跳到正文
  1. 技能速查
  2. 动手和清理
  3. /pstack:unslop

/pstack:unslop

自动调用

一句话去掉文字里的「AI 味」:套话、空话、排比、破折号。

什么时候用
写 README、PR 说明、提交信息,或嫌 AI 的回复太啰嗦。
这样说
/pstack:unslop 改一下 README 的新内容,不要破折号

用它和不用它

同一件事,直接跟 AI 说,和点名用 /pstack:unslop,AI 的做法不一样:

直接说用 /pstack:unslop
写「好读的 SQL」这种只讲感觉的句子换成机制或数字,比如「改列名会让构建失败」
写「专家认为」,不说是哪位专家写出具体来源,写不出就删掉
开头夸「好问题」,结尾问「还需要别的吗」删掉客套和奉承,直接回答

文档写了什么

原文分 8 节,约 850 个英文单词,中文译文约 1050 字,读完约 4 分钟。点一节直接跳到全文。

开头先扫一遍下面列出的 AI 腔,再改写,意思不变,语气照原文想要的来。

  1. 1要找出并改掉的套路规则带固定编号,别的技能按编号引用,删掉的规则留着空号不补。
  2. 2内容删掉「突出了」「体现了」这类空 -ing 短语,「专家认为」要写出来源或删掉。
  3. 3用词AI 词换成大白话,「serves as」改回「is」,不硬凑三项,同一个东西只用一个名字。
  4. 4风格完全不用破折号,少用冒号和粗体,标题用句子式大小写,删装饰 emoji,用直引号。
  5. 5沟通类文字删掉「希望对你有帮助」这类客套和「好问题」这类奉承,直接回答。
  6. 6废话「in order to」改成「to」,层层含糊压成一个「may」,空泛结尾换成具体计划或事实。
  7. 7行话把 substrate、north star、flywheel 这类听着专业的比喻名词换成具体的词。
  8. 8说人话让每句写出机制或数字,拆长句,改成主动语态,删副词和比喻,箭头缩写写回完整句子。

看写法亮点读全文

写法亮点

挑了几句原文:每句让 AI 做什么、为什么这样写管用、你写提示词时怎么借来用。

  1. 给每条规则编固定的号

    Rule numbers are stable ids that other skills cite.

    规则编号是固定的 id,别的技能会引用它们。

    让 AI
    别的技能可以直接说「按第 13 条改」,增删规则后也不会对错号。
    为什么管用
    编号固定下来,引用就短又准,不会因为清单改动而错位。
    借来用
    把团队的写作规矩编上号,评审时直接说:「这段违反了第 4 条。」
  2. 写出一眼能扫到的特征

    Catch "is/are/was/were + past participle" and name the actor

    找出「is/are/was/were + 过去分词」,点出是谁做的

    让 AI
    它能一句句挑出被动句,并补上动作到底是谁做的。
    为什么管用
    给出看得见的字面特征,AI 就知道该搜什么,不用凭感觉判断。
    借来用
    想去掉某种毛病,写出它的样子:「找出所有『进行 + 动词』,比如『进行优化』,改成『优化』。」
  3. 用「搬到别处」来测空话

    if the sentence could appear unchanged in another project's docs, it says nothing about this one.

    如果这句话原封不动地放进另一个项目的文档里也说得通,它对这个项目就什么都没说。

    让 AI
    放到哪个项目都成立的套话会被删掉,留下的句子都在说这个项目自己。
    为什么管用
    「空话」不好判断,「搬过去还成立吗」是一个马上能做的小实验。
    借来用
    让 AI 改简介时加一句:「换成竞品的名字也成立的句子,删掉。」
  4. 讲出毛病从哪来

    An adverb propping up a weak verb means the verb is wrong.

    一个靠副词撑着的弱动词,说明动词选错了。

    让 AI
    它不只删掉「快速地」「显著地」,还会换一个本身就准确的动词或一个数字。
    为什么管用
    说出毛病的来源,AI 就能去改根子,也能处理清单里没写到的例子。
    借来用
    提要求时带上原因:「标题太长,通常是一个标题想说两件事,把它拆开。」

全文

AI 读到的就是这份 SKILL.md。可以看中文、看原文,或者一段一段对照着看。

descriptionCut AI tells from any writing.

description从任何文字里去掉一看就是 AI 写的痕迹。

Unslop

Unslop(去掉 AI 味)

Edit text to remove AI patterns. Scan for the patterns below, then rewrite. Preserve meaning and match the intended tone.

编辑文字,去掉 AI 的写作套路。先按下面的套路扫一遍,再改写。保留原意,贴合原本想要的语气。

Patterns to detect and fix

要找出并改掉的套路

Rule numbers are stable ids that other skills cite. A removed rule leaves a gap.

规则编号是固定的 id,别的技能会引用它们。删掉的规则会留下一个空号。

Content
内容
  1. Superficial -ing phrases. "highlighting...", "ensuring...", "reflecting...", "showcasing...", "fostering...". Delete or expand with real sources.
  2. Vague attributions. "Experts believe", "Industry reports suggest", "Some critics argue". Name the source or delete.
  1. 肤浅的 -ing 短语。 「highlighting...」「ensuring...」「reflecting...」「showcasing...」「fostering...」。删掉,或者用真实的来源展开。
  2. 含糊的出处。 「Experts believe」「Industry reports suggest」「Some critics argue」。写出来源,或者删掉。
Language
用词
  1. AI vocabulary. Additionally, crucial, delve, enduring, enhance, fostering, garner, interplay, intricate, landscape (abstract), pivotal, showcase, tapestry (abstract), testament, underscore, vibrant. Replace with plain words.
  2. Fancy ways to say "is". "serves as", "stands as", "boasts", "features". Say "is" or "has".
  3. "Not just X, but Y." State the point directly instead.
  4. Rule of three. Don't force ideas into groups of three. Use the natural number.
  5. Synonym cycling. Protagonist, main character, central figure, hero all in one paragraph. Pick one, repeat it.
  6. False ranges. "from X to Y" where X and Y aren't on a meaningful scale. List topics directly.
  1. AI 词汇。 Additionally, crucial, delve, enduring, enhance, fostering, garner, interplay, intricate, landscape(抽象义), pivotal, showcase, tapestry(抽象义), testament, underscore, vibrant。换成平常的词。
  2. 把「is」说得花里胡哨。 「serves as」「stands as」「boasts」「features」。直接说「is」或「has」。
  3. 「Not just X, but Y.」 直接说出要点。
  4. 三个一组。 不要硬把想法凑成三个一组。有几个就写几个。
  5. 同义词轮换。 一段里 Protagonist、main character、central figure、hero 全都出现。选一个,一直用它。
  6. 假的范围。 「from X to Y」,而 X 和 Y 并不在一个有意义的刻度上。直接把话题列出来。
Style
风格
  1. Em dash overuse. Avoid em dashes entirely. Use periods or commas only (no parentheses, no en dashes, no hyphen-as-dash substitutes). If a thought needs separation, end the sentence or use a comma.
  2. Colon overuse. Colons are fine before a list or example, not as mid-sentence connectors. "If you're coming from traditional automation: instead of registering event handlers, you describe conditions" adds nothing with the colon. Rewrite to let the point stand on its own without comparison framing. "Describing when the scheduler should fire works best as plain English." Same meaning, no crutch punctuation.
  3. Boldface overuse. Don't bold every proper noun or acronym.
  4. Inline-header lists. The tell is a bold label and colon that restates the line: "Performance: Performance improved...". Convert those to prose. A bold lead-in that ends in a period, names the item, and is followed by genuinely new detail ("Schema in TypeScript. Tables live in one file.") is fine, not a tell.
  5. Title case headings. Use sentence case.
  6. Decorative emojis. Remove from headings and bullets.
  7. Curly quotes. Use straight quotes.
  1. 破折号用太多。 完全不用破折号(em dash)。只用句号或逗号(不用括号,不用 en dash,也不拿连字符冒充破折号)。一个想法需要隔开时,就结束这句话,或者用逗号。
  2. 冒号用太多。 冒号放在列表或例子前面没问题,但不要拿它当句子中间的连接词。「If you're coming from traditional automation: instead of registering event handlers, you describe conditions」这句的冒号什么都没加。改写成让要点自己站得住,不要那种拿来对比的铺垫。「Describing when the scheduler should fire works best as plain English.」意思一样,没有靠标点撑着。
  3. 加粗用太多。 不要把每个专有名词或缩写都加粗。
  4. 带行内标题的列表。 破绽是一个加粗标签加冒号,又把这一行的内容重复一遍:「Performance: Performance improved...」。把这种改成正常的句子。一个以句号结尾、点出这一项的名字、后面跟着真正新内容的加粗开头(「Schema in TypeScript. Tables live in one file.」)没问题,不算破绽。
  5. 标题每个词首字母大写。 改用句子式大小写。
  6. 装饰用的 emoji。 从标题和列表项里删掉。
  7. 弯引号。 用直引号。
Communication artifacts
沟通类文字
  1. Chatbot phrases. "I hope this helps!", "Let me know if...", "Of course!", "Certainly!", "Found the smoking gun!" Remove.
  2. Sycophantic tone. "Great question! You're absolutely right!" Respond directly.
  1. 聊天机器人用语。 「I hope this helps!」「Let me know if...」「Of course!」「Certainly!」「Found the smoking gun!」删掉。
  2. 讨好的语气。 「Great question! You're absolutely right!」直接回应。
Filler
废话
  1. Filler phrases. "In order to" becomes "To". "Due to the fact that" becomes "Because". Delete "It is important to note that".
  2. Excessive hedging. "could potentially possibly be argued that it might" becomes "may".
  3. Generic conclusions. "The future looks bright." State specific plans or facts.
  1. 凑字的短语。 「In order to」改成「To」。「Due to the fact that」改成「Because」。删掉「It is important to note that」。
  2. 过度含糊。 「could potentially possibly be argued that it might」改成「may」。
  3. 空泛的结尾。 「The future looks bright.」写出具体的计划或事实。
Jargon
行话
  1. Abstract metaphor nouns. Substrate (base), wedge, wedge in (add), vector (way, method), locus, vantage, nexus, primitive as a noun, harness as a metaphor, surface as in "API surface", bedrock, scaffolding as a metaphor, modality, paradigm, gold-plating (more than the job needs), ratchet as a metaphor (the mechanism's real name, or a limit that only tightens), evacuate for moving code (move out), endgame (the last phase), north star, flywheel. These read as technical but usually have a plainer concrete word. Pick the concrete word.
  1. 抽象的比喻名词。 Substrate(base)、wedge、wedge in(add)、vector(way、method)、locus、vantage、nexus、当名词用的 primitive、当比喻用的 harness、「API surface」里那种 surface、bedrock、当比喻用的 scaffolding、modality、paradigm、gold-plating(more than the job needs)、当比喻用的 ratchet(换成这个机制的真名,或者「一个只会收紧的限制」)、说挪代码时用的 evacuate(move out)、endgame(the last phase)、north star、flywheel。这些读起来很技术,但通常都有更平实、更具体的词。选那个具体的词。
Plain speech
说人话
  1. Say what it does, not how it feels. "the database stays close at hand", "SQL you can read", "types that follow your schema" name a feeling. The fix names the mechanism or a number: ".toSQL() returns the exact string sent to the database", "a column rename fails the build". Ask what the sentence tells the reader to do or know, then write that. If you can't restate it as a concrete instruction, fact, or number, cut it. One more check: if the sentence could appear unchanged in another project's docs, it says nothing about this one. Cut it.
  2. Shorten or split dense sentences. If the reader has to backtrack to parse a sentence, break it in two or drop clauses. One idea per sentence.
  3. Active voice. Prefer it. Catch "is/are/was/were + past participle" and name the actor: "queries are validated" becomes "the compiler validates queries", "the file is parsed by the loader" becomes "the loader parses the file". Passive is fine only when the actor is unknown or genuinely doesn't matter.
  4. Cut adverbs, or use a stronger verb. "runs quickly" becomes "is fast" or the number, and "significantly improves" becomes the measured delta. An adverb propping up a weak verb means the verb is wrong.
  5. Prefer the plain word. "utilize" and "leverage" become "use", "facilitate" becomes "help", "numerous" becomes "many", "in the event that" becomes "if". The fancier synonym is rarely clearer.
  6. Mannered prose. Metaphor or flourish where a literal phrase exists: aphorisms ("wire it or delete it"), rhetorical fragments for effect, personified code ("the plan holds it"), figurative verbs ("rides along", "stands on"), stock framing phrases. "A dial worth turning" becomes "a parameter worth varying". Say what you mean. Rule 26 covers the metaphor nouns.
  7. Over-compression. Dropped articles, verbless fragments, symbol-speak, and abbreviations that make the reader decode instead of read. "Parser rejects bad date → exit 2, no write" becomes "The parser rejects a bad date, exits with code 2, and writes nothing." Write whole sentences with their articles and verbs, and spell out arrows and abbreviations.
  1. 说它做什么,不说它给人什么感觉。 「the database stays close at hand」「SQL you can read」「types that follow your schema」说的是一种感觉。改法是写出机制或一个数字:「.toSQL() returns the exact string sent to the database」「a column rename fails the build」。问问这句话要读者去做什么、知道什么,然后就写那个。如果你没法把它改写成一条具体的指令、事实或数字,就删掉。再查一项:如果这句话原封不动地放进另一个项目的文档里也说得通,它对这个项目就什么都没说。删掉。
  2. 缩短或拆开太密的句子。 如果读者得回头重读才能看懂一句话,就拆成两句,或者删掉几个从句。一句话一个意思。
  3. 用主动语态。 优先用它。找出「is/are/was/were + 过去分词」,点出是谁做的:「queries are validated」改成「the compiler validates queries」,「the file is parsed by the loader」改成「the loader parses the file」。只有不知道是谁做的、或者是谁做的确实无关紧要时,才可以用被动语态。
  4. 删掉副词,或者换一个更有力的动词。 「runs quickly」改成「is fast」或者直接给数字,「significantly improves」改成量出来的差值。一个靠副词撑着的弱动词,说明动词选错了。
  5. 用平常的词。 「utilize」和「leverage」改成「use」,「facilitate」改成「help」,「numerous」改成「many」,「in the event that」改成「if」。更花哨的同义词,很少更清楚。
  6. 做作的文字。 明明有字面的说法,却用比喻或修辞:格言式的话(「wire it or delete it」)、为了效果而写的残句、把代码拟人化(「the plan holds it」)、比喻性的动词(「rides along」「stands on」)、套路化的开场框架。「A dial worth turning」改成「a parameter worth varying」。有话直说。比喻名词归第 26 条管。
  7. 压缩过度。 省掉冠词、没有动词的残句、拿符号代替话、缩写,让读者得去解码,而不是阅读。「Parser rejects bad date → exit 2, no write」改成「The parser rejects a bad date, exits with code 2, and writes nothing.」写完整的句子,带上冠词和动词,把箭头和缩写都写成字。

原文作者 Lauren Tan(poteto),MIT 许可。译文由本站翻译,核对到 pstack 镜像 97de40d。 在 GitHub 上看

本站是 pstack 的非官方入门教程,不是官方文档;pstack 由 poteto 开发,命令和行为以官方仓库为准。教程原始源码见 GitHub。

由 airali部署,基于小小东的 MIT 授权教程改编。