Write technical blogs and long-form with AI
Skills used in this guide (3)
写文章、指南、博客、教程与 Newsletter，并依据样稿或品牌规范还原稳定的行文风格。适合需要超过一段篇幅、且要求语气一致与结构可信的成稿。
npx skills add affaan-m/ECC -g创建新技能、编辑既有技能，或部署前验证技能可用时使用。
npx skills add obra/superpowers -g为 X、LinkedIn、TikTok、YouTube、Newsletter 等平台搭建原生内容体系，产出帖子、长推、脚本、内容日历，并把一份素材干净地改编到多平台。
npx skills add affaan-m/ECC -g
The worst tech blog is "copy the docs" or "fluff padding". The point of AI: give it your real experience/pitfalls (the unique value), let it handle structure, examples, polish — not invent stuff you never did.
Who this is for / not for
For:
- Have hands-on experience to distill into articles
- Stuck at "can't start" or "messy structure"
- Want steady output, consistent voice
Not for:
- Never did it but want AI to invent substance (fiction)
- Publish without edits (you're the author)
- Clickbait only, no value
Before you start
- Material: projects, pitfalls, code you did.
- Reader: beginner / peer, depth.
- Platform: blog / newsletter / Zhihu, length & tone.
Steps
1. Install the skill
article-writing is a long-form skill. Copy its folder into your client's skills directory (same as any Agent Skill):
| Client | Directory |
|---|---|
| Claude Code | ~/.claude/skills/ |
| Codex | ~/.codex/skills/ |
| Cursor | ~/.cursor/skills/ |
| Trae | ~/.trae/skills/ |
| CodeBuddy | ~/.codebuddy/skills/ |
| Tongyi Lingma | ~/.lingma/skills/ |
2. Give material, fix structure
Write "survive flash sale with Redis" for junior backends.
Material: we used Redis+Lua for stock deduct on 11.11, hit oversell pit (code).
Outline first (pain→approach→pit→review), one runnable snippet per section,
don't copy docs, write what we really did.3. Clarity with writing-skills
After draft, run writing-skills to cut redundancy, add transitions, sharpen conclusions.
4. You finalize
Add real data/screenshots, delete fabricated parts, set title and hook.
Pitfalls
| Pit | Symptom | Fix |
|---|---|---|
| Copy docs | No value | Feed real pitfalls |
| Fluff | No takeaway | One runnable example/section |
| Messy | Unreadable | Outline first |
| Fabricate | Wrote undone | Mark [verify] outside material |
| No hook | Swipe early | Open with counterintuitive/pain |
FAQ
No material?
Hard. Best with your project/pit; pure AI fabricates.
Structure?
Pain→approach→pit→review is a solid skeleton.
Consistent voice?
Yes — give one sample for AI to match tone/rhythm.
Code needed?
Yes, runnable snippets beat description.
Title?
From reader pain; specific beats abstract for CTR.
Related guides
Continue across these scenes:
- Produce a Xiaohongshu post that actually gets taps
- Write cold emails and outreach with AI
- Translate and localize docs with AI
相关源码
- affaan-m/ECC · article-writing source
- obra/superpowers · writing-skills source
- affaan-m/ECC · content-engine source
- Agent Skill open format spec
一句话结论
结论：技术博客喂真实经验/坑让 AI 管结构润色，别让它编没做过的事，你才是作者。
References
Related Scenes
Social content AI pipeline (translated)
The biggest lever in social content is the topic — communities widely agree "topic decides 60%+ of a post's success". Have AI mine last-7-day hits for direction first, then run topic → script → image/video → multi-platform → review. Far better than staring at a blank page.
Produce a Xiaohongshu post that actually gets taps
Xiaohongshu posts fail in a specific way: the content is fine and nobody taps. The first decision point is the cover and title, not the body. So reverse the order. Lock the cover line first, then write the content.