Token导航 LogoToken导航TokenDH.com
开发需要联网github未标认证来源可访问许可证需确认审计通过

docs-generator文档生成器

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

1,283

周安装

54

GitHub Stars

68

下载量

449
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:docs-generator(文档生成器)
来源仓库:https://github.com/luongnv89/skills
仓库路径:skills/docs-generator
安装命令:
npx skills add https://github.com/luongnv89/skills --skill docs-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。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/luongnv89/skills --skill docs-generator

简介

docs-generator 用于辅助文档、README 和内容稿件的整理与改写。

  • 适合提炼结构、补齐章节或统一术语,保留项目已有事实与路径。
  • 通过 npx skills add 命令安装,避免将未确认信息写成确定结论。
  • 涉及对外文案时需控制语气,防止过度营销或夸大能力描述。
  • 适用宿主包括 Codex、Claude、Cursor、Gemini CLI,接入前应确认版本、权限和运行环境要求。

SKILL.md

Documentation Generator

Restructure and organize project documentation for clarity and accessibility.

Prerequisites

This skill requires (validate each before starting; if any check fails, stop and ask the user):

  • Git working tree: clean or stashable. Run git status first; if dirty, back up via git stash push -u -m "pre-docs-sync" before any sync that could rebase or overwrite local changes.
  • Tools required: git >= 2.30, a Markdown-aware editor, and mermaid-cli (or a renderer) only if diagrams must be exported. Confirm availability with git --version and command -v mmdc.
  • Repo permissions: read access for analysis; write access only when the user explicitly approves commits. For read-only repos, fall back to emitting a diff or inline summary instead of writing files.
  • Branch state: an origin remote that is reachable (git remote get-url origin); if missing, do not attempt rebase/pull — ask the user.
  • Backups: confirm the repo is pushed or otherwise backed up before any restructure that deletes or moves doc files. Pair every destructive rm/git mv with a prior git status check and explicit user confirmation; never run git reset --hard, git push --force, or rm -rf without a dry-run preview and user approval.
  • Safety defaults: prefer dry-run previews (git mv -n, rm -i) and require user confirmation before any irreversible action.

If any prerequisite fails, halt and surface the blocker to the user rather than proceeding.

Repo Sync Before Edits (mandatory)

Before creating/updating/deleting files in an existing repository, sync the current branch with remote:

branch="$(git rev-parse --abbrev-ref HEAD)"
git fetch origin
git pull --rebase origin "$branch"

If the working tree is not clean, stash first, sync, then restore:

git stash push -u -m "pre-sync"
branch="$(git rev-parse --abbrev-ref HEAD)"
git fetch origin && git pull --rebase origin "$branch"
git stash pop

If origin is missing, pull is unavailable, or rebase/stash conflicts occur, stop and ask the user before continuing.

Workflow

0. Create Feature Branch

Before making any changes:

  1. Check the current branch - if already on a feature branch for this task, skip
  2. Check the repo for branch naming conventions by running git branch -r | head -20 (e.g., feat/, feature/, etc.)
  3. Create and switch to a new branch following the repo's convention, or fallback to: feat/docs-generator

1. Analyze Project

Read the codebase to identify:

  • Project type: Library, API, web app, CLI, microservices
  • Architecture: Monorepo, multi-package, single module
  • User personas: End users, developers, operators
  • Existing docs: Scan for README files, docs/ folder, inline comments, docstrings
  • Gaps: List what documentation exists vs. what is missing

2. Restructure Documentation

Root README.md - Streamline as entry point:

  • Project overview and purpose
  • Quickstart (install + first use)
  • Modules/components summary with links
  • License and contacts

Component READMEs - Add per module/package/service:

  • Purpose and responsibilities
  • Setup instructions
  • Testing commands

Centralize in docs/ - Organize by category (select applicable):

docs/
├── architecture.md      # System design, diagrams
├── api-reference.md     # Endpoints, authentication
├── database.md          # Schema, migrations
├── deployment.md        # Production setup
├── development.md       # Local setup, contribution
├── troubleshooting.md   # Common issues
└── user-guide.md        # End-user documentation

3. Create Diagrams

Use Mermaid for all visual documentation:

  • Architecture diagrams
  • Data flow diagrams
  • Database schemas

4. Review and Validate

  1. Verify all internal links resolve correctly
  2. Check that code examples in docs are syntactically valid
  3. Confirm no orphaned docs (files not linked from anywhere)
  4. Present a summary of changes to the user before committing

Present changes to user for approval. Do not commit unless the user explicitly asks.

Expected Output

After running this skill on a mid-size Node.js API project, you should see:

  • A clean root README.md with project overview, quickstart, module links, and license
  • Per-package README.md files for each service or library
  • A docs/ folder with relevant files such as architecture.md, api-reference.md, deployment.md, and development.md
  • Mermaid diagrams embedded in architecture and data-flow docs
  • A validation summary listing all internal links checked and any gaps found

Acceptance Criteria

A run passes when all of the following are true:

  • Root README.md contains an overview, a quickstart (install + first use), and links to component or docs/ files.
  • A docs/ directory exists with at least one of: architecture.md, api-reference.md, deployment.md, development.md — and only the files relevant to the project type.
  • Every internal Markdown link resolves to an existing file (no broken [text](path) references).
  • No orphaned files: every docs/*.md is reachable from README.md or another docs/ page within one hop.
  • Mermaid diagrams in architecture or data-flow docs render without syntax errors (mmdc validation if available).
  • No commits land on main/master; all changes are on a feature branch following the repo's naming convention.

Edge Cases

  • No existing documentation: Skill generates from scratch using code analysis. Starts with README.md and adds docs/ files based on project complexity.
  • Conflicting or outdated docs: Flags conflicts to the user. Prefers code-derived information over stale docs; marks outdated sections for review.
  • Monorepo with many packages: Limits per-package README creation to packages with actual public APIs or user-facing functionality; skips auto-generated or build-output packages.
  • Private or secret-adjacent content: Never documents credentials, tokens, or internal-only endpoints beyond what already exists in code comments.
  • Read-only repository: If git write access is unavailable, outputs documentation as a diff or inline summary rather than committing files.

Step Completion Reports

After completing each major step, output a status report in this format:

◆ [Step Name] ([step N of M] — [context])
··································································
  [Check 1]:          √ pass
  [Check 2]:          √ pass (note if relevant)
  [Check 3]:          × fail — [reason]
  [Check 4]:          √ pass
  [Criteria]:         √ N/M met
  ____________________________
  Result:             PASS | FAIL | PARTIAL

Adapt the check names to match what the step actually validates. Use for pass, × for fail, and to add brief context. The "Criteria" line summarizes how many acceptance criteria were met. The "Result" line gives the overall verdict.

Skill-specific checks per phase

Phase: Branch Setup — checks: Branch creation, Repo sync

Phase: Project Analysis — checks: Project analysis, Gap identification

Phase: Documentation Restructure — checks: Doc restructure, Diagram creation

Phase: Validation — checks: Validation pass, Link verification

Error Handling

No existing documentation found

Solution: Generate documentation from scratch based on code analysis. Start with README.md and add docs/ files based on project complexity.

Conflicting or outdated docs

Solution: Flag conflicts to the user. Prefer code-derived information over stale docs. Mark outdated sections for user review.

Guidelines

  • Keep docs concise and scannable
  • Adapt structure to project type (not all categories apply)
  • Maintain cross-references between related docs
  • Remove redundant or outdated content
  • Preserve any existing docs that are still accurate

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

35.74%
按下载量换算160

Claude

33.01%
按下载量换算148

Cursor

18.27%
按下载量换算82

Gemini CLI

8.74%
按下载量换算39

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。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