Updated 3mo ago
by tzengyuxio
skills
A collection of custom skills for Claude Code and compatible AI coding agents
47 GitHub stars (47)
Steady at 47 stars — no new stars in the 16 days since Sep 30.
- 30 skills (30 skills found in its SKILL.md files)
- 3 forks (3 forks)
- 3mo ago last update (Last updated 3mo ago)
Install
Skills CLI
npx skills add tzengyuxio/skills
Asks which skills and agents to install. Just one: npx skills add tzengyuxio/skills --skill bagu-essay — add -g for every project.
Claude Code
git clone https://github.com/tzengyuxio/skillscp -r "skills/bagu-essay" ~/.claude/skills/
Copies the bagu-essay skill into ~/.claude/skills/ for every project (use .claude/skills/ for one project). Repeat for the other skills listed below.
Plugin
/plugin marketplace add tzengyuxio/skills
Run inside Claude Code if the repository ships a plugin marketplace, then pick plugins from /plugin.
Skills can run code on your machine — read the SKILL.md first.
What it does
In the author's own words, from each skill's SKILL.md — showing the first 30 of 47 files.
-
bagu-essay
Transform any article into Ming-Qing eight-legged essay (八股文 / 經義文 / 明清制藝)—— a strict eight-section structure inherited from the imperial examination system. Eight sections required: 破題、承題、起講、入手、起股、 中股、後股、束股。 The four 股(起股/中股/後股/束股) MUST be paired parallel couplets (兩兩對仗、字數相當、虛實相對、平仄錯綜)。 Voice is 代聖人立言: speak as Confucius / Mencius / Zhu Xi would, never first-person 「我」。 Allusions to 四書五經 expecte…
bagu-essay/SKILL.md on GitHub (opens in a new tab) -
chinese-sonnet-style
Transform any modern article into a 中文十四行詩 (Chinese sonnet) — a strict 14-line metrical poem in modern Chinese, modelled after 馮至《十四行集》(1942) with formal scaffolding from the two canonical Western forms: Petrarchan (義式 8+6, 押 ABBAABBA CDECDE 或 CDCDCD, volta 在第 9 行) and Shakespearean (英式 4+4+4+2, 押 ABAB CDCD EFEF GG, volta 在第 9 或第 13 行, 結句 couplet 收). Output strictly obeys 14 lines; each line 10–1…
chinese-sonnet-style/SKILL.md on GitHub (opens in a new tab) -
chinglish
Transform Chinese text into Chinglish (中式英語) — English that is heavily influenced by Chinese grammar, word order, and thought patterns, producing the characteristic "中式英文" style found on signs, menus, instructions, and everyday speech. Can also take English text and re-render it through a Chinese-thinking lens to produce Chinglish. Applies a 25-item checklist covering article errors, copula dropp…
chinglish/SKILL.md on GitHub (opens in a new tab) -
concrete-42
Transform any article into 「義大利麵就應該拌 42 號混凝土」體(concrete-42 pseudo-professional jargon-soup style)— Traditional Chinese, paragraph-style 論述外殼包裝跨領域名詞拼貼,每段強制至少 3 個跨領域具體名詞同時出現 (數學定理/工業規格/食物/歷史人物/物理化學概念/經濟術語),用偽因果 連接詞(因此/顯然/必然導致/由此可見/綜上所述)串成一本正經但前後不通 的論述,結論與開頭主題完全脫節。源自抖音「延邊刺客」卧龍鳳雛網路梗 (2024)。Primarily user-invoked via "/concrete-42"; also triggers on 「義大利麵」「42 號混凝土」「答非所問」「偽專業」「術語堆砌」「改寫成 42 號混凝土體」「偽專業話術…
concrete-42/SKILL.md on GitHub (opens in a new tab) -
dewesternise
Review and rewrite Traditional Chinese text to remove Europeanized (歐化) writing patterns that arise from over-literal translation of Western sentence structures. Applies a 25-item checklist covering passive voice, abstract subjects, clause ordering, 的/們/性/地 overuse, calque prepositions, weak verbs, pronoun/tense overuse, long pre-modifiers, and more. Works on markdown files or inline text. Trigge…
dewesternise/SKILL.md on GitHub (opens in a new tab) -
dongbei-mandarin
Transform text into Dongbei Mandarin (東北話) — the Northeast Chinese variant of Mandarin spoken across Heilongjiang, Jilin, and Liaoning, shaped by long contact with Manchu, Mongolian, Russian, and a strong oral-performance tradition (二人轉、 小品、《鄉村愛情》). Features the all-purpose verb 整, the intensifier 賊, signature lexicon (嘎哈/咋整/老鐵/妥妥的/杠杠的/必須的/可勁兒/瞅), 老 X / 小 X prefixes, moderate 兒化, sentence-final p…
dongbei-mandarin/SKILL.md on GitHub (opens in a new tab) -
eileen-chang-style
Transform any article into Eileen Chang-style (張愛玲體) Chinese prose — Traditional Chinese, 400–1000 characters, 蒼涼基調 (the world quietly sinks, never erupts), 舊上海/戰時香港 mise-en-scène, female gaze on fabric and family power, 反高潮收尾 (cool off at emotional peak), 細節暴力 (one tiny detail carries an entire feeling), 古典-通俗夾雜 with low- saturation colours (暗黃、月白、灰青)。 Use this skill whenever the user asks for 張…
eileen-chang-style/SKILL.md on GitHub (opens in a new tab) -
feihua-literature
Rewrite any article into Feihua Literature (廢話文學) — the 2021–2022 Chinese internet meme genre that produces grammatically perfect, tonally earnest sentences carrying zero new information. Hallmarks: synonym substitution (「聽君一席話,如聽一席話」), tautology / contradiction (「這個西瓜很甜,因為它不苦」), and Q-keyword echo (「Q: 你為什麼遲到? A: 因為我沒有準時。」). The text must read as if it were earnestly delivering truth while in fa…
feihua-literature/SKILL.md on GitHub (opens in a new tab) -
gulong-style
Transform any article into Mid-Era Gulong (中後期古龍《多情劍客無情劍》 1969 之後) wuxia narrative — Traditional Chinese, 200–600 characters with extreme short-sentence rhythm (avg ≤12 chars), embedded aphorisms, mandatory contrast tension (殺手 vs 殺手 / 生 vs 死 / 動 vs 靜 / 年輕 vs 年老), no battle process. Designed as the duality counterpart of the future `jinyong-style` (章回招式派) — for that style use jinyong-style when i…
gulong-style/SKILL.md on GitHub (opens in a new tab) -
haiku-style
Compress and distill any modern article into a Japanese 俳句 (haiku) in Traditional Chinese translation. Hybrid output modes: Light/Medium produce one (or two) standalone haiku from the most condensed moment of the source; Heavy produces a 3–7-haiku 連作 sequence in the manner of 松尾芭蕉《奧之 細道》 (prose 俳文 interleaving multiple haiku). Strictly observes the three pillars: 5-7-5 mora structure (in Chinese…
haiku-style/SKILL.md on GitHub (opens in a new tab) -
hanjify
Transform English text into Hanjified English — replacing words with CJK characters (Hanzi/Kanji/Hanja) while preserving English grammar structure. Uses a hybrid morphology where inflectional suffixes stay as Roman letters and derivational suffixes/prefixes become Hanzi. Produces three output versions: pure Hanjified, HTML Ruby annotation, and original text. Works on markdown files or inline text…
hanjify/SKILL.md on GitHub (opens in a new tab) -
hemingway-style
Transform any article into a Chinese rendering of Hemingway-flavored prose (中譯海明威腔) — Traditional Chinese, 短句為主、動詞主導、拒絕修飾, 外部動作敘述,情緒壓在地表下,以冰山理論(七分水下、一分水上)承載強烈 情感。風格參照《老人與海》《白象似的群山》《雨中的貓》《一個乾淨明亮 的地方》《大二心河》《戰地春夢》《我們的時代》。歐化句法(「於是」「然 後」「他走進去。他坐下。他點了酒。」這類短句並列)在這個 skill 是 feature 不是 bug — 翻譯腔即是中譯海明威腔的指紋之一。Primarily user-invoked via "/hemingway-style"; also triggers on "改寫成海明威", "海明威體", "海明威腔", "冰山理論寫作"…
hemingway-style/SKILL.md on GitHub (opens in a new tab) -
hongkong-chinese
Transform text into Hong Kong Chinese (港式中文) — a written register shaped by Cantonese substrate, English contact, and Hong Kong-specific social vocabulary. Features Hong Kong-specific terms (巴士/的士/士多/雪櫃), transliterations (荷里活/三文治/梳化), Cantonese lexicon (食飯/飲茶/睇/搞掂/啲/嘢), frequent English code-mixing (開 meeting / send email), AB/BA word-order flips (質素/私隱/擠擁), and — at Heavy intensity — optional f…
hongkong-chinese/SKILL.md on GitHub (opens in a new tab) -
humanize
Remove AI-generated writing tics (grandiose wording, tricolon abuse, vague attribution, hollow parallel structures) and rewrite flagged passages in a more natural voice. Works on markdown files or inline text; checklist is tuned for Traditional Chinese but principles apply to any language. **Scope boundary**: humanize targets *AI-flavored tone*. If the issue is Europeanized sentence structure (翻譯…
humanize/SKILL.md on GitHub (opens in a new tab) -
jingjing-style
Transform text into 晶晶體 — a Taiwan colloquial register (起源於 2016 年 媒體名媛口語現象) characterised by a Chinese sentence frame sprinkled with unnecessary English single words or short phrases (amazing / busy / fashion / life style / nice / really 等),其炫耀感正來自「明明可以用中文卻用英文」。 Three intensity levels (Light 每 4–5 句 1 處 / Medium 每 2–3 句 1 處 / Heavy 每句 1–2 處)。Useful for stylistic experiments, satire of code-switc…
jingjing-style/SKILL.md on GitHub (opens in a new tab) -
jinyong-style
Transform any article into Mid-Era Jinyong (中期黃金期金庸 1957–1969, 《射鵰》《神鵰》《倚天》《天龍》《笑傲》流派) wuxia narrative — Traditional Chinese, 400–1200 characters with 章回式長句 (avg ≥18 chars), mandatory martial-art moves with names, embedded classical poetry, and historical anchoring (specific dynasty + place). Designed as the duality counterpart of `gulong-style` (短句留白派) — for that style use gulong-style. Excludes…
jinyong-style/SKILL.md on GitHub (opens in a new tab) -
koanify
Transform any article into a narrative-style Zen koan (敘事故事型禪宗公案) in the 白隱「是這樣嗎?」 tradition — Traditional Chinese, white-vernacular with light classical particles, 100–400 characters, no commentary appended. This is the narrative-story type, NOT the short Q&A koan (機鋒問答型 趙州狗子). Primarily user-invoked via "/koanify"; also triggers on explicit phrases like "改寫成公案", "做一則公案", "rewrite as a Zen koan".
koanify/SKILL.md on GitHub (opens in a new tab) -
laogao-style
Use this skill to rewrite articles, stories, or any text into the Chinese YouTube mystery-narration genre — the "未解之謎 / 都市傳說 / 失落古文明 / 神秘解說" voiceover style popularized by 老高與小茉 and similar channels. Trigger broadly whenever the user wants text reshaped to sound like a Chinese mystery/conspiracy/lost-civilization YouTube video: 老高體、老高與小茉風格、laogao style、Mr & Mrs Gao style、雙人解說 (narrator + reactor)…
laogao-style/SKILL.md on GitHub (opens in a new tab) -
liaozhai-tale
Transform any article into a 聊齋體 classical Chinese short tale (蒲松齡 《聊齋誌異》流派) — 200–800 character pure 文言 narrative with mandatory supernatural transposition (one element becomes 狐/鬼/妖/精/神/異獸/物精), 史傳式 opening, and a conditional 「異史氏曰」 commentary if the source carries satire or moral tone. Primarily user-invoked via "/liaozhai-tale"; also triggers on "改寫成聊齋", "寫成聊齋體", "蒲松齡體", "志怪短篇", "rewrite as a…
liaozhai-tale/SKILL.md on GitHub (opens in a new tab) -
lu-xun-style
Transform any article into 魯迅體(早期白話雜文)—— a 1920–30s vernacular voice fusing 文白夾雜 (early baihua spine with classical function-words 然而/便/罷了/也罷/未免/實乃 sprinkled in), 反諷距離 (the narrator stands at the scene yet outside it, neither endorsing nor mocking),標誌虛詞密集(然而、便、罷了、未免、實在、想來、 只是、倒是、至於、究竟),「我」字第一人稱抒情主體偶現,短段落為主 (50–150 字),破折號連接思考停頓,結尾不下定論(留問號/留靜默/留動作)。 Output is Traditional Chinese essay-style prose m…
lu-xun-style/SKILL.md on GitHub (opens in a new tab) -
mainland-mandarin
Transform text into Mainland China Mandarin (中國大陸普通話) — the everyday spoken and media register of the People's Republic of China. Features PRC-specific everyday vocabulary (視頻/質量/出租車/公交/軟件/互聯網/視 頻/屏幕/鼠標), administrative-life terms (戶口/社保/落戶/北漂/公積金), social-platform terms (微信/朋友圈/微博/熱搜/B站/彈幕/抖音/小紅書/拼多 多/美團), and 2020-2025 internet slang (內卷/躺平/yyds/破防/絕絕子/家人 們誰懂啊/友友們/666/拿捏/鴿了/翻車/emo 了/蚌埠住了/cityci…
mainland-mandarin/SKILL.md on GitHub (opens in a new tab) -
postmodern-lens
Rewrite an article through a postmodern lens — events stay intact, but the narrative's interior (its master story, its narrator's authority, its assumed binaries, its claim to be "the" version) gets re-narrated so the article looks at itself with suspicion: which big story is smuggling itself in, who gets to say it, what other version was edited out, and what stays unsayable. This is a **lens** s…
postmodern-lens/SKILL.md on GitHub (opens in a new tab) -
prc-bureaucratese
Transform text into PRC party-state bureaucratese (中共公文體 / 黨政公文體) — the institutional writing style of Mainland China's Party and government documents, news bulletins, and official speeches. Features high-frequency action verbs (落實/加強/推進/深化), adverb-verb stacks (紮實推進/深入貫徹/全面落實), four-character parallel phrases (不忘初心、牢記使命), numerical political slogans (三個務必/四個全面/五位一體), New-Era ideological vocabula…
prc-bureaucratese/SKILL.md on GitHub (opens in a new tab) -
psychoanalytic-lens
Rewrite an article through a psychoanalytic lens — events stay intact, but the narrative's interior (motives, meanings, narrator's asides) is re-narrated using Freudian / Lacanian vocabulary so that the article itself looks at its own subject through the analytic ear. This is a **lens** skill, not a style skill: the surface language need not change shape, but the *interpretive frame* does. Trigge…
psychoanalytic-lens/SKILL.md on GitHub (opens in a new tab) -
ptt-bbs-style
Transform any modern article into a 台灣 PTT 鄉民體 BBS post —— complete with a category-tagged title (例如 [問卦]/[新聞]/[爆卦]/[心得]/[討論] /[黑特]/[Re: 問卦]), a colloquial 鄉民 body that opens with 自介起手 式(肥宅我/弟弟我/妹妹我), drops in 鄉民詞 (484/笑死/ㄎㄎ/中肯/ zzz/RR/QQ/3Q/喇/啦/齁), and ends with a 求支援/求戰/文長慎入 closing line, followed by a stack of 5–15 simulated 推文/噓文/→ replies written by anonymized fake IDs (rookie123/fatdog2024/…
ptt-bbs-style/SKILL.md on GitHub (opens in a new tab) -
qiongyao-style
Transform any article into a Mid-Era Qiongyao (中期瓊瑤,《在水一方》 《一簾幽夢》流派) romance — 300–800 character modern-vernacular romance with a mandatory love axis, a structural obstacle, high emotional density, optional poetry inlay (max 1–2 at peak moments), and an ending that preserves the ache. Use only for romance narrative — the grandiose framing, tricolons, and exclamations are signature features of the…
qiongyao-style/SKILL.md on GitHub (opens in a new tab) -
roc-bureaucratese-classic
Transform text into classical-era Republic of China bureaucratese (中華民國 1960s-80s 公文體) — the distinctive semi-classical administrative writing style of the ROC government before the gradual 白話化 of the 1990s. Features single-character classical vocabulary (之/於/以/係/為/俾/爰), archaic connectives (茲/惟/嗣/俟/頃/殆/甫), hierarchical honorific terms (鈞/貴/本/職/台端/鈞長), fixed procedural formulas (查照/鑒核/核備/照辦/奉悉/敬悉…
roc-bureaucratese-classic/SKILL.md on GitHub (opens in a new tab) -
shakespeare-style
Transform any article into Mid-20th-century Zhu-Shenghao-style Chinese translation of Shakespearean tragedy (朱生豪式中譯莎劇腔) — Traditional Chinese, 300–1000 characters, half-classical-half-vernacular voice with dense metaphors (natural imagery + abstract concepts), expressive exclamatory rhythm, and a tragic-conflict spine. Primarily targets the four great tragedies (Hamlet / King Lear / Othello / Mac…
shakespeare-style/SKILL.md on GitHub (opens in a new tab) -
shanghainese
Transform text into Shanghainese (上海話 / 滬語) — the Wu Chinese variety spoken in Shanghai, belonging to the Taihu subgroup (太湖片) of Wu (吳語). Features Wu pronouns (儂/阿拉/伊/伊拉), Wu negation (弗/勿/覅), sentence-final particles (呃/嘞/伐/哦呦), Wu adverbs (蠻/老/嘎), signature lexicon (結棍/霞氣/嗲/戇/洋盤/淘漿糊/軋鬧猛/噶三胡), cultural vocabulary (弄堂/石庫門/阿婆/爺叔/小囡), and 海派 register(派頭/有腔調). Three intensity levels from light Wu-f…
shanghainese/SKILL.md on GitHub (opens in a new tab) -
sherlock-holmes-style
Transform any article into 朱生豪式中譯維多利亞偵探小說腔——Traditional Chinese rendering of late-19th-century Conan Doyle detective fiction (Sherlock Holmes canon). 第一人稱華生回憶體開場 + 福爾摩斯演繹獨白 + 案件三幕(委託登門→現場勘察→演繹揭曉)+ 維多利亞物件詞彙(煤氣燈/懷錶/菸斗/馬車/哈德森太太/蘇格蘭場)+ 中譯翻譯腔為 feature。世界觀層走西洋通名(倫敦/貝克街/鄉間別莊/ 火車車廂),不可換為中華具體專名。風格參照《血字的研究》《四簽名》 《冒險史》《回憶錄》。Primarily user-invoked via "/sherlock-holmes-style"; also triggers on "改寫成福爾摩斯", "福爾…
sherlock-holmes-style/SKILL.md on GitHub (opens in a new tab)
No skill matches that filter.
Review before you install
A skill is a set of instructions and scripts your agent will follow, and it can run code on your machine. Read the SKILL.md and any scripts first — this listing is not a security review.