Token导航 LogoToken导航

Agent Skill · 运维和基础设施

blog-writing(博客写作)

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。

blog-writing 是运维和基础设施类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为权限未标注;许可证为 MIT,第三方扫描已通过,按安装量在运维和基础设施类的 2,449 个 Skill 里排第 1,476。

总安装 544周安装 22GitHub Stars 9下载 171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
MIT

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:blog-writing(博客写作)
来源仓库:https://github.com/delorenj/skills
仓库路径:skills/blog-writing
安装命令:
npx skills add https://github.com/delorenj/skills --skill blog-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。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/delorenj/skills --skill blog-writing

这个 Skill 是做什么的

运维和基础设施类 Skill 处理部署、监控、日志和云资源,多数需要较高的执行权限。

blog-writing 用于辅助文档与 Markdown 内容的整理改写,适合在 Codex、Claude、Cursor、Gemini CLI 中需要结构化输出技术文案时使用。它提炼章节框架、统一术语并检查链接有效性,保留项目既有事实不虚构结论。使用时避免过度营销表述,保持专业克制语气。

安装前建议确认是否修改原始文件及对外发布文案的合规审查流程。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。

来源仓库内路径:skills/blog-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/delorenj/skills --skill blog-writing;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

    权限没有明确标注,第一次使用时只给一个无关紧要的任务,观察它到底读了哪些文件和网络。

常见问题

6 条

blog-writing 怎么安装?

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

blog-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。

blog-writing 需要什么权限?

权限级别没有明确标注。第一次使用时建议只给无关紧要的任务,观察它实际读写了哪些内容。

blog-writing 是免费的吗?

来源标注的许可证是 MIT。本站不代收任何费用,也不托管安装包;能否商用、要不要署名,以许可证原文为准。

blog-writing 用的人多吗?

公开数据是累计安装 544 次,近一周安装 22 次,下载 171 次,GitHub 9 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。

blog-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。
  • 3 个目录同时收录,说明这个 Skill 流传比较广。
  • 许可证明确为 MIT,商用和分发边界好判断。
  • 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

  • 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。
  • 需要处理运维和基础设施类任务、又不想每次都手写提示词的人。
  • 需要许可证清楚、并且有第三方扫描结果的团队项目。

不太适合

  • 需要厂商级技术支持和 SLA 的生产系统。

运维和基础设施类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
brainstorming

头脑风暴 · Codex、Claude

3,148,776173,131操作浏览器MIT通过
copywriting

文案写作 · Codex、Claude

1,996,72425,583需要联网MIT通过
openclaw-secure-linux-cloud

OpenClaw secure linux cloud 命令行 · Codex、Claude

1,853,08840接触敏感数据未明确未通过
xget

xget 命令行 · Codex、Claude

1,799,28040执行命令未明确未通过
requesting-code-review

代码审查请求 · Codex、Claude

1,590,144173,103权限未标注MIT通过
blog-writing

博客写作 · 本页

5449权限未标注MIT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

Jarad's Blog Voice

Core Voice Patterns

Sentence Structure:

  • Productive rambling: longer, winding sentences that add layers of context
  • Preemptively address skeptics mid-thought with fourth-wall breaks
  • Add specific technical context even when making broad claims
  • Example: "...ok if you're gonna obsess over the accuracy of my estimate, I'll use t-shirt sizes instead of hours/weeks - i'm well aware of the lack of meaningful estimates in both pre- and post-agentic era - but what i'm saying is there is an undeniably amazing almost supernatural improvement..."

Concrete Over Abstract:

  • Use specific actions: "trawled GitHub every morning" NOT "pushed boundaries"
  • Use specific tools/people: "Matt Wolfe, MattVidPro, Claude" NOT "popular YouTubers"
  • Use vivid personal analogies: "boss staring at you while you work" NOT "incubation phase"
  • Show insider knowledge: "GitHub pulse > YouTube hype"

Tone Elements:

  • Direct, almost confrontational: "Use your brain and curate them yourself"
  • Data-focused even in failure: "Data is data"
  • Dark self-interest angle: "stashing dynamite in our doomsday bunkers"
  • Self-aware about exaggeration: acknowledge imprecision before critics do

NEVER Use:

  • "This isn't X, it's Y" profound-sounding structures
  • Cliche transitions: "here's the kicker", "here's where it gets interesting", "and then something happened"
  • Abstract action verbs without details: "experimented relentlessly", "pushed boundaries", "tried to break things"
  • Overly polished blog-speak
  • Clean, explanatory metaphors like "incubation phase"

Header Style:

  • Minimalistic: 2-4 words maximum
  • When read in sequence, headers tell their own story
  • No explanatory subtitles like "And Why That's Beautiful"
  • Example progression: "Flat Charts" → "The Lab" → "Spring 2024" → "The Numbers" → "The Window"

Content Strategy

Opening:

  • Address the skeptic's question directly
  • Don't try to be clever - just state what they said

Body:

  • Share concrete personal experiences with specific details
  • Break fourth wall to preempt criticism
  • Name tools, people, communities to show you're in the trenches
  • Let sentences run long when adding nuance
  • Bury practical tips in the rambling

Addressing Critics:

  • Do it mid-paragraph, not in separate defensive sections
  • Use self-deprecating acknowledgment before they can attack
  • Then pivot to the actual point

Closing:

  • Direct callback to opening question
  • Honest about self-interest and the dark future
  • End with something that feels human and imperfect

Examples

Bad: "The shovelware isn't missing. It's incubating." Good: "I would say this is more accurately 'an incubation phase'. Side effects include tons of garbage code, extra long cycles devoted to theory - stuff that's usually in textbooks - except we didn't write them yet."

Bad: "I was hitting these incredible 'a-ha' moments weekly." Good: "I was on a roll, building stuff day and night - literally, as in I didn't sleep much anymore."

Bad: "Experimented relentlessly. Pushed boundaries. Tried to break things." Good: "While everyone was busy making fun of claude's shitty sense of humor, I looked at every single failure as progress. Data is data. When everyone was eating up the tools they saw Matt Wolfe or MattVidPro talk about, I just cast my line into the github sea every morning and got a pulse on the community - guess what - there are SO many more quiet non-youtube developers out there making tools at 10x speed than can be reported. Use your brain and curate them yourself."

Bad: "So yeah, Lars - the explosion is coming. We're just busy learning how to detonate it properly." Good: "So yeah, Lars - the explosion is coming. We're just all busy quietly mining dynamite while making sure to stash some for ourselves in our doomsday bunkers."

平台分布

Claude Code

31.95%
按下载量换算55

OpenCode

25.52%
按下载量换算44

windsurf

16.07%
按下载量换算27

trae

12.96%
按下载量换算22

Cursor

7.33%
按下载量换算13

Codex

3.68%
按下载量换算6

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

权限未标注

当前来源未能明确判断权限范围,默认进入异常复核队列。

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。

继续浏览同类 Skills