Best Writing Skills for Claude Code
Skills for writing, translation, content auditing, and humanizing AI text. Evaluated on six dimensions — full scorecards with written rationale on every detail page.
The best writing skills for Claude Code, ranked by six-dimension evaluation: baoyu-translate (9.4/10), baoyu-xhs-images (9.4/10), baoyu-post-to-wechat (9.3/10), baoyu-wechat-summary (9.3/10), baoyu-youtube-transcript (9.3/10). Every score has published rationale. How we score →
Install the top 3 in one line
curl -fsSL https://skill123.me/install/baoyu-translate.sh | bash
curl -fsSL https://skill123.me/install/baoyu-xhs-images.sh | bash
curl -fsSL https://skill123.me/install/baoyu-post-to-wechat.sh | bash
Verified evaluation, one-line install. Or tell your agent: install the writing skill from skill123.me.
This skill should be used when the user asks to "translate", "翻译", "精翻", "translate article", "tr…
Mature three-mode translation pipeline with EXTEND.md preferences, deterministic markdown chunking, parallel subagent drafting and a critique-revision-polish loop; no When-Not boundary in description.
Generates infographic image card series with 12 visual styles, 8 layouts, and 3 color palettes. B…
Outstanding orchestration skill: 484-line SKILL.md over a deep references tree, mandatory confirmation gate, image-1 anchor chain for consistency, batch policy, backup rule, prompt files as reproducibility record. No scripts, so security surface is text-only and clean.
Posts content to WeChat Official Account (微信公众号) via API or Chrome CDP. Supports article posting …
Mature tool-wrapper skill with pre-flight permission check, three publishing methods, multi-account support, detailed troubleshooting and tested scripts; all network calls go to official WeChat domains with opt-in Telegram QR delivery.
Summarizes WeChat group chat highlights into a structured digest using the local wx-cli binary (h…
Exceptionally thorough digest skill: three-round workflow with mandatory grep-based attribution audit, incremental mode, anti-injection memory rules and completion checklist; only minor sandbox-bypass and size concerns.
Downloads YouTube video transcripts/subtitles and cover images by URL or video ID. Supports multi…
Solid tool wrapper with InnerTube-to-yt-dlp fallback chain, caching, chapter and sub-agent speaker workflows and a thorough error table; cookie use is opt-in, documented and YouTube-scoped.
Fetch any URL and convert to markdown using baoyu-fetch CLI (Chrome CDP with site-specific adapte…
Robust tool-wrapper around a vendored baoyu-fetch CLI with site adapters, blocking first-time preference setup, post-run quality gate and interaction wait modes; cookie sidecar is local and domain-scoped.
When the user wants to optimize content for AI search engines, get cited by LLMs, or appear in AI…
Posts content to Weibo (微博). Supports regular posts with text, images, and videos, and headline a…
Mature browser-automation skill: bilingual triggers, six coordinated scripts with package.json and lockfile, placeholder and image-count verification, careful CDP-only Chrome kill guard, and mandatory manual publish review; anti-bot flag and third-party npm deps cost some security margin.
Designs distributed system architectures, decomposes monoliths into bounded-context services, rec…
High-quality architect skill: per-step validation checkpoints, five scoped references with Load-When table, runnable multi-language examples; only lacks When-Not in description.
Generate a 7-day LinkedIn content plan from a theme, audience, and pillars. Produces per-day post…
Pure planning skill, no scripts or network. Strong 3-pillar framework with quantified guardrails, formula-to-pillar mapping, founder-edition variant, worked 7-day example, and an inbound-readiness checklist that doubles as output verification.
Score comparison
How the top 10 compare across the six evaluation dimensions (0–10). Full written rationale for every score is on each skill's page.
| Skill | Overall | Trigger | Structure | Workflow | Content | Engineering | Security |
|---|---|---|---|---|---|---|---|
| baoyu-translate | 9.4 | 9.0 | 9.3 | 9.2 | 9.0 | 10.0 | 10.0 |
| baoyu-xhs-images | 9.4 | 8.0 | 10.0 | 10.0 | 8.0 | 10.0 | 10.0 |
| baoyu-post-to-wechat | 9.3 | 8.5 | 9.3 | 9.6 | 9.0 | 10.0 | 9.5 |
| baoyu-wechat-summary | 9.3 | 9.0 | 9.3 | 9.6 | 9.0 | 10.0 | 9.0 |
| baoyu-youtube-transcript | 9.3 | 8.5 | 8.7 | 9.6 | 9.0 | 10.0 | 10.0 |
| baoyu-url-to-markdown | 9.2 | 8.0 | 9.3 | 9.6 | 9.0 | 10.0 | 9.5 |
| ai-seo | 9.2 | 9.5 | 9.0 | 8.5 | 9.5 | 9.0 | 10.0 |
| baoyu-post-to-weibo | 9.1 | 8.5 | 9.3 | 9.6 | 10.0 | 9.0 | 8.5 |
| microservices-architect | 9.1 | 7.5 | 10.0 | 8.8 | 9.0 | 10.0 | 10.0 |
| linkedin-content-planner | 9.1 | 9.0 | 9.3 | 8.4 | 10.0 | 8.0 | 10.0 |
FAQ
What is a Claude Code writing skill?
A package that teaches your agent a writing workflow — grammar checking, multilingual translation, content quality audits. It loads when your request matches its trigger.
Which skill removes the AI tone from text?
humanizer — it rewrites AI-sounding text using a catalog of specific tells, with fact-preservation checks so your claims survive the rewrite.
Can these skills translate professionally?
Yes — baoyu-translate scored 9.4 and handles multilingual translation with tone preservation. English-first alternatives are listed too.
Every skill here was scored on the same six-dimension rubric — the full scorecard with written rationale is on each skill's page. Browse the full Writing skills collection (including skills outside the top 10), or see all Writing skills.