Token导航 LogoToken导航

Agent Skill · 研究检索

user-guide-writing(用户指南写作)

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。

user-guide-writing 是研究检索类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为操作浏览器;许可证信息还不明确,第三方扫描已通过,按安装量在研究检索类的 73,675 个 Skill 里排第 35,207。

总安装 2,122周安装 85GitHub Stars 11下载 687核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
unknown

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:user-guide-writing(用户指南写作)
来源仓库:https://github.com/akillness/oh-my-skills
仓库路径:skills/user-guide-writing
安装命令:
npx skills add https://github.com/akillness/oh-my-skills --skill user-guide-writing
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/akillness/oh-my-skills --skill user-guide-writing

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题,通常会让助手先检索再总结。

用于编写客户或管理员使用的操作指南,聚焦 workflow 完成路径。适合 getting-started、tutorial 与 FAQ 等场景,提供最小配套文档集。强调任务导向与步骤清晰,避免冗长理论描述。

安装命令:npx skills add https://github.com/akillness/oh-my-skills --skill user-guide-writing 使用时应基于实际产品界面,确保截图与说明一致。

来源仓库内路径:skills/user-guide-writing

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 Codex、Claude、Cursor、Gemini CLI;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 npx skills add https://github.com/akillness/oh-my-skills --skill user-guide-writ…;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它需要操作浏览器,先在测试目录或测试账号里跑,确认没问题再放到正式项目。

常见问题

6 条

user-guide-writing 怎么安装?

在终端执行 npx skills add https://github.com/akillness/oh-my-skills --skill user-guide-writing。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

user-guide-writing 支持哪些 AI 工具?

公开资料标注支持 Codex、Claude、Cursor、Gemini CLI。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

user-guide-writing 需要什么权限?

权限级别是操作浏览器。该 Skill 可能涉及浏览器控制能力,使用时可能读取或操作网页内容,需要在受控环境中确认权限边界。

user-guide-writing 是免费的吗?

公开资料没有标注明确许可证,收费和商用限制都无法确认。正式使用前请到来源仓库核对授权说明。

user-guide-writing 用的人多吗?

公开数据是累计安装 2,122 次,近一周安装 85 次,下载 687 次,GitHub 11 星。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

user-guide-writing 安全吗?

第三方扫描结果为通过,但这只覆盖已知的静态检查项,不能替代自己读一遍源码和安装命令。

值得选的地方和要先想清楚的地方

值得选的地方

  • 同时适配 4 种工具,换客户端不用重新找替代品。
  • 第三方扫描已通过,常见的高风险行为没有命中。

先想清楚的地方

  • 许可证没有明确标注,能不能商用、能不能再分发都不确定。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 Codex,想直接扩展它能力的人。
  • 需要处理研究检索类任务、又不想每次都手写提示词的人。

不太适合

  • 需要明确授权才能用到公司项目里的场景。
  • 需要厂商级技术支持和 SLA 的生产系统。

研究检索类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
find-skills

寻找技能 · Codex、Claude

31,199,81316,486操作浏览器MIT有提醒
self-improving-agent

自我改进 Agent · OpenClaw

16,464,0023,396接触敏感数据MIT-0通过
soultrace

灵魂追踪 · Codex、Claude

7,588,94412需要联网未明确有提醒
self-improving

自我完善 · OpenClaw

6,829,2841,046需要联网MIT-0通过
microsoft-foundry

微软代工厂 · Codex、Claude

6,614,144764调用外部服务未明确有提醒
user-guide-writing

用户指南写作 · 本页

2,12211操作浏览器未明确通过

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

User Guide Writing

Use this skill when the deliverable is customer-facing or admin-facing documentation that helps someone complete a workflow in the product.

user-guide-writing is the documentation-cluster anchor for:

  • getting-started / first-success guides
  • tutorials and guided walkthroughs
  • task-based how-to articles
  • FAQs and support-oriented answer sets
  • release-facing help updates after shipped UI/workflow changes
  • small guide sets that combine one primary page with the minimum supporting companion docs

Read these support docs before choosing the mode or boundary:

When to use this skill

  • A product needs a getting-started guide that gets a new user to the first successful outcome.
  • A feature launch needs a tutorial or how-to article for real product users.
  • A help center needs a task-based article, FAQ, or troubleshooting-friendly walkthrough.
  • Support pain keeps recurring and the answer should become durable customer-facing documentation.
  • A workflow changed and the existing user-facing steps, screenshots, warnings, or prerequisites need to be updated.
  • A request really needs one guide plus a tiny companion packet instead of a huge mixed-purpose document.

When not to use this skill

  • The main job is an internal spec, architecture doc, ADR, runbook, migration plan, or builder-facing implementation guide → use technical-writing.
  • The main job is API reference, SDK docs, webhook docs, developer quickstarts, or developer-portal content → use api-documentation.
  • The main job is release notes, CHANGELOG.md, migration announcements, or shipped-change summaries → use changelog-maintenance.
  • The main job is a deck, launch presentation, workshop slide set, or visual review artifact → use presentation-builder.
  • The main job is product positioning, marketing messaging, lifecycle email copy, or launch copy → use marketing-automation.
  • The main job is building in-app onboarding mechanics or product-tour UX → use the relevant product/UX skill first, then use this skill for the durable written companion docs.

Instructions

Step 1: Classify one primary mode

Normalize the request before drafting.

user_guide_mode:
  primary_mode: getting-started | tutorial | how-to | faq | release-help-update
  audience: end-user | admin | manager | mixed | unknown
  experience_level: first-time | familiar | advanced | mixed | unknown
  docs_surface: help-center | docs-site | in-app-companion | pdf | unknown
  source_of_truth: product-ui | release-notes | support-tickets | sme-notes | mixed | unknown
  maintenance_need: one-off | recurring | stale-doc-rewrite | launch-critical
  output_shape: single-page | guide-plus-faq | refresh-packet | guide-set | unknown

Use one primary mode per run:

  • getting-started → shortest path to first success
  • tutorial → guided learning with context and milestones
  • how-to → one practical task completed quickly
  • faq → concise repeated questions plus route-outs to deeper guides
  • release-help-update → patch user-facing docs after shipped UI or workflow change

Step 2: Confirm audience, task, and route-outs

Answer these before writing:

  1. Who will follow the doc, and what role or permissions do they have?
  2. What exact task or outcome should they achieve?
  3. Which prerequisites, plan limits, or environment assumptions could break the flow?
  4. Which neighboring skills must stay out of scope?

Quick route-out table:

If the request sounds like...Use
"Write the architecture doc / runbook / migration plan"technical-writing
"Write the developer portal / SDK quickstart / webhook docs"api-documentation
"Summarize what shipped this release"changelog-maintenance
"Write onboarding docs / tutorial / FAQ / help article"user-guide-writing
"Make a launch deck / training deck / walkthrough presentation"presentation-builder

Step 3: Gather the minimum user-facing evidence

Do not draft from vibes alone. Pull the smallest credible evidence set first:

  • current UI labels, navigation, and states
  • prerequisites, permissions, and plan / edition assumptions
  • expected result after each important step
  • likely blockers, empty states, and branching conditions
  • screenshots or screenshot placeholders only where they reduce confusion
  • support questions, search terms, or recurring failure signals if available
  • release changes that made the current guide stale

If details are incomplete, label assumptions explicitly instead of pretending the guide is verified.

Step 4: Choose the smallest fitting structure

Use references/mode-structures.md and keep only the sections the chosen mode needs.

Rules:

  • If the request is one user outcome, prefer a single page.
  • If the task needs one main guide plus recurring blockers, use guide plus FAQ.
  • If the real work is patching stale pages after a release, use a refresh packet instead of rewriting a whole manual.
  • If one draft starts doing onboarding, daily usage, troubleshooting, and release notes all at once, split it into a small guide set.

Step 5: Emit the smallest useful artifact packet

Default output shapes:

  • single-page → one getting-started guide, tutorial, how-to, or FAQ page
  • guide-plus-faq → one main guide plus a short FAQ or troubleshooting appendix
  • refresh-packet → changed steps, stale screenshot list, affected companion docs, and assumptions to verify
  • guide-set → a bounded set such as getting-started + FAQ or tutorial + how-to follow-up

Do not ship a broad handbook when the request only needs one page and a short sync list.

Step 6: Apply user-doc writing rules

  • Lead with the task and audience, not implementation details.
  • Name prerequisites early: role, permissions, plan limits, data required, browser/app/version assumptions.
  • Prefer observable actions: what the user clicks, types, sees, and receives.
  • Keep one major outcome per page.
  • Call out branching conditions like role-specific UI or alternate paths.
  • Give the user a success check so they know the task worked.
  • Keep troubleshooting near the task instead of burying it in a generic appendix.
  • Use screenshots intentionally at confusing transitions, not after every click.
  • Link outward deliberately: FAQ → full guide, release-help update → affected articles, getting-started → next task.

Step 7: Run the maintenance and sync check

Use references/workflow-checklist.md and references/maintenance-signals.md before finalizing.

Verify:

  1. The page type matches the user job.
  2. Prerequisites and blockers appear before risky steps.
  3. Screenshot placeholders are justified and trackable.
  4. Related guides, FAQs, release updates, or support macros that must stay in sync are named.
  5. Route-outs to technical-writing, api-documentation, and changelog-maintenance remain explicit when the request drifts.

Step 8: Return a brief or the finished artifact

Preferred brief shape before full drafting:

# User Guide Writing Brief

## Mode
- Primary mode:
- Why it fits:
- Audience:
- Output shape:

## Source material used
- Product truth / UI evidence:
- Support or release signals:
- Assumptions / gaps:

## Planned artifact packet
1. main page
2. companion FAQ / refresh list / sync note

## Writing notes
- Key user outcome:
- Known blockers / branching conditions:
- Route-outs kept out of scope:

If the user already asked for the finished artifact, produce the selected page or packet directly with the matching structure.

Examples

Example 1: First-success onboarding

Input

Write a getting-started guide for new workspace admins inviting their team and creating the first project.

Good output direction

  • mode: getting-started
  • output shape: single-page
  • include prerequisites, step order, success check, and next steps
  • keep internal implementation notes out of scope

Example 2: Release-driven doc refresh

Input

Update our help-center article for exporting reports because the Export button moved into the Reports header and only admins can choose XLSX now.

Good output direction

  • mode: release-help-update
  • output shape: refresh-packet
  • call out changed UI, role differences, screenshot refreshes, and affected companion docs
  • keep release-note summarization out of scope

Example 3: Support-to-FAQ conversion

Input

Turn these repeated billing support replies into a short FAQ for workspace owners.

Good output direction

  • mode: faq
  • output shape: single-page or guide-plus-faq
  • keep answers short and link to the deeper task guide where needed
  • surface role / plan caveats early

Example 4: Boundary with developer docs

Input

Refresh our public webhook quickstart and auth troubleshooting page for external developers.

Good output direction

  • route to api-documentation
  • explain that the main job is published developer-facing API docs, not end-user product guidance

Best practices

  1. Choose the page type before writing the body.
  2. Prefer one user outcome per page and one primary mode per run.
  3. Use the smallest useful packet instead of a giant all-in-one guide.
  4. Put prerequisites, permissions, and blockers before the user gets stuck.
  5. Keep screenshots intentional and easy to refresh.
  6. Use support/search/release signals to decide what to patch next.
  7. Route internal docs, API docs, and release notes out instead of stretching the skill boundary.
  8. Split mixed-purpose drafts into a guide set when the page starts teaching too many jobs.

References

平台分布

Codex

34.24%
按下载量换算235

Claude

31.31%
按下载量换算215

Cursor

17.92%
按下载量换算123

Gemini CLI

9.5%
按下载量换算65

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

操作浏览器

该 Skill 可能涉及浏览器控制能力,使用时可能读取或操作网页内容,需要在受控环境中确认权限边界。

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。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills