata
  • Glob: Find all chapter directories and reference files
  • Write: Create reference files
  • Edit: Update chapter files and mkdocs.yml

No additional scripts or assets are required for this skill.

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

Codex

32.3%
按下载量换算73

windsurf

21.8%
按下载量换算49

Claude Code

16.41%
按下载量换算37

OpenCode

13.03%
按下载量换算29

Gemini CLI

8.97%
按下载量换算20

Antigravity

3.94%
按下载量换算9

安全审计

Gen Agent Trust Hub

未通过

Socket

通过

Snyk

可疑

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