writing-artifacts
Compare original and translation side by side
🇺🇸
Original
English🇨🇳
Translation
Chinesewriting-artifacts
writing-artifacts
A production line: raw thoughts in, publishable artifact out. You orchestrate
stations; each station's method lives in its own SKILL.md — read it at that
stage, never work from memory of it.
一条标准化生产线:输入原始想法,输出可发布的成品文档。你需要统筹各个环节;每个环节的方法都存放在对应的SKILL.md中——执行到该环节时请查阅文档,切勿凭记忆操作。
Stage 0 — frame (one question round, then move)
阶段0 — 框架搭建(一轮问题确认后推进)
Pin three things before writing anything: the SUBJECT (one sentence), the
READER (who must understand it), and the OUTPUT FORMAT. Default format by
destination: quick share → · something the user will look at twice or
show someone → page or Artifact · visual metaphor / hero image needed →
add an image stage. If the user already said the format, don't ask.
mdhtmlGenre branch, decided here: in-repo SOFTWARE documentation (README, API
docs, tutorials, how-to guides — anything living under the repo) swaps the
structure station: run (Eight Rules + Diataxis, docs/
location and linking rules) as Stage 2 instead of the writing-beats/shape
pair, keeping Stage 1 mining and Stage 3 discipline unchanged.
Everything else — problem writeups, incidents, decisions, long-form prose —
takes the default line below.
documentation-writingstop-slop开始写作前需锁定三个要点:主题(一句话概括)、读者(目标受众)、输出格式。根据使用场景默认格式:快速分享→ · 用户会反复查看或展示给他人的内容→页面或Artifact · 需要视觉隐喻/主图的内容→添加图片生成环节。如果用户已指定格式,则无需询问。
mdhtml在此环节确定内容类型分支: 仓库内的软件文档(README、API文档、教程、操作指南等仓库下的所有文档)需替换结构环节:将阶段2替换为(八大规则+Diataxis框架、文档位置与链接规则),阶段1的信息挖掘和阶段3的规范保持不变。其他所有内容——问题报告、事件记录、决策说明、长篇散文——均遵循下方默认流程。
documentation-writingstop-slopStage 1 — mine (diverge)
阶段1 — 信息挖掘(发散阶段)
writing-fragmentsadhd调用:将所有与主题相关的信息以原始片段形式输出——证据、时间线、感受、不成熟的想法——无需构建结构。若面对的是完全模糊的想法(而非已知事件),先调用进行平行发散,再整理筛选后的内容为片段。
writing-fragmentsadhdStage 2 — structure (converge)
阶段2 — 结构构建(收敛阶段)
Pick ONE:
- — when the reader must be LED somewhere: problem narratives, incident writeups, decision rationales. Grounds each term before a beat leans on it.
writing-beats - — when the material is already roughly ordered and just needs shaping paragraph by paragraph.
writing-shape
选择其中一种方式:
- — 当需要引导读者理解内容时:问题叙事、事件记录、决策依据。在每个关键节点前先明确相关术语。
writing-beats - — 当素材已有大致顺序,仅需逐段梳理结构时。
writing-shape
Stage 3 — tighten
阶段3 — 精简优化
edit-articlestop-slop对初稿执行,全程以作为评判标准(不仅限于此阶段):无冗余内容、无模糊表述、无AI生成的拖沓语气。
edit-articlestop-slopStage 4 — render (format branch)
阶段4 — 格式渲染(格式分支)
- / plain prose → deliver the tightened text directly; done.
md - page / Artifact / diagram / image → hand the FINISHED TEXT to
htmlas the content contract and let IT compose the pipeline (it owns authority selection, imagegen delegation, and the screenshot-evidence quality loop). Never style inline yourself; the text is frozen content by this point — design changes layout, not wording.using-design-skills
- / 纯文本 → 直接交付精简后的文本;流程结束。
md - 页面 / Artifact / 图表 / 图片 → 将最终文本交付给
html作为内容约定,由其统筹后续流程(包括权限选择、图片生成委托、截图证据质量校验)。切勿自行设置内联样式;此时文本已为固定内容——设计仅调整布局,不修改措辞。using-design-skills
Boundaries
边界规则
- Words first, pixels second: never enter Stage 4 with an untightened draft — redesigning slop produces well-dressed slop.
- One pass back is allowed: if rendering reveals a structural hole (a section that can't be visualized because it was never actually explained), return to Stage 2 for THAT section only, then re-render.
- This skill owns written deliverables about a subject — prose AND in-repo
software docs (the genre branch in Stage 0 picks the structure station).
A plan page from a discussion → directly.
html-plan
- 先文字后视觉:绝不带着未精简的初稿进入阶段4——为冗余内容做设计只会产出包装精良的无效内容。
- 允许返回修改一次:若渲染过程中发现结构漏洞(某部分因未明确说明而无法可视化),仅针对该部分返回阶段2调整,然后重新渲染。
- 本技能负责与主题相关的书面交付物——散文及仓库内的软件文档(阶段0的内容类型分支会选择对应的结构环节)。讨论中的计划页面→直接调用。
html-plan